Arab Social Media - MBRSG

Social Media and the Internet of Things Towards Data-Driven Policymaking in the Arab World: Potential, Limits and Concerns

Arab Social Media Report 2017

7th Edition

Social Media and the Internet of Things Towards Data-Driven Policymaking in the Arab World: Potential, Limits and Concerns

The Arab Social Media Report series was initiated by the Governance and Innovation Program at the Mohammed Bin Rashid School of Government (formerly the Dubai School of Government) in 2011. It aims to enhance regional understanding of the impact of social media on societies, development and governance in the Arab region. During the past years, the findings of the series have contributed to global policy and scholarly discourses. They have informed decision making across multiple disciplines by providing regional data and in-depth analysis on the role of social media in development and digital transformations in the Arab region. The key objective of the report is to explore the following questions: • How are societies and governments using social media across the Arab region? • What are the demographic, gender and language usage trends? • What factors affect the adoption of social media in the Arab states? • What is the impact of the phenomena on public engagement, participation and social inclusion? • How is social media impacting governance and public policy in the region?

This Report was Authored by: Fadi Salem Mohammed Bin Rashid School of Government

You can re-use any material included in this report as long as proper citation and referencing are applied. To cite this publication please use the following format:

Salem, F. (2017). The Arab Social Media Report 2017: Social Media and the Internet of Things: Towards Data-Driven Policymaking in the Arab World (Vol. 7). Dubai: MBR School of Government.

The views expressed in this report are those of the author and do not necessarily reflect those of the trustees, officers and other staff of the Mohammed Bin Rashid School of Government (MBRSG) and its associated entities and initiatives. For questions or media enquiries please direct emails to the authors at: [email protected] or [email protected]. To follow the author on Twitter: @FadiSalem

To access additional datasets and charts, not published in this report, join the Arab Social Media Report (ASMR) community and register (at no cost) online at: www.ArabSocialMediaReport.com Non-registered members can download the full report, and follow ASMR social networking groups through the website.

Introduction

The power and influence of social media has continued to grow globally over the past decade. During this period, the scope and scale of usage have changed dramatically, from the early days when social media was primarily fulfilling social needs of connectivity, communication, socialization and entertainment, to the current era, where social media applications are seen as important tools of governing, development, diplomacy and business. More recently, informational flows taking place through social media have been informing—and misinforming—public opinion and influencing policy development and political communication. For example, in 2015 and 2016, social media played a critical role in shaping public opinion internationally and on national levels, during numerous major events with global implications. During the past few years, as applications of “big data” and artificial intelligence continued to mature, the level of sophistication and influence of social media expanded further. As societal penetration rates increased, these data-driven applications started to provide deep insights into public views, sentiments, needs, behaviors and activities in numerous countries at unprecedented granular levels. The newfound insights harnessed through social media created new opportunities, as well as new risks. During the past year, social media also played a highly influential role in what has been described as a “post truth” era in policymaking, diplomacy and political communication1. For example, social media “bots” arguably played a key role in influencing public opinion globally, whether on the political or public policy levels. Such bots rely heavily on big data analytics, artificial intelligence and machine learning algorithms, not just in gathering and crunching public views and sentiments, but more so in pro-actively influencing public opinions, decisions and behaviors. These rising social media activities were coupled with a flood of “fake news” techniques, that may have played a role in manipulating public perceptions and beliefs at large scales. Facebook was a primary example during 2016, where highly influential “fake news” flourished and penetrated public opinion in the US and Europe, which arguably had influential outcomes on global policy and the world order. By early 2017, Twitter has also been utilized as an influential tool of governing in the highest public office in the US and beyond. These uses undermined traditional information mediums, triggered foreign policy crises, impacted political communication and disrupted established policy formulation cycles.

11.

By the end of 2016, the term “post-truth” (or “post-factual”) became mainstream in many languages. For example, the Oxford University Press Dictionaries, selected the term as the “word of the year” that year. “Post-truth” refers to circumstances in which objective facts are less influential in shaping public opinion than perceptions and appeals to emotion and personal belief. Social media was seen as the primary medium enabling this phenomenon to flourish and expand.

Social Media and the Internet of Things - Towards Data-Driven Policymaking in the Arab World: Potential, Limits and Concerns

7

On the other hand, the digital revolution has expanded the horizon of possibilities for development, governance and policymaking. The ever-expanding Internet connectivity of people and objects, and the maturity of big data and artificial intelligence applications are together ushering what has been described as the “fourth industrial revolution”. This new disruptive transformation is characterized by a fusion of inter-connected technologies where the digital, physical and biological worlds converge. This inter-connectivity is generating—and consuming—an enormous amount of data that is changing the ways policies are conducted, decisions are taken and day-to-day operations are carried out. Within this context, ‘big data’ applications are increasingly becoming critical elements of policymaking. Coupled with the rise of a critical mass of social media users globally, this ubiquitous connectivity and data revolution is promising major transformations in modes of governance, policymaking and citizen-government interaction. In the Arab region, there are already numerous experiments and applications where data from social media and the “Internet of Things” (IoT) are informing and influencing government practices as sources of big data, effectively changing how societies and governments interact2. These two sources, among others, are already influencing policymaking, public engagement and other means of citizen-government interactions in the region. The sources of big data are numerous. These include, open government data, financial transactions data, geo-spatial data, societal “lifestyle” data, social media data, data generated by Internetenabled devices (IoT), sensors data, machine-to-machine data, data from security systems, among many other sources. In this edition of the Arab Social Media Report series, we explore the potential limitations and concerns related to this phenomenon by focusing on two key sources of big data: social media and IoT data. Observations and anecdotal evidence from public sector and decisionmaking organization in the region suggest that there is limited understanding of the real potential, the limitations, and the public concerns surrounding these big data sources in the Arab region. This report contextualizes the findings in light of the socio-technical transformations taking place in the Arab region, by exploring the growth of social media and building on past editions in the series. The objective is to explore and assess multiple aspects of the ongoing digital transformation in the Arab world and highlight some of the policy implications on a regional level. More specifically, the report aims to better inform our understanding of the convergence of social media and IoT data as sources of big data and their potential impact on policymaking and governance in the region. By using a triangulation of research approaches it digs into several critical dimensions of this phenomenon. Ultimately, in light of the availability of massive amount of data from physical objects and people, the questions tackled in the research are: What is the potential for data-driven policymaking and governance in the region? What are the limitations? And most importantly, what are the public concerns that need to be addressed by policymakers while they embark on next phase of the digital governance transformation in the region? The report has two main parts. In the first part, we explore the questions discussed in the previous paragraphs through a regional survey spanning the 22 Arab countries. In the second part we continue the tradition set in the previous editions of the Arab Social Media Report series by exploring the growth and usage trends of influential social media platforms across the region, including Facebook, Twitter, Linkedin and, for the first time, Instagram. The findings highlight important changes—and some stagnation—in the ways social media is infiltrating demographic layers in Arab societies, be it gender, age and language. Together, the findings provide important insights for guiding policymakers, business leaders and development efforts. More specifically, these findings can contribute to shaping directions and informing decisions on the future of governance and development in the Arab region.

22.

Definitions of “social media”, “IoT” and “big data” as the key terms discussed in the paper, are included in Annex 3

8


1. Social Media and IoT for DataDriven Policymaking: Potential, Limits and Concerns of Big Data and Governance in the Arab Region Based on a regional survey, this section presents some of the key findings on the uses of social media data, IoT data and their convergence as a source of big data. It also explores public views and perceptions on the potential role of this data in influencing policymaking. The findings primarily highlight the public concerns on personal data use by governments and businesses, the limitations facing data-driven policymaking and potential policy responses and actions to tackle these concerns and limitations.

1.1. Social Media Data and Public Policy Despite the many promises, the value of social media in public policy in the region has numerous limitations. The section provides key findings of the survey related to these limitations.

•

False Personal Information: 15% of respondents said that they provide false information on social media.

•

Closed Social Media Accounts: Around 44% of respondents in our survey said that they have closed down or abandoned at least one social media account in 2016.

•

Location services:

•

•

A third of respondents actively disable location-based services when using all social media platforms.

•

Around 40% allow location services only on some social media platforms.

•

One in five respondents said that they intentionally allow location services on all social media platforms they use.

Multiple Accounts: Across the Arab region, 46% of respondents said that they own multiple accounts in at least one social media platform.


9

Social Media for Data-Driven Policymaking – Exploring the Limitations 1) Flawed Representation: Even when social media penetration reaches 90% of the population in some countries, social media data may still provide flawed representation of societal breakdowns. This is especially true if used in isolation in big data analysis as a source of representative voice online. One example is the common regional behavior among social media users of owning multiple accounts on a single social media platform with diverse usage trends in terms of sharing sentiments or opinions. For example, in our survey, 46% of respondents said that they own multiple accounts in at least one social media platform. These users may have different “voices” or varied representation of their views on each of these accounts. The reasons behind this behavior may be related to political and cultural restrictions or the contrasting objectives that may exist for people using social media platforms. In terms of platforms, more than 34% of respondents said that they own more than one Facebook account (representing 73% of those who said that they own more than one social media account on a single platform). Other social media platforms have considerably less ownership of multiple accounts according to our survey respondents. Around 18% said that they own more than one account on Google Plus, 10% on Instagram and Twitter each, and just above 5% on LinkedIn. While this behavior differs from one country to another, this is one example of how social media demographic data can be misleading or non-representative.

More than One Account: More ThanWhich One Account: Which Social Media Platform You than Own One More Account than One in? Account in? Social Media Platform Do You OwnDo More

Do you own multiple acounts (more than one) in any social media platform?

34.1%

Yes 46%

No 54%

18.0%

10.7%

10.0% 5.6% 1.6%

Facebook

Google+

Instagram

Twitter

LinkedIn

Others

Twitter

trest


gle+

kedin

utube

10

tagram

cebook

2) False Information: One of the key sources of insight data derived from social media is the ability to segment people’s views, behaviors and sentiments based on their demographic information Social Media monthly active users (survey treemap) shared over social media. Such data includes age, gender, language and other demographic and behavioral data that is usually directly shared by social media users or can be easily extracted. However, in the age of mass online surveillance, “fake news” and large-scale privacy breaches, it is no secret that many social media users provide false information when using social media platforms. Moreover, the number of fake accounts, or even machine-operated accounts, or “bots”, is difficult to assess but is arguably large.

Social Media monthly active users (survey treemap)

Linkedin

Youtube

Facebook

Instagram

Twitter

One important question for social media-driven policymaking here is: how much false information is out there? And what personal information shared online is false? According to our survey, around 15% of respondents in the Arab region openly said that they provide false personal information on all or some social media platforms (10% on some platforms and 5% on all platforms). Meanwhile, around 76% of users said that they always provide accurate personal information when using social media platforms.

Pintrest

Google+

What false personal information do people provide? According to the survey respondents who said that they do provide false personal information on social media, the person’s “name” is the top false piece of information provided in the Arab region. Around 61% of those respondents said that they provide a false name (9% of all social media users), followed by 40% who said that they provide false birthday or age (6% of all respondents). Around 35% said that they provide false location, such as the city or country (5% of total), 17% said that they provide false social status, such as if married or single (2.6% of total) and 13% said they provide false gender information (less than 2% of total social media users in our survey).

False Information: What False Information do you Usually Provide on Social Media? (out of (out of respondents who said What False Information do you Usually who Provide Social Media? respondents saidonthey provide false they provide false information) information)

When using social media, do you provide false personal information (gender, name, age, etc.)? Prefer not to say, 9.24% FALSE info on SOME social media platforms, 9.58%

60.9% FALSE info on ALL social media platforms, 5.14%

ACCURATE info on ALL my social media platforms, 76.04%

39.8% 35.1%

17.3% 13.0% 4.9%

Name

Age or birthday

Location, city or country

Social status (e.g. married, single)

Gender

Other personal info

On which social media platforms do you provide false personal information?


11

On what social media platforms do people in the region usually provide false information? Around 40% of those who said that they provide false personal information when using social media are doing so on multi-purpose social media platforms such as Facebook and Google Plus. Around 16% said they provide false personal information on Twitter and other microblogging platforms, 11% on photo-sharing platforms such as Instagram, and another 11% on video-sharing platforms such as Youtube.



40.4%

16.0%

Facebook, Google+ and other Multi-purpose platforms

Twitter and other Microblogging platforms

15.5% 10.8%

10.7%

Instagram, Pintrest and other Phototsharing platforms

Youtube, Vine and other Video sharing platforms

Other social media platforms

Such false personal information is usually fed into demographic analysis of social media data within electronic participation initiatives or big data-driven policymaking cycles. Accounting for and correcting for such misrepresentative demographic data is critical when social media data informs policy formulation cycles.

Users abandoned) Social Media Accounts During the Past YearAs (of out Total Survey One important noteWho to beClosed made (or here, is that the data presented by the survey is self-reported. such, it would be wise to consider the possibility of different levels of social desirability biases in responses provided in some countries or by some segments in society. For example, it is difficult to tell exactly how many of the respondents who said that they always provide accurate personal 22.4% on social media (76% of respondents) are being truthful and how many are trying to be information Did you close down or stop using any of your politically or socially correct in their answers. Moreover, social some—or arguably most—of the 10% of media accounts during the past year? respondents who said that they prefer not to answer the question may very well opted to not give an answer because they do provide false personal information online, but prefer not to say this openly for obvious reasons. Yes No 3) Validity of Over-Time Data: In addition to the risk of under-representation (or over-representation) 44% 56% of societal voices, sentiments and opinions, actual possession and levels of activity of social 11.9% media accounts are fluid and vary with time. People do change behaviors and activity, “migrate” to alternative platforms and even close or abandon social media accounts over time. When extracting longitudinal or panel data about certain segments in society at different points of time through social 8.4% media platforms, considerable amount of 7.5% noise can be introduced in the analysis. For example, no less than 44% of respondents around the region said that they have closed down or abandoned at least one social media account in 2016. More specifically, around 22% of respondents said that they 3.9%year, 12% closed or abandoned a Twitter closed or abandoned a Facebook account during that account. Other accounts people said they stopped using include: 2.7% Instagram (8%), Google Plus (8%) 1.7% and LinkedIn (4%). Around half of those who said that they have closed or abandoned at1.5% least one of their social media account, said that they closed a Facebook account. Almost 1 out of 4 said that they closed or abandoned a Twitter account, 1 out of 5 an Instagram account, and 1 in 10 stopped Facebook Twitter Instagram Google+ LinkedIn Youtube Pintrest Other using a LinkedIn account.

12


Users Who Closed (or abandoned) Social Media Accounts During the Past Year (of out Total

Users Who Survey Respondents)

Closed (or abandoned) Social Media Accounts During the Past Year (of out Total S

22.4% Did you close down or stop using any of your social media accounts during the past year?

Yes 44%

No 56%

11.9%

8.4% 7.5%

3.9% 2.7%

Facebook

Twitter

Instagram

Google+

LinkedIn

Youtube

1.7%

1.5%

Pintrest

Other

4) Geolocation: Much of big data-driven analyses, especially for development and policymaking, rely on location as a critical attribute. For example, geo-locating social media users is important when Do weight you usually allow location-based services (GPS other deciding on the that should be given to their views or sentiments withand regards to government services in certain areas, their proximity certain using events social and if they are living in the country, or one location services)to when media apps? of its regions at all. Identifying geolocations of users is also as important for big data-driven decisions made in the private sector, such as customer service, investment options and advertisement decisions, among many other uses. However, locations on social media can be intentionally or unintentionally Not Aware misguiding. For example, according to our survey a third of social media users in the Arab region 8.4% intentionally and actively disable location-based services when using any social media platforms. ENABLE on ALL Meanwhile, around 40% said that they consciously allow location services 21.2% only when using some social media platforms while disabling location services on others. A fifth of respondents in the Arab region said that they intentionally allow their location to be identified, by enabling location-related services on all platforms. DISABLE on ALL 31.9%

Overall, the findings of the survey suggest the existence of a healthy level of awareness about the availability and ability to control location-services when using social media in the region. Only 9% of respondents said that they are unaware of the location services functions or how to enable or disable them on social media. on SOME Additionally, the type of social media services seems to affect Enable the38.5% behaviors and preferences of users when deciding on sharing their location. Expectedly, of those who said that they intentionally disclose their location when using social media platforms or allow their location to be shared, the majority enables this on location-driven or maps-based social media platforms, such as Foursquare. Meanwhile, only 30% said that they allow location services on multi-purpose social media platforms such as on Facebook or Google Plus. Around 12% of respondents said that they allow location to be shared when using photo-sharing apps, such as Instagram or Pinterest; and only 4% when using microblogging services such as Twitter.

How Do You Usually Express Your Sentiments or Views When Using Social Media (e.g. satisfaction, happiness, disapproval, anger, etc.)?


13

When using social media (on mobile or other devices), on which platforms do you allow location-based services (e.g. GPS)? (out of respondents who said they Enable location services)

Do you usually allow location-based services (GPS and other location services) when using social media apps? Not Aware 8.4% ENABLE on ALL 21.2%

DISABLE on ALL 31.9%

43.1%

Enable on SOME 38.5%

30.9%

12.0%

Map-based Apps (Foursquare, etc.)

Multi-purpose social media apps (Facebook, Google+)

Photo sharing apps (Instagram, Pintrest,etc.)

4.2%

4.1%

4.6%

Microblogging apps (e.g. Twitter)

Social gaming apps (e.g. Pokémon Go)

Other social media apps or platforms

Finally, while public or private institutions and authorities can still identify and locate users through numerous complex methods, the users’ behaviors highlighted in our findings add additional noise to the data and provide multiple layers of complexity, regardless of the methods used. Creating robust and segmented analysis on where public views or sentiments are coming from is important when analyzing social media data. This helps avoid voices becoming amplified or underrepresented among different segments of society. As such, it is critical to adjust and correct for such noise in algorithms or analyses that rely on social media as a big data source in decision-making or public policy design. Personal IoT Devices: On your smatphone, do you use apps that connect with any physical devices online (IoT devices)?

Sentiments and Views on Public Policies and Social Media – Trends and Limitations According to the findings of the Arab Social Media Report series, people in the Arab region have 24.5% Yes 23.7% adopted social media applications in all aspects of 47% their lives, including interacting with their No do people in the governments regarding public services3. However, beyond government services, 53% region use social media to express their views and opinions on public policy issues? 12.7%

According to our survey, a majority of social media users in the Arab region (58%) said that they express their views or sentiments regarding their government’s policies using social media. This 6.5% 6.0% includes for example, expressing their satisfaction, happiness,3.6% anger and disapproval textually or 3.5% visually. Around 86% of those who use social media to express views on government policies, said 0.6% that they use Facebook to do so. Only 28% said that they use Twitter, and another 28% use messaging apps, suchTV:asApps WhatsApp onAppgovernment than 10% respondents Cameras:to express Video gamesview Cars: Wearable policies. Physical Less Home Other of Devices: to each Appsof the console: Appscategories connecting to of devices: Apps media security electrical AppsGoogle Plus, said that connecting they use other social platforms, such as Internetconnecting to connecting to Internetconnecting to systems: Apps appliances: connecting to enabled TV camerasor (e.g.video-sharing gaming enabled Apps other InternetLinkedin, photo-sharing sites.car watches, sport connecting to 33.

14

webcam, consoles bracelets, security connecting to enabled surveillance, healthcare systems (e.g. thermostats, physical products, door locks, devices The Arab Social Media Report baby (Vol. 6), Citizen Engagement and Public Services in the Arab World: The Potential A/Cs, of Social Media. Dubai: Governance and monitoring) smart glasses, garage refrigerators, Innovation Program, Governance and Innovation Program, MBR School of Government – Available at www.ArabSocialMediaReport.com etc. systems, etc.) light bulbs, washing machines, etc.


8.4%

7.5% Do You Use Social Media to Express Your Sentiments or Views on Government Policies or Services?

Do You Use Social Media to Express Your Sentiments 3.9%or Views on Government Policies or 2.7% Services?

Facebook

Twitter

Prefer not to say 6%

Instagram

Google+

LinkedIn

Youtube

1.7%

1.5%

Pintrest

Other

No 36%

Yes Do you usually allow location-based services (GPS and other 58% location services) when using social media apps?

Not Aware 8.4% ENABLE onin ALL However, how do people usually express their sentiments on social media the region in general? 21.2% A large majority of our survey respondents (76%) said that they do so by clicking on corresponding buttons when available, such as “Like”, “Dislike” or “Favorite”. Around 50% said that they express views or sentiments on social media using “emoticons”, “emojis” or “smileys”. Across the region, When Expressing Your Views on Governmentusing Policies or Services, What Social Media DISABLE on ALL around 30% express their views by typing expressions Arabic text and characters, while another 31.9% Platforms do You Usually 26% do so using Latin characters, primarily in the English or Use? French languages. Internet “slangs” or “lolspeak” come next in the ranks of how people prefer to express their views on social media in the region (23%). Meanwhile, around 15% said that they combine Arabic and English text while they express their views on social media. Only less than 10% of respondents said that they use “hashtags” Facebook as a way to express sentiments on social media. Transliteration of Arabic text using Latin85.8% characters, Enable on SOME a formerly common practice during earlier days of Internet development, Twitter 27.8% 38.5% is still preferred by around 8% of respondentsMessaging in expressing their sentiments on social media. These expression styles on social Apps (Whatsapp, Snapchat, etc.) 27.5% media in the region, especially in light of the diverse linguistic usage trends online, may increase the Photo-sharing platforms (e.g. Instagram, Flickr) 9.9 % complexity of efforts to harness representative samples of voices or views online. On the technical Youtube, Vine) in the region, 9.9 % level, the diverse styles Video-sharing of online(e.g. expression which vary from region and country and even city to another, also make relying Google+ on big data8.6algorithms and methods in sentiment analysis % more complex and less reliable. Linkedin, Bayt or other professional social media platforms

8.4 %

Blogs (write or comment government or services) 6.9 % How DoonYou Usuallypolicies Express Your Sentiments or

Views When Using Social Media

How Do You Usually Express Sentiments Views When Using Social Media (e.g. (e.g.Your satisfaction, happiness,or disapproval, anger, etc.)? Audio-sharing platforms (e.g. SoundCloud) 1.8 % satisfaction, happiness, disapproval, anger, etc.)? Other social media platforms

1.3 %

75.8%

49.7% 29.1%

25.9%

When Unhappy or Unsatisfied23.0% with a Government Policy, How do You 15.5% 10.2% 8.5% Usually Express Your Views on Social Media? 2.4% Click on Emoticons, Arabic text buttons [e.g. emoji or [e.g. “smileys”, “‫]”رﻫﻴﺐ“ ”راﺋﻊ‬ “Like” or I openly express my opinion directly “Favorite”] [e.g. :) or :-D]

English/ French text [e.g. “cool”, “sad”]

Internet slangs or “lolspeak” [e.g. LOL, 5air, ‫ﻫﻬﻬﻪ‬, ISA

Mix Arabic and other language 21% [e.g. “ ‫ﺣﻠﻮ‬ lol”]

Hashtags Transliterations Other ways [e.g.#fail] of Arabic in Latin characters (e.g. gameel) 29%

I self-censor my comments and expressions

I use sarcasm in my comments

4%


Do You Use Social Media to Express Your Sentiments or

I indirectly imply my opinion

15

When Expressing Your Views on Government Policies or Services, What Social Media

When Expressing Your Views on Government or Use? Services, What Social Media Platforms doPolicies You Usually Platforms do You Usually Use?

Facebook

85.8%

Twitter

27.8%

Messaging Apps (Whatsapp, Snapchat, etc.)

27.5%

Photo-sharing platforms (e.g. Instagram, Flickr)

9.9 %

Video-sharing (e.g. Youtube, Vine)

9.9 %

Google+

8.6 %

Linkedin, Bayt or other professional social media platforms

8.4 %

Blogs (write or comment on government policies or services)

6.9 %

Audio-sharing platforms (e.g. SoundCloud)

1.8 %

Other social media platforms

1.3 %

These findings shed light on people’s general preferences while expressing their sentiments and views on social WhenHowever, Unhappyaor Unsatisfied with a Government How do You media in the region. much more important HowPolicy, open are people in the question arises when considering online social media region when they express Usually Express Your Views on Social Media? interactions as input into policymaking: How open are their negative views on social people in the region when they express their negative media, specifically when views on social media, specifically when unsatisfied or unsatisfied or unhappy with express my opinion directly unhappy withI openly government policies? In our survey, only 21% government policies? 29% 29% of respondents said that they openly and directly express their I views social and media when unhappy or self-censoron my comments expressions unsatisfied with government policies. Meanwhile, around a quarter of respondents said they selfcensor their comments, expressions or opinions on social media in such cases. Around 11% said that I use sarcasm in my comments 4% they use “sarcasm” when expressing their negative views on social media about government policies or services, and another 11% said that they try to indirectly imply their views. I indirectly imply my opinion

11% These online behaviors are problematic when relying on social media as a big data source for Other ways analyzing public sentiments and opinions in the region. In light of these regional findings, the reliability of sentiment analysis algorithms across wide parts in the region should be questioned. Regardless of how sophisticated sentiment and behavior analysis algorithms are, the capability of 24%such algorithms I prefer not to say 11% to peel away diverse cultural and linguistic layers and extract meaning from social media interactions in the region should be examined thoroughly, especially in Regardless of how policymaking circles. For example, it is difficult to imagine sophisticated sentiment and extracting meaningful outcomes from a tweet or a post behavior analysis algorithms when a social media user in the Arab region expresses are, the capability of such a view indirectly using a sarcastic tone, through a algorithms to peel away combination or Arabic and Latin characters in a local diverse cultural and linguistic dialect, combined with an Internet slang and a bunch of layers and extract meaning Views on Governments' Practices on Social Media emojis. While some algorithms’ accuracy may increase from social media interactions if they are fine-tuned and adjusted to behaviors of small in the region should be populations, until textual big data analytics mature with examined thoroughly, Bots: Trying to influence your views by using social media “chatbots” or 53% 14%in policymaking 22% 9% 3% advanced contextual abilities, the findings on these especially paid users of social media behaviors suggest limited location potential of using textual big circles. Geo-tracking: Tracking geographic and movement using social 6% 40% 20% 20% 15% data media analytics across the social media realm for policy Linked Data: Linking social media data with other sources of personal 40% 18% 24% 13% 5% data (e.g. phone development indata) the region.

16

Behaviors Analysis: Tracking personal behaviors on social media

36%

Prediction: Predict your future behaviors or activities based on your social media data

36%

Frequency: Tracking frequency of your use of social media Sentiment Analysis: Analyzing “sentiments” on social media (e.g. if

30%

18% 16% 16%

24% 27%

16%

6%

15%

6%

17%

5%


31%

When Unhappy or Unsatisfied with a Government Policy, How do You

When Unhappy or Unsatisfied with a Government Policy, How do You Usually Express Your Usually Express Your Views on Social Media? Views on Social Media?

I openly express my opinion directly

21% 29%

I self-censor my comments and expressions

I use sarcasm in my comments

4%

I indirectly imply my opinion 11% Other ways

24%

I prefer not to say

11%

Public Concerns on Existing Government Practices on Social Media Governments around the world are proactively experimenting with social media data and interacting with social media users in a variety of ways. The spectrum of uses varies from benign and positive uses, to manipulative and Views aggressive ones. For example, some governments may rely on social on Governments' Practices on Social Media media data for trend and sentiment analysis, feeding into policymaking with the objective of aligning policy development with public needs, concerns and preferences. Others capture and store massive amounts of activities and personal data on social media users for profiling and behaviors analysis. Tryingcommon to influence your views by usingin social media “chatbots” One Bots: of the practices recent years,oris moving beyond listening14% and analyzing; 53% 22% 9% instead 3% paid users of social media governments are utilizing artificial intelligence and machine learning through software “bots” with Geo-tracking: Tracking geographic location and movement using social 40% 20% 20% 15% 6% media the objective of influencing public perceptions and beliefs, and direct policy outcomes. Similarly, Linked Data: Linking social media data with other sources of personal 40% 18% 24% 13% 5% in another growing data (e.g. phone data) practice, some authorities recruit social media users, whether widely followed Behaviors Analysis: Tracking personal behaviors on social mediausers for the same purposes. We surveyed public views “influencers” or large masses of unidentified 36% 18% 24% 16% 6% around the region regarding these practices by presenting them with brief descriptions and related Prediction: Predict your future behaviors or activities based on your 36% 16% 27% 15% 6% social media data scenarios. Frequency: Tracking frequency of your use of social media

30%

16%

31%

17%

5%

Trends analysis, where “sentiments” generalontrends are by29% analyzing public data available on social Sentiment Analysis: Analyzing social media (e.g.extracted if 15% 30% 18% 7% you’re happy, unsatisfied, angry, etc.) media, was the only big data-driven government practice supported by a majority of respondents. Social Network Analysis: Identify social groups you belong to in society 29% 16% 30% 18% 7% Around 51% percent either support through analyzing networks of people you interact withthis practice without concerns (17%) or support it with some Profiling:(36%). ClassifyingMeanwhile, your profile based a onmajority social media of behaviors concerns respondents expressed their rejection all other 27% 15% 32% in general 18% to8% practices presented. Other government practices examined are listed from the least to most in terms Public Data: Capturing and storing public social media data 25% 24% 23% 22% 5% of people’s disapproval: Trend Anslysis: Analyzing general trends based on publicly available social media data

9%

12%

1. Capturing and storing public data from social media.

26%

2. Profiling and classifying individuals based on social media behavior. I strongly reject it

I reject it but I see some value

36%

17%

Neutral

I support it but I have concerns I strongly it without concerns 3. Identifying social groups thesupport individual belongs to through analyzing their social networks online.

4. Analyzing social media users’ sentiment, such as their level of satisfaction, disapproval, etc. 5. Tracking the frequency of social media use. 6. Predicting future activities based on social media data. If the use of your personal and socialactions media data by governments leads to any of the following 7. Tracking personal behaviors of social media users. beneficial outcomes, would you support it?

8. Linking social media data with other sources of personal data.

Social Media and the Internet of Things - Towards Data-Driven Policymaking in the Arab World: Potential, Limits and Concerns Better Policies (e.g. better policies on education, healthcare, transport, etc.)

5.6% 7.5%

18.0%

34.7%

17 34.2%

When Unhappy or Unsatisfied with a Government Policy, How do You Usually Express Your Views on Social Media? 9. Tracking location or movement of social media users. 10. Using social media bots to try to influence public perceptions, views, beliefs or sentiments. I openly express my opinion directly 21% 29% Other than general trend analysis using social media data, respondents who said that they support I self-censor my comments and expressions any of these practices without concerns did not exceed 8% for each. Government attempts to influence public Government attempts to I use sarcasm “bots” in my comments views by using social or recruited users on social 4% influence public views media received the highest percentage of negative by using social “bots” or views by our I indirectly surveyimply respondents (53% strongly reject my opinion recruited users on social this use, and 14% reject it while seeing some value). media received the highest 11% In general, around 20 to 30 percent of respondents Other ways percentage of negative views said that they have no opinion on these practices. This by our survey respondents response may be reflective of lack of full understanding 24% I prefer not to say 11% of these practices, or because the respondents in this case intentionally preferred not to comment and held back their opinions on these practices. Overall, taken collectively, the findings strongly suggest that the public across the Arab region is generally concerned about the use of personal social media data by governments, and the activities some government take on social media.

Views on Governments’ Practices on Social Media Views on Governments' Practices on Social Media

Bots: Trying to influence your views by using social media “chatbots” or paid users of social media

53%

Geo-tracking: Tracking geographic location and movement using social media Linked Data: Linking social media data with other sources of personal data (e.g. phone data)

14%

22%

40%

20%

20%

40%

18%

24%

Behaviors Analysis: Tracking personal behaviors on social media

36%

Prediction: Predict your future behaviors or activities based on your social media data

36%

18%

24%

16%

27%

9% 3% 15%

6%

13%

5%

16%

6%

15%

6%

17%

5%

Frequency: Tracking frequency of your use of social media

30%

Sentiment Analysis: Analyzing “sentiments” on social media (e.g. if you’re happy, unsatisfied, angry, etc.)

29%

15%

30%

18%

7%

Social Network Analysis: Identify social groups you belong to in society through analyzing networks of people you interact with

29%

16%

30%

18%

7%

18%

8%

Profiling: Classifying your profile based on social media behaviors

27%

Public Data: Capturing and storing public social media data

25%

Trend Anslysis: Analyzing general trends based on publicly available social media data

I strongly reject it

9%

12%

16%

15% 24%

31%

32% 23%

26%


22%

36%

5% 17%

Neutral

I support it but I have concerns

I strongly support it without concerns

However, despite the clear negative views about using personal social media data by governments, our survey respondents’ views changed if the usage of personal social media data by the government If the use of your personal social media by governments leads to any of the following led to positive policy-related outcomes. Whendata asked about their level of support for using personal beneficial outcomes, would you support it? social media data in some scenarios, a majority of respondents supported government use of their social media data if it leads to positive policies. Around 34% of respondents supported, without concern, using personal social media data by government, if this leads to better education, healthcare or transport policies. Another 35% also supported this, but with some concerns. In total, 69% of respondents support the use of their social media data if this leads for example, to better policies. 18

Better Policies (e.g. better policies on education, healthcare, transport, etc.)

5.6% 7.5%

18.0%

34.7%

34.2%


Trend Anslysis: Analyzing general trends based on publicly available social media data


9%

12%

26%


36%

17%

Neutral

It is noteworthy that the level of support for using social media data in policymaking among our survey respondents is slightly higher than concerns their support for using personal I support it but I have concerns social media data for I strongly support it without government service delivery. On the other hand, only 13% rejected using personal social media data by government, even if it leads to better policies.

If the use ofIfyour personal socialsocial media data bybygovernments leads toofany of the following the use of your personal media data governments leads to any the following beneficial outcomes, would you support it? would you support it? beneficial outcomes,

Better Policies (e.g. better policies on education, healthcare, transport, etc.)


5.6% 7.5%

18.0%

34.7%


I Strongly Support It Without Concerns

34.2%

Neutral


What Policies? Data-driven Public Policy Formulation by Governments Survey respondents were presented with some practices where social media data interact with public policies in different government sectors. The scenarios presented were suggested based on existing practices of using social media in data-driven policy formulation globally, whether such practices existed in the region or not. Education Policy: Tracking public behaviors or sentiments on social media to develop better educational policies (e.g. students’ activities, learning behaviors, school attendance, language use, academic performance, etc.)

5% 6%

19%

33%

36%

In general, there is a wide support for using social media in all public policy development scenarios Tourism: Track social media datato by shoppers and touristsof related to hospitality, presented. Around 60 70 percent respondents in4% 6% general 35%their social 21% supported 34% the use of tourism and entertainment to develop better tourism and marketing policies (e.g. on facilities, infrastructure, mediasentiments data and byopinions governments in mostetc.) categories of public policy formulation. Based on the scenarios presented, respondents supported themedia use oftosocial in the development cycle of38% the following Health Policy: Tracking and analyzing behaviors of social users develop media 5% 7% 19% 32% better health policies (e.g. sleeping habits, eating behaviors, diseases spread, etc.) policies respectively: health policies (70%), education policies (69%), tourism policies (69%), Transport Policy: Tracking geographic or location data through social media policies (64%). However, while most respondents transport policies (68%), and security and safety accounts to develop better transport policies (e.g. traffic congestions, crowd 6% 7% 19% 33% 35% movement, etc.)use of social media data in security and safety related policies, these types of policies supported the also received theApply largest shareof social of negative responses. Around 20% of respondents said that they Security and Safety: mass surveillance media data to develop better safety and security policies (e.g. mass-surveying, collecting and analyzing social 8% 10%related 18% 31%formulation. 33% rejectmedia thedata use oftotheir policy Across the aiming enhancesocial security). media data in security or safety board, out of those who supported the use of social media data in public policy formulation, almost Social Policy: Tracking discussions, analyzing sentiments and trends by 7% 8% 23% 35% 27% half ofcommunities them said theyto develop support but they have concerns. of socialthat media users better it social policies or social cohesion (e.g. youth, women, minorities, nationalities, special needs, etc.)

Meanwhile, scenarios where social media isusing used or economic policies—while still largely Economic Policy: Track financial or economic behaviors and trends social for social 8% 10% 24% 35% 24% media to develop better economic or financial policies. viewed positively—received less support by respondents. Around 59% of respondents said that they support the use of their social media data in developing economic policies, and 62% in social policies. I strongly reject it I reject it but I see some value Neutral I support it but I have concerns



Views on Private Sector Practices on Social Media

19

Government Policies and Social Media: In each of the following scenarios, how would you describe your views about the following uses of social media by government for public policy formulation?

Education Policy: Tracking public behaviors or sentiments on social media to develop better educational policies (e.g. students’ activities, learning behaviors, school attendance, language use, academic performance, etc.)

5% 6%

19%

33%

36%

Tourism: Track social media data by shoppers and tourists related to hospitality, tourism and entertainment to develop better tourism and marketing policies (e.g. sentiments and opinions on facilities, infrastructure, etc.)

4% 6%

21%

34%

35%

Health Policy: Tracking and analyzing behaviors of social media users to develop better health policies (e.g. sleeping habits, eating behaviors, diseases spread, etc.)

5% 7%

Transport Policy: Tracking geographic or location data through social media accounts to develop better transport policies (e.g. traffic congestions, crowd movement, etc.)

6% 7%

Security and Safety: Apply mass surveillance of social media data to develop better safety and security policies (e.g. mass-surveying, collecting and analyzing social media data aiming to enhance security).

8%

19%

19%

10%

Social Policy: Tracking discussions, analyzing sentiments and trends by communities of social media users to develop better social policies or social cohesion (e.g. youth, women, minorities, nationalities, special needs, etc.)

7% 8%

Economic Policy: Track financial or economic behaviors and trends using social media to develop better economic or financial policies.

8% 10%


32%

33%

18%

35%

31%

23%

33%

35%

24%

27%

35%



38%

24%

Neutral


Views on Private Sector Practices on Social Media Language: Identify your language using your social media data Employability: Assess employability, fit for a job or career potential based on your social media data Personalization: Suggest personalized products or services based on your social Private Sector Use of Social Media Data media data Sentiments: Assess your “sentiments” (how satisfied, dissatisfied, etc.) about

12% 16% 16%

39%

11% 13% 15%

23% 28 %

21%

17% 19%

30% 28%

13%

products or servicesrespondents were also presented with scenarios on the use of social media data by For comparison, Advertizing: Provide you with better advertizing based on your social media 14%private 25% 27% 15% 19% based the private sector and businesses. These scenarios were also on existing practices by data sector entities inbythe region related tohabits) using of social media data.32% Interests: Assessglobally your interestsor (e.g. capturing your internet surfing 10% 23% 15% 20% 18%

14%

29 %

26%

13%

Health: Assess the state of your healththe (e.g. largest eating habits, sleeping habits, etc.) 13% media 22% social 30%of personal 13% 22 % The practices that received support in terms of private sector use based on your social media data data were: Assessing employability and fit for the job of social media users (49% supported and 29% Location: Track your geographic location and movement using social media 5% 16% 22% 21% 36 % rejected), personalization of services and products based on social media data (41% supported and Status: Assess your financial status (e.g. credit scoring by banks, loans, 15%to 6% 22% 18% 38% uses: sentiment 21% Financial rejected). Public opinion was split regarding the following analysis assess etc.) using your social media data customer satisfaction with products and services (39% support and 32% reject), advertising based Commercial Use: Sell any of this information (based on your social media data) 9% 5% 17% 10% 59% to other companies, governments or individuals for commercial purposes on social media uses (39% support and 34% reject). Meanwhile, the largest share of negative views by respondents was expectedly related to theI reject commercial use I strongly reject it it but I see some valueof social media Neutral users’ personal data, where such data is sold or bought (69% rejected it with 59% strongly reject it). Assessing financial it but I have concerns I strongly negative support it without concerns status and location based onI support social media data also received views by the respondents.

20

Arab Social Media Report 2017 Which of the Following Would Make You Less Concerned About Use of Personal Data by Governments or Businesses through IoT Devices or Social Media?



Language: Identify your language using your social media data

12%

Employability: Assess employability, fit for a job or career potential based on your social media data

16%

Personalization: Suggest personalized products or services based on your social media data

16%

Sentiments: Assess your “sentiments” (how satisfied, dissatisfied, etc.) about products or services

18%

Advertizing: Provide you with better advertizing based on your social media data

19%

Interests: Assess your interests (e.g. by capturing your internet surfing habits)

20%

15%

Health: Assess the state of your health (e.g. eating habits, sleeping habits, etc.) based on your social media data

22 %

13%

Location: Track your geographic location and movement using social media Financial Status: Assess your financial status (e.g. credit scoring by banks, loans, etc.) using your social media data

15% 14% 15%

26%

13%

27%

25%

14% 10%

23%

32%

13%

22%

30% 22%

16%

5%

22%

15%

6%

10%

59%



29 %

18%

38%

13%

28%

28 %

21%

36 %

17% 19%

30%

23%

13%

Commercial Use: Sell any of this information (based on your social media data) to other companies, governments or individuals for commercial purposes I strongly reject it

21%

39%

11%

17%

9% 5%

Neutral


1.2. IoT Data and Public Policy In addition to its numerous other promises, the Internet of Things (IoT) can also be a rich source for data for policy and decision-making. In the region, the rapid proliferation of devices and sensors that Which of the Following Would Make You Less Concerned About Use of Personal Data by Governments or Businesses connect to the Internet are a key source ofthrough dataIoTthat public and private entities are exploring, both Devices or Social Media? for business and public policy objectives. Smart mobile phones, wearable devices, internet-enabled entertainment systems, home appliances, connected vehicles and sensors are just a few elements of the massive network of interconnected devices that generate enormous amount of rich data. The 70% promises are huge, but so are the limitations, risks and public concerns. 60% 50% 40% 30%

•

Human-IoT Connectivity: Half of the respondents in the Arab region (47%) have apps installed on their smart phones that interact with physical IoT devices

20% 10%

0%

•

Around 55% of in theCode region have 5 Internet-enabled IoT Penetration: Nonto If You Can2 to Ability of Data Laws Priorhouseholds Consent Ability to Knowing Knowing Commercial Approve Conduct (AUP, Download and Remove your on What May Who Uses What Data is devices and another 6 Done to 10 devices on our Usessurvey. Only Sectors that All Your Data based Fair Use) Regulations be withinternet-connected Data have from the Data 25% Collected Repositories

Your Data

Availablity

Availability

May Use your Data

•

Most Common IoT Devices: Around 65% of respondents said that they own an Internetenabled device that classifies asSocial anMedia entertainment IoT device, with Internet-enabled TVs being the largest category (20%).

•

Wearable IoT Devices: Only 8% own wearable devices such as smart watches, wristbands, etc. that connect to the Internet

•

IoT Appliances: Less than 7% own physical appliances that connect to the Internet, such as security systems, smart meters and home electrical appliances

•

Connected Vehicles: Internet-connected are You owned other Internet-enabledcars Devices Useby less than 4% of the respondents in the Arab region.

Connections or Services Usually Activated on Your Phone, Tablet or

83.4% 66.3% Social Media and the Internet of Things - Towards Data-Driven Policymaking in the Arab World: Potential, Limits and Concerns

21

Agree

39.0%

33.7%

27.7%

Neutral

The majority of respondents around the region positively viewed the increased Internet-enablement 15.8% Disagree 15.1% of everything. However, around 60% 19.8% said that while they view this growth positively, they also have 14.1% some concern.12.5% Meanwhile, around 23% viewed the universal connectivity in negative light (18% had 11.6% Strongly Disagree level negative view but said that they see some This finding indicates a healthy 11.1% 6.9% 6.1% value in this trend). of awareness in the region about both the potential and risks associated with IoT. For example, the Collecting my Data: Concerned that Monitoring: Concerned I am being Selling Mye Data: Concerned my expansion of IoT infrastructure is increasing the likelihood magnitude my personal info and activities are monitored as everything around me personalof infoacould be bought and of threats occurring. These by different entities online connected to the internet sold as everything around me range collected between breaches of privacy, mass surveillance, proliferation of ransomware and malware, to connected to the internet cybersecurity threats that are life-threatening on the personal level and highly risky for businesses and national interests. On the brighter side, the promises put forward by IoT implementations for increased government responsiveness, efficiency, and better adapted public policies are widely acknowledged as well.

How do youHow feeldo about a future everything around youyou would bebeconnected to you feel about where a futurealmost where almost everything around would the Internet and interacting with other devices? connected to the Internet and interacting with other devices?

Very Positive, 8.6%

Positive, but I have some concerns, 59.4%

Very Negative, 4.3%

Negative, but I see some value, 18.3% Neither positive nor negative, 9.4%

The Limitations of IoT in Policymaking in the Arab Region

Do You Usually Allow anyin Devices AppsMobile to Connect withphones Your penetration Internet connectivity is expanding rapidly the Arabor region. and smart 4 are among the highest globally in some of Accounts? the Arab world . With expanding smart city Social parts Media transformations, digital government and electronic participation initiatives, experiments with IoT applications are 54.6% already taking place in the region. These experiments view IoT both as a source 47.9% of big data for decision making, as well as an enabler of citizen-government interactions, customer45.4% engagement and data-driven policymaking. Despite their widely acknowledged promises, there are still numerous limitations for IoT applications in public policy formulation. For example, how connected are individuals in the region in terms of IoT devices and objects? How feasible are IoT implementations as reliable sources of data and enablers of engagement? 6.7%

Scope of IoT Connectivity in the Arab Region Devices Phoneof Internet Only Phone Apps around (no Only Devices (no Phone According to our AND survey users, 55% of households inNothing the allowed regionto have 2 to 5 Apps are allowed devices) Apps) interact with my Social Internet-enabled devices (other than computers and laptops) and another 25% have 6 to 10 internetMedia Accounts

44.

22

The Arab World Online 2017. Mohammed Bin Rashid School of Government, Dubai. Arab Social Media Report 2017

30%

Video Games Console (Internet-enabled)

20%

Music player (Internet-enabled)

10%

Cameras (Internet-enabled)

11.5% 10.8% 8.3%

0% Smart watches 4.8% connected devices5. These include all types of Internet-enabled devices, whetherNonpersonal or Ability to If You Can Code of Data Laws Prior Consent Ability to Knowing Knowing Personal car (Internet-enabled) 3.9% Commercial Approve according Conduct (AUP, Download andthe IoT May Remove your on What most Who Uses What Data is otherwise. Entertainment devices dominate of devices in the region to our Uses Only All Your Data Sectors that Fair Use) be Done with Regulations Data from the Data Collected Home electrical appliances (Internet-enabled) 2.9% May Use Availability Availablity Your Data Repositories survey respondents. Around 65% of respondents said that they own an Internet-enabled device that your Data Physical security systems (Internet-enabled) classifies as an entertainment device (TV, video 2.3% games console, set-top box, cameras, music player, Children toys (Internet-enabled) 1.8% toys, etc.), with Internet-enabled TVs being the largest category Social Media IoT (more than 20% of all respondents). Wristband 1.8% This is followed by computer peripherals that connect directly to the Internet, such as printers, Smart meters cameras and other devices. Only around 8%1.4% of respondents said that they own wearable devices Other wearable devices 0.9% such as smart watches, wristbands, etc. Meanwhile, less than 7% own physical appliances that Other IoT devices 0.3% connect to the Internet, such as security systems, smart meters, home electrical appliances and less None of the above 1.6% than 4% own internet-connected vehicles. While uptake of IoT devices is increasing, the potential for utilizing the networks of personal IoT devices in data-driven policymaking is becoming more feasible. However, so far, the types of devices and ways of usage may limit the feasibility of utilizing Connections oraServices Usually on Your or in the region. the network of IoT devices as meaningful and Activated reliable source of Phone, societalTablet big data

other Internet-enabled Devices You Use

Internet-Enabled Devices Ownership per Household in the Arab Region 2017 (survey self-reported) Internet-Enabled Devices Ownership per Household in the Arab Region 2017 (survey self-reported)

83.4% 66.3% 54.5%

25.5% 19.7% 25.3%

WiFi

9.0%

Only 1

Mobile data (GPRS,2G,3G,4G,..)

2 to 5

GPS

Bluetooth

6.1%

6 to 10

11 to 15

5.6%

3.5%

Sensors (movement, light,..)

NFC

3.4%

0.6% Other data connections

Not sure

2.2%

0.7%

0.5%

0.2%

0.2%

1.2%

16 to 20

21 to 25

26 to 50

51 to 100

More than100

None

Internet-enabled Devices Ownership by Type in the Arab Region 2017 (survey self-reported) Internet-enabled Devices Ownership by Type in the Arab Region 2017 (survey self-reported)

Personal IoT Devices and Apps: On your smatphone, do you use apps that connect with any physical devices online (IoT devices)? 20.8% TV (Internet-enabled) Computer Peripheral (Internet-enabled Printer, etc.)

15.2%

Set-top box (Internet-enabled)

12.1%

Video Games Console (Internet-enabled)

11.5%

Music player (Internet-enabled)

10.8%

Cameras (Internet-enabled)

8.3%

Smart watches

4.8%

Personal car (Internet-enabled)

3.9%

Yes

Home electrical appliances (Internet-enabled) 47%

2.9%

Physical security systems (Internet-enabled) Children toys (Internet-enabled) Wristband Smart meters Other wearable devices Other IoT devices None of the above

55.

2.3%

No 53%

1.8% 1.8% 1.4% 0.9% 0.3% 1.6%

It is worth reminding that our research surveys Internet users in the region, which by definition are connected to the Internet one way or another.

Does the place you work or study at has any Internet-enabled Internet-Enabled Devices Ownership Householdsystems? in the Arab Region 2017 commercial orper industrial


(survey self-reported)

23

70% 60% 50%

Modes of 40%Connectivity According to the survey respondents, people in the region usually connect to the Internet through 20% Wi-Fi (83%) when using Internet-enabled devices, such as phones, laptops, tablets, among others. 10% Data connectivity over mobile subscriptions with service providers, such as through 3G, 4G or LTE 0% connections, comes next with 66% of respondents. Other connectivity services or modes of connection NonAbility to If You Can Code of Data Laws Prior Consent Ability to Knowing Knowing Approve Download Conduct (AUP, and Bluetooth May(25%), Remove your on What Who Uses What Data is that are usually enabled by users include GPS (20%), sensors such Commercial as motions and Uses Only All Your Data Sectors that Fair Use) be Done with Regulations Data from the Data Collected May Use Availability Availablity side, Your Data Repositories light (5%), and NFC connections (3.5%). On the positive these modes of your connection can provide Data data that may feed into decision and policymaking. However, they do also increase the likelihood of privacy or security breaches for individuals if Media no sufficient technological or legal precautions are put Social IoT in place. These findings may suggest that increasing the potential of harnessing personal IoT devices as sources of societal big data in policymaking can be better achieved if large-scale Wi-Fi coverage is available. Alternatively, if the coverage and cost of mobile data connectivity is made affordable across the region this potential will increase as well. 30%

Connections orConnections Services Usually Activated onActivated Your Phone, Tablet or other Internet-enabled or Services Usually on Your Phone, Tablet or Devices You Use

other Internet-enabled Devices You Use

83.4% 66.3%

25.5% 19.7%

WiFi

Mobile data (GPRS,2G,3G,4G,..)

GPS

Bluetooth

5.6%

3.5%

Sensors (movement, light,..)

NFC

3.4%

0.6% Other data connections

Not sure

IoT-to-Mobile Connectivity Interactivity between Internet-enabled IoT devices and human beingsbyisType facilitated through Internet-enabled Devices Ownership in the Arab Regiondifferent 2017 mediums, such as computers, control panels(survey andself-reported) mobile devices. The primary interface or medium of interactivity between people and IoT devices is the smart phone. More specifically, this takes place through specifically-developed software applications (or apps) installed on the smart phones and designed to control or interact with physical IoT devices. Likewise, mobile apps are also the primary TV (Internet-enabled) 20.8% medium for interaction between physical devices and social media accounts. Computer Peripheral (Internet-enabled Printer, etc.)

Set-top respondents, box (Internet-enabled) survey use

15.2%

According to our of mobile apps to IoT devices in the region is Gameshalf Console (Internet-enabled) widespread.Video Around of the respondents in our survey (47%) said11.5% that they use apps on their smart Music player (Internet-enabled) phones that interact with physical Internet-enabled devices. This 10.8% highlights a growing convergence Cameras (Internet-enabled) 8.3% between human social networks and IoT devices in the region, mediated by smart mobile apps. Smart watches 4.8% However, which IoT devices do most people interact with through mobile apps? Entertainment Personal car (Internet-enabled) 3.9%survey. Internet-enabled televisions, cameras IoT devices have the largest share according to our Homeconsoles electrical appliances (Internet-enabled) 2.9% that people interact with through mobile apps and gaming are the top three IoT devices systems (Internet-enabled) 2.3% by internet-enabled cars (6.5% of respondents), (25%, 24%Physical and security 12% respectively). This is followed Children toys (Internet-enabled) 1.8% wearable IoT devices (6%) and security devices (3.6%) and connected home appliances (3.5%). Wristband

Smart meters Other wearable devices Other IoT devices None of the above

24

interact 12.1% with

1.8% 1.4% 0.9% 0.3% 1.6% Arab Social Media Report 2017

Personal IoT Devices: Which of the Following Physical Devices do You Connect with Online through Apps Installed on Your Smart Phone?

Personal IoT Devices: On your smatphone, do you use apps that connect with any physical devices online (IoT devices)?

24.5%

Yes 47%

23.7%

No 53%

12.7%

6.5%

6.0% 3.6%

3.5% 0.6%

TV: Apps connecting to Internetenabled TV

Cameras: Apps connecting to cameras (e.g. webcam, surveillance, baby monitoring)

Video games console: Apps connecting to gaming consoles

Cars: App connecting to Internetenabled car

Wearable devices: Apps connecting to watches, sport bracelets, healthcare products, smart glasses, etc.

Physical security systems: Apps connecting to security systems (e.g. door locks, garage systems, etc.)

Home Other Devices: electrical Apps appliances: connecting to Apps other Internetconnecting to enabled thermostats, physical A/Cs, devices refrigerators, light bulbs, washing machines, etc.

Data derived from mobile phone triangulation is already used as a rich data source by mobile service providers, as well as certain government authorities in the region. These findings provide a snapshot of people’s modes of connectivity in the region, their preferences and behaviors. Coupled with other sources of behavioral and usage data, these sources can help inform decisions on utilizing IoT and social media data for public engagement and policy development. As the volume and uptake of personal IoT devices increase in the region, expanding the utility of this big data source will rely on many regulatory, economic and infrastructural factors, which remain limited. However, the use of such data is still limited and restricted in scope, and provides a partial and fragmented digital picture of society. Public Concerns on IoT in the Arab Region In addition to the limitations highlighted above, public concerns about the use of IoT data by government are widespread. The majority of respondents in our survey are concerned about their personal data and activities being collected by different entities online (65% agree or strongly agree and 19% disagree or strongly disagree). Similarly, people in our survey are also concerned about monitoring and surveillance with increased connectivity of everything (63% agree or strongly agree) and 59% are concerned about their personal data being used for commercial uses by different entities, such as selling or buying personal data (25% disagree or strongly disagree). While these views and concern differ from one country to another, a majority of people in the region are still living either in countries with under-developed IoT infrastructure or in rural areas with limited connectivity, these findings highlight certain concerns that may expand with the increased introduction of IoT infrastructure across the region.


25

31.8%

Strongly Agree

23.4%

31.3% Agree

Public Concerns about Internet-Connected Things in the Arab Region 39.0% Public Concerns about Internet-Connected Things in the Arab Region 27.7% 33.7%

15.1% 31.8% 12.5%

15.8%

19.8% 23.4%

Disagree Strongly Agree

31.3% 14.1%

6.9%

11.6% 6.1%

Collecting my Data: Concerned that 33.7% my personal info and activities are collected by different entities online

39.0% I am being Monitoring: Concerned monitored as everything around me connected to the internet

15.1%

19.8%

12.5%

Neutral

Strongly Agree Disagree

11.1% Selling Mye Data: Concerned my 27.7% personal info could be bought and sold as everything around me connected to the internet

Neutral

15.8%

Disagree

14.1%

6.9%

11.6% 6.1%

Collecting my Data: Concerned that my personal info and activities are collected by different entities online

Monitoring: Concerned I am being monitored as everything around me connected to the internet

Strongly Disagree

11.1% Selling Mye Data: Concerned my personal info could be bought and sold as everything around me connected to the internet

How do you feel about a future where almost everything around you would be connected to the Internet and interacting with other devices?

1.3. The Convergence of Social Media and IoT Data - Potential, Limitations and Concerns on Big Data for Public Policy The previous sections highlighted the potential, limitations, and concerns about using either social Positive, Positive, I have a futureaccording do you feel about where almost everything youThe would be presented in 8.6% media or IoTHow separately in but policymaking to our regionalaround survey. findings some concerns, the1)Internet interacting other devices? the previous sectionsconnected suggest that: there isand a potential forwith increasing the uses of big data in policy 59.4% to Very potentially a number of making in the region, both from social media and IoT sources, 2) that there are Negative, limitations facing data-driven policymaking, which are specifically related to 4.3% the diverse cultural, social and political contexts of the Arab region, and 3) that there are wide-spread concerns among the public in terms of using personal data through social media or IoT by governments and Very businesses. This however, Positive, we explore the potential, is also coupled with a healthy level of awareness about the risks. In this section, Positive, but I have 8.6% limitations and concerns relating IoT data Negative, but I as big data sources for some concerns,to the convergence of social media and see some value, policy formulation. 59.4% Very

positiveRegion The State of IoT and Social Media Interaction inNeither the Arab

18.3%Very Negative, 4.3%

nor negative, 9.4%

The lines between the virtual and physical are blurring rapidly. Social media accounts and IoT devices are interacting either by themselves or through human mediation. Home appliances are updating Facebook status, banking apps are operated by tweets and personal cars are controlled byI social media accounts Negative, but see some value, social media accounts and posting social media updates. In other words, alongside their human owners, 18.3% are increasingly being inhabited by pieces of software, such as IoT-driven “bots”. Likewise, internetpositive enabled devices are increasingly being operated,Neither updated or triggered by social media activities. Within nor negative, this context, what are the usage habits of social media and IoT devices in the region? 9.4%

Do You Usually Allow any Devices or Apps to Connect with Your Social Media Accounts?

Do You Usually Allow any Devices or Apps to Connect with Your Social Media Accounts? 54.6% 47.9%

45.4%

Do You Usually Allow any Devices or Apps to Connect with Your Social Media Accounts? 54.6% 47.9% 6.7%

45.4%

Devices AND Phone

Only Phone Apps (no

Only Devices (no Phone

Nothing allowed to

Apps are allowed

devices)

Apps)

interact with my Social Media Accounts

6.7% Arab Social Media Report 2017

26 Devices AND Phone

Only Phone Apps (no

Only Devices (no Phone

Nothing allowed to

Apps are allowed

devices)

Apps)

interact with my Social

Social Media Accounts? 54.6% 47.9%

Around 45% of respondents around the region said that they don’t allow any interactions to take place on their behalf with their personal social media accounts. However, more than half of the respondents (55%) said that they do allow their social media accounts to interact with physical devices and software apps they use (like phones, Internet-enabled appliances, wearables, connected cars, etc.). This includes software apps installed on smart phones as well as physical devices owned by users. The majority of those who allow interactions to take place with their social media accounts (48% out of 55% of total respondents) said that they only allow apps on their phones or devices to do so. Those who allow interaction between 6.7% physical IoT devices and their social media accounts are around 7% of respondents. 45.4%

What IoT devices people Arab region usuallyOnly allow to connect social media accounts? Devicesdo AND Phone in theOnly Phone Apps (no Devices (no Phonewith their Nothing allowed to As with the general trend on the use of Internet-enabled devices in the region highlighted in previous Apps are allowed devices) Apps) interact with my Social questions, the largest portion of respondents who said that they allow suchMedia interactions, do so with Accounts entertainment devices (such as music players, cameras, TVs, etc.) This is followed by cameras, smart watches, wristbands and cars.

What Internet-enabled do Usually Allow Connect with Social Media Accounts? WhatDevices Internet-enabled Devices do Usuallyto Allow to Connect withYour Your Social Media Accounts? 23.0%

11.3%

9.4% 5.8%

Entertainment devices (e.g.music player, TV, gaming consoles, Internet radios, etc.)

Cameras (e.g.webcam, baby monitoring camera, surveillance cameras)

Smart Watches (e.g. Apple Watch)

Wristbands oractivity trackers (e.g. Fitbit, smart wristbands, etc.)

5.5%

Car or vehicles

4.6%

Home appliances (e.g. light bulbs, refrigerators, security systems, etc.)

3.9%

Sensors (e.g. temperature, light, movement sensors, etc.)

3.2%

Other wearable devices (e.g.Smart Glasses, VR headset)

0.5% Other devices

Towards Inclusive Data-Driven Public Policy: Addressing Public Concerns on Big Data in the Arab Region In the Future, Would You Allow the Following Kinds of Interactions Between Your Social Media Account and Internet-enabled Devices and Objects?

Previous findings demonstrate a broad spectrum of public concerns about practices affecting personal data From the citizen’s viewpoint, when 9.4% generated by social media and IoT devices. Based on 13% Yes 17.9% governments or businesses try to the findings, the survey also explored public perceptions harness personal data to extract regarding potential policy responses to these concerns. Maybe 32.5% 41.1% meaning, the key concerns are 34.1% From the citizen’s viewpoint, when governments or related to loss of control, breaches businesses try to harness personal data to extract No of privacy, and misrepresentation meaning, the key concerns are related to loss of control, or misuse of data. breaches of privacy, and misrepresentation or misuse 54.6% 49.4% 48% of data. These concerns are understandably justified in the era ofIoTbig data and data-driven governance. TheIoT-to-IoT: Allow your devices to to People: Allow your IoT to Social Media: Allow your devices to interact withother devices to share information on interact with other devices that survey respondents were presented with a set of possible scenarios and responses to assess public people you know your social media accounts don’t belong to you acceptance of each policy response. What policy responses can be taken to ease public concerns about big data related to social media and IoT?


27

Based on the survey findings, possible policy responses can be categorized in three main groups: 1) transparency, 2) user control of data, and 3) data-related regulations. Relatively large percentages of the survey respondents said that each of the following would provide them with enough assurances or make them less concerned about their personal social media or IoT data being collected, retained and analyzed by governments or businesses: I - Transparency: The top two potential policy responses that respondents said would make them less concerned about the use of their personal data by government and businesses are related to transparency measures: What and who; ability to know what data is collected, and ability to know who has access to the data. The following are the possible transparency-related policy responses people supported in the Arab region: a. Knowing What Data: If the user was given the right to know exactly what data is collected and retained about them through social media and IoT devices b. Knowing Who Can Access the Data: If users were given a list of entities (companies or government agencies) that may have access to personal data generated by social media or Internet-connected devices. c. Getting the Data: If users can get a complete copy or download all personal data collected through social media Internet devices. d. Mapping IoT: If public maps about the locations of sensors and cameras are available, with regards to IoT devices such as sensors and cameras. This may be important in easing concerns with regards to smart cities developments. II - Control: a. Ability to Remove Data: Users ability to remove personal data from government or commercial data repositories was one of the top three responses highlighted by respondents as a measure to ease their concerns across the region. b. Consent: Prior explicit user consent on what may be done with the personal data collected was highlighted as an important factor that can make people less concerned. This was the only possible policy response that more respondents supported for IoT data than for social media data. c. Approving Who Uses the Data: If users have the ability to approve and control which sectors can use their social media or IoT personal data then that will make it more likely for more people to share data (e.g. only educational uses, research, public health, NGOs, government, etc.) III - Regulations: a. Data Laws: Availability of regulatory frameworks and appropriate data laws and regulations that protect the personal data of individuals when such personal data is used in policymaking. b. Code of Conduct: Availability of clear “fair use” policy, “ethical framework” or other acceptableuse policy (AUP) can ease public concerns and influence willingness to share personal their data. c. Certified privacy policies: If the privacy and data policies applied are approved or certified by independent standards body (e.g. academic board, non-government expert committee, etc.) d. Anonymization: If there are regulations forcing entities to anonymize personal data or delete any data that identify individuals.

28


Which of the Following Would Make You Less Concerned About Use of Personal Data by Which of the Following Would Make You Less Concerned About Use of Personal Data by Governments or Businesses Governments or Businesses through IoTthrough Devices ororSocial Media? IoT Devices Social Media?

70% 60% 50% 40% 30% 20% 10% 0% Knowing What Data is Collected

Knowing Who Uses the Data

Ability to Remove your Data from Repositories

Data Laws and Regulations Availablity

Prior Consent on What May be Done with Your Data

Social Media

If You Can Code of Conduct (AUP, Download All Your Data Fair Use) Availability

Ability to Approve Sectors that May Use your Data

NonCommercial Uses Only

IoT

In addition to the policy responses above, the survey findings highlight the following two important points: Views on Personal Data: Regardless of Usually the medium, personal dataPhone, is viewed in the Connections or Services Activated on Your Tablet or same light with regards to possible responses. Respondents around the region had almost identical views on possible other Internet-enabled You Use was captured from their social responses to their concerns, regardless if the personal Devices data in question media accounts or through IoT devices they use or interact with. 83.4% Public Mistrust: Overall, there is no single possible response that gained universal acceptance around the 66.3% As a possible policy response region. This may relate to the uncertainty around the uses to ease public concerns, of personal data harnessed through social media and IoT. anonymization of data received It also may imply limited confidence or trust by the public in the least number of positive the region that any of the possible25.5% responses can ease their responses. concerns about their data being misused. 19.7% For example, as a possible policy response to ease public concerns, 5.6% 3.5% 3.4% 0.6% anonymization of data received the least number of positive responses, which may indicate limited public trust in the capacity or willingness of government entities to Mobile data Sensors Other data WiFi GPS Bluetooth NFC Not sure (GPRS,2G,3G,4G,..) (movement, light,..) connections enforce such measure.

Finally, big data presents numerous promises for public policy development. However, data-driven policymaking also raises a wide range of concerns. The findings of this regional survey demonstrated a broad spectrum of public concerns about personal social media and IoT data in the Arab region. As tools of digital governance mature, social media and IoT data promise to enhance data-driven policymaking. If such opportunities are to be harnessed in the region, the policy responses discussed here need to be examined contextuallyInternet-enabled to help address and ease suchbyconcerns. Devices Ownership Type in the Arab Region 2017 (survey self-reported)

Way Forward: The Future of People, Social Media and IoT Networks TV (Internet-enabled) 20.8%continues As will be demonstrated in the next section of the report, social media adoption in the Arab region Computer Peripheral (Internet-enabled Printer, etc.) and trends point towards convergence 15.2%between social media to grow across the board. The behaviors box (Internet-enabled) 12.1% and Internet-enabledSet-top devices. This presents a new frontier for data-driven policymaking as well as a wider Video Games Console (Internet-enabled) 11.5% horizon of concerns. However, as the Arab region continues to adopt Internet technologies, will this trend Music player (Internet-enabled) 10.8% continue going forward? Cameras (Internet-enabled)

8.3%

Smart watches

4.8%

Personal car (Internet-enabled) Home electrical appliances (Internet-enabled)

3.9% 2.9%


Physical security systems (Internet-enabled) Children toys (Internet-enabled)

2.3%

1.8%

29

IoT devices form physical hierarchies and social networks on their own. They also converge with other hierarchical or human social networks to form a hybrid network of devices and people, in their virtual or physical being. Kept separated, the two networks of 1) human 6.7% social connections, digitized and “datafied” through social media technologies and platforms, and 2) physical network of IoT devices, may serve complimentary purposes. However, by converging in a hyper-connected digital world, they provide unlimited possibilities, as well as raise numerous concerns and unimaginableNothing risks. allowed The convergence of Devices AND Phone Only Phone Apps (no Only Devices (no Phone to IoT devices and social media is creating new opportunities for data-driven policymaking. The big data Apps are allowed devices) Apps) interact with my Social generated by massive networks of IoT devices and social media networks can provide an unprecedented Media Accounts granular and nuanced picture of society. They can also enable wide-scale security breaches, erosion of privacy, mass surveillance, among other risks. The boundaries are rapidly blurring between the individuals’ physical being in real-life, their presence within the virtual social media realm and their interaction with the physical network of Internet-enabled objects. People, social media accounts and IoT objects are interacting heavily and generating unprecedented amount of data. With regards to data, these three networks, the real-life social network of people, the virtual social network of social media users andDevices accounts, andAllow the “social network” hierarchy of interconnected What Internet-enabled do Usually to Connect with Youror Social Media Accounts? IoT objects, are becoming one. Based on these 23.0%trends, we asked respondents about future tendencies and the ways of interaction between these three networks, people, social media and physical IoT devices. According to our survey, a clear majority of people is hesitant about allowing increased connectivity between these three networks in the 11.3% 9.4% said that they would not allow their IoT devices to interact with other future. Around 48% of respondents 5.8% 5.5% 4.6%50% said 3.9% that in 3.2% people they know (only 18% said they would). Similarly, around the future they would not 0.5% allow IoT devices to share data with social media accounts (10% said they would). The type of interaction that received the largest negative wasCar IoT-to-IoT (or machine-to-machine) Around Entertainment Cameras Smartresponse Wristbands Home Sensors (e.g. Other interactions. or vehicles Other devices devices (e.g.webcam, Watches (e.g. oractivity appliances temperature, wearable 55% of respondents said that in the future, they will not allow internet devices they own to interact with (e.g.music baby Apple Watch) trackers (e.g. (e.g. light light, devices player, TV, monitoring Fitbit, smart Of course, these bulbs,responses movementimply (e.g.Smart other internet devices (13% said they would). that people have a level gaming camera, wristbands, refrigerators, sensors, etc.) Glasses, VR of control on such interactions in the future, which is one security of the critical policy questions as the world consoles, surveillance etc.) headset) Internet radios, cameras) systems, etc.) becomes ever etc.) more connected. The survey also highlights an obvious level of uncertainty about how people will react as these changes continue to emerge. Around a third of respondents on average, were not sure about how they will react in the future with regards to interactions between people, social media and IoT. These findings provide insight into levels of public readiness for the next phase of the data-driven revolution.

In the Future, Would You Allow the Following Kinds of Interactions Between Your Social In the Future, Would You Allow the Following Kinds of Interactions Between Your Social Media Media Accounts and Internet-enabled and Devices Objects? Account and Devices Internet-enabled and Objects?

17.9%

34.1%

9.4%

13%

41.1%

32.5%

Yes

Maybe

No

48%

49.4%

54.6%

IoT to People: Allow your devices to interact withother people you know

IoT to Social Media: Allow your devices to share information on your social media accounts

IoT-to-IoT: Allow your devices to interact with other devices that don’t belong to you

Finally, this new hybrid universe of connected humans and devices require further multidimensional and multi-disciplinary research and policy responses. This report provides an in-depth exploratory look into the possibilities, limitations and public concerns of two key sources of big data in the Arab region. The findings can hopefully increase public awareness and inform policy responses while the region embraces “smart” government and smart city approaches as routes towards building future governance models. 30

Would the Following Make You Less Concerned About Governments or Arab Social Media Report 2017 Businesses Using Your Personal Data Collected through IoT Devices?

2. Mapping Social Media in the Arab World: 2010-2017 Over the past six years, the Arab Social Media Report series has been providing in-depth analysis on social media trends, growth and demographic breakdowns across 22 Arab countries. During this period, the findings of the series highlighted critical emerging transformations taking place on a regional level. The series has also been informing regional and international policy formulation, strategic business decisions and scholarly discourse related to development, socio-economic policy, and digital governance and its associated transformations in the Arab region. The 7th edition of the Arab Social Media Report continues to explore the growth and influence of key social media platforms in the Arab region. It follows from the previous six editions in covering the growth and usage trends of Facebook, Twitter and LinkedIn in the region as some of the most widely used platforms in their respective categories. However, for the first time, this edition also explores usage trends related to Instagram, one of the fastest growing photo-sharing social media platforms in the region. This section of the report provides a detailed view of the changes on the social media horizon in the Arab World since the previous edition of the report. In addition, it puts the changes over the past two years in context by mapping key elements of the growth trends in the Arab region in the period between 2010-2017. The data presented in this section on demographic breakdowns, such as age, language and gender were collected periodically for all 22 Arab states between January 2016 and January 2017. This was combined with over-time growth data which was collected periodically between 2010 and 2017 as part of previous editions of the Arab Social Media Report series.

Social media in the Arab world is gaining more ground, growing increasingly localized, becoming less youthful, while remaining a maledominated medium.

In addition to the regional survey, this updated in-depth analysis of social media transformations expands the coverage and scope of the series, with the objective of enriching our understanding of the cross-societal and citizen-government interactions in the region. The findings of the 2017 edition of the Arab Social Media Report highlights strong and continued growth of social media uptake and influence across Arab societies. Overall, the findings of this edition indicate that social media in the Arab world is gaining more ground, growing increasingly localized, becoming less youthful, while remaining a male-dominated medium. Social Media and the Internet of Things - Towards Data-Driven Policymaking in the Arab World: Potential, Limits and Concerns

31

Global Social Media Highlights: •

Facebook6: Facebook remains the most popular social media platform globally. By the third quarter of 2016, it had 1.79 billion monthly active users, an increase of 16% since 2015. Out of those, 1.66 billion people use the platform through mobile devices monthly, an increase of 20% year-over-year. Approximately 85% of those who use Facebook daily are outside North America.

•

Twitter7: By the third quarter of 2016, Twitter had 317 million monthly active users around the world, a 3% increase over 2015. Out of those, 85% are actively using the platform on their mobile phones. The majority of Twitter users (79%) are based outside the US.

•

LinkedIn8: Almost 467 million people have LinkedIn accounts, an 18% growth since 2015. Out of those users, 106 million are active on monthly basis and around 65 million of them use the platform through their mobile devices.

•

Instagram9: Instagram has 300 million active users globally, primarily used through mobile devices.

•

Youtube10: Over one billion people use Youtube, half of them using their mobile devices.

Social Media in the Arab Region The following is a snapshot of some key findings from the 2017 edition of the Arab Social Media Report: •

Arabic Language: Arabic language use on social media has increased substantially during the past two years in the region. For the first time ever, the Arabic language is now used in a majority of social media activities in the Arab region. On average, the rate of using Arabic language in social media activities stands at 55%, up from 43% two years ago on Facebook, while 72% of all tweets in the region are now posted in Arabic.

•

Language by Region: In terms of language use on a regional level, North African countries witnessed the strongest growth of Arabic language use over social media during the past two years. In contrast, the rates of using Arabic language on social media in all six Gulf countries have dropped during the past two years.

•

Youth: Social media users in the region remain largely youthful. On average, 64% of social media users in the region are under 30 years old. However, this percentage has slightly, but consistently, decreased over the past six years where early adopters of social media in the region are maturing.

•

Age Balance: Compared to the rest of the Arab region, the Gulf countries largely have more balanced age breakdowns of young users who are under 30 and those who are over 30 years old, indicating maturity of usage across age groups in society.

66. 77. 88. 99. 1010

32

For the first time ever, the Arabic language is now used in a majority of social media activities in the Arab region.

Facebook 2016: http://www.fb.com/ Twitter 2016: http://www.twitter.com/ LinkedIn 2016: http://www.linkedin.com/ Instagram 2016: http://www.instagram.com/ YouTube 2016: http://www.youtube.com/ Arab Social Media Report 2017

•

Gender Gap: The gender gap persists among Arab social media users with only 1 out of 3 social media users being female at best, which is lower than the global average. This imbalance has remained almost unchanged over the past six years.

•

Regional Gender Balance: On a regional level, only 1 out of 4 users in the Gulf region is a woman (24.6%), compared to 1 out of 3 in the other parts of the Arab region on average (32.3%). The North Africa region has a slightly better gender balance of social media users compared to the Arab average, with more than 36% of users being females.

•

Gender-Balance Per Platform: The ratio of women to men in the region is almost identical on Facebook and Twitter (1 to 3). Meanwhile, career-focused social media platforms such as LinkedIn has lower percentage of women among its users.

•

Best and Worst Countries on Gender Balance:

•

•

Facebook: Palestine has the best gender balance on Facebook, Oman has the worst.

•

Twitter: Bahrain has the best gender balance on Twitter, Yemen has the worst.

•

LinkedIn: Lebanon has the best gender balance on LinkedIn, Saudi Arabia has the worst.

Regional Penetration rates: The Gulf countries largely dominate the top five spots in terms of penetration rates of key social media platforms among their populations, including Facebook, Twitter, LinkedIn and Instagram.

2.1 Facebook in the Arab World 2010-2017 Highlights - Facebook in the Arab Region •

Most Popular: Facebook is by far the most popular social media platform in the region with around 156 million users by early 2017, up from 115 million a year earlier.

•

Growth: Overall, the number of Facebook accounts in the Arab region has more than doubled over the past three years.

•

Regional Penetration: By early 2017, the penetration rate of Facebook accounts among the Arab population in the region stands at 39%, up from 28% a year earlier.

•

New Users: Over the past two years, the largest growth in terms of number of new users joining Facebook was in Egypt (14 million), followed by Algeria (9.4 million) and Iraq (7.3 million).

•

Penetration growth: Between 2014 and 2016, penetration rates of Facebook grew by 11.4 percentage points per country on average in the Arab region. The four countries of Qatar, UAE, Algeria and Libya grew by more than 20 percentage points each.

•

Users Activity: One out of three persons in the Arab region owns a Facebook account. However, 1 out of 5 persons uses that account on daily basis.

•

Top Countries: In terms of penetration rates, Qatar and the UAE have, by far, the highest penetration rates of Facebook accounts, with almost 95% penetration rate in each country. They have a 20-point lead over Bahrain (ranked 3rd), which is followed by Kuwait and Jordan.

•

Largest Share: In terms of distribution, around half of all Facebook users in the Arab region are based in three countries: Egypt, Saudi and Algeria. Around 23% of all Arab Facebook users are in Egypt alone.


33

20

Remaning

15

UAE

Morocco

Syria users Most Active Countries: Qatar has the most active users on daily basis, out of all Facebook 10 in the country, followed by the UAE and Tunisia. Users in Yemen are the least active daily. Tunisia

•

Jordan

5

Most Active Region: Overall, users in the Gulf region are the most active daily. Almost 2 out of Lebanon 3 persons living in the Gulf region own a Facebook account; and 1 out of 3 persons uses that 0 account on daily basis.

•

Jan. '11

Jan. '12

Jan '13

Jan. '14

Jan. '15

Jan '16

Jan '17

•

Youth: Around 64.3% of Facebook users in the region are under 30 years old, dropping from 67% two years earlier, and from 70% in 2012. There is a slow, but steady, uptake of Facebook by users over the age of 30 in the region.

•

Women: The percentage of female users of Facebook remains persistently lower than the global average. By 2017, only 32.3% of Facebook users in the Arab region are women.

Millions

2.1.1.3.5 Facebook Growth and Penetration Rates in the Arab RegionSudan Libya

3.0

Kuwait Facebook remains the most popular social media platform in the Arab region by a far margin. Qatar 2.5 Collectively, the number of Facebook users in the Arab region has steadily increased over the past Yemen six years as per the findings of the Arab Social Media Report series, reaching more that 156 million in 2.0 Oman 2017. This number represents a year-on-year increase of close to 41 million, up from 115 million in the Palestine 1.5 beginning of 2016. Individually, the growth rates in several countries fluctuated over this time period, Bahrain influenced by multiple political, societal and economic factors. For example, after the popular uprising Somalia 1.0 in Egypt in 2011, there was a noticeable increase in growth rates of Facebook users in the Mauritania country, as 0.5 the public, the government and political powers increasingly used social media as a communication, Djibouti mobilization and engagement medium. The growth rate—while still moving upwards—slightly slowed Comoros 0.0 down starting 2013. In contrast,Jan military conflicts in Jan. other'15countries in theJan region, have negatively Jan. '11 Jan. '12 '13 Jan. '14 Jan '16 '17 affected the growth rates of social media during different phases of the conflicts. In Syria, Yemen and Palestine there were noticeable slumps in social media growth trend at multiple points over the years, which seem to coincide with intensifying levels of the conflict in each country. Meanwhile, market changes in other countries, may have affected social media growth, and Facebook growth specifically. For example, Algeria’s Facebook adoption rates increased fastest in 2015-2016, where increased Internet access and more competition in the telecommunication market may have had an effect. These observations highlight potential areas for additional research on the impact of political, economic and societal dynamics on social media adoption and uses, and vice versa.

Millions

Growth of Facebook Users in the Arab Region between June 2010 and Jan 2017

160 140 120 100 80 60 40 20 0

34

Jun ‘10

Jan ‘11

Jun ‘11

Jan ‘12

Jun ‘12

Jan ‘13

Jun ‘13

Jan ‘14

Jun ‘14

Jan ‘15

Jun ‘15

Jan ‘16

Jun ‘16

Jan ‘17


Millions Millions

Growth of Facebook Users in the Arab States (Top 10 Facebook Populations - Jan 2011 to Jan 2017)

35

Egypt

30 35

Saudi Egypt Algeria Saudi Iraq Algeria Remaning Iraq Morocco Remaning UAE Morocco Syria UAE Tunisia Syria Jordan Tunisia Lebanon Jordan

25 30 20 25 15 20 10 15 5 10 0 5 0

Jan. '11

Jan. '12

Jan '13

Jan. '14

Jan. '15

Jan '16

Jan '17

Jan. '11

Jan. '12

Jan '13

Jan. '14

Jan. '15

Jan '16

Jan '17

Lebanon

Growth of Facebook Users in Remaining Arab Countries (Excluding Top 10 Facebook Populations - Jan Millions Millions

2011 to Jan 2017) 3.5

Sudan Libya

3.0 3.5

Kuwait Sudan Qatar Libya

2.5 3.0

Yemen Kuwait Oman Qatar

2.0 2.5

Palestine Yemen Bahrain Oman

1.5 2.0

Somalia Palestine Mauritania Bahrain

1.0 1.5 0.5 1.0 0.0 0.5 0.0

Djibouti Somalia Comoros Mauritania

Jan. '11

Jan. '12

Jan '13

Jan. '14

Jan. '15

Jan '16

Jan '17

Jan. '11

Jan. '12

Jan '13

Jan. '14

Jan. '15

Jan '16

Jan '17

Djibouti

Comoros

In terms of the actual accounts, the average number of Facebook accounts per country stands in 2017 at 6.7 million account. In 7 out of 22 Arab countries the number of Facebook accounts exceeds the regional average. These countries are: Egypt, Saudi, Algeria, Iraq, Morocco, UAE and Syria. Egypt still leads by far in terms of number of Facebook users with close to 34.5 million accounts.

MillionsMillions

In terms of penetration rates of Facebook accounts, the regional penetration rate of Facebook accounts stands at 34% on average. The majority of countries are above the regional average with Qatar and the UAE remaining in the lead and crossing the 90 penetration rate for the first time. The 160 two countries have almost identical penetration rates of almost 95 percent. This is more than 20 140 160

120and the Internet of Things - Towards Data-Driven Policymaking in the Arab World: Potential, Limits and Concerns Social Media 140 100 120

35

10 0

Jun ‘10

5

Jan ‘11

Jun ‘11

Jan ‘12

Jun ‘12

0.84 1.02 1.6 0.07 0.18 0.44

0

2.2

1.9

Jan ‘13

2.5

Jun ‘13

Jun5.1 ‘14

Jan ‘14

6.8

Jan ‘15

3.4

3.2

3.1

2.8

6.3

8.7 Jun ‘15

Jan ‘16

Jun ‘16

Jan ‘17

om ia a B lia ah P rain al es tin e O m an Y em en Q at ar K uw ai t Li by S a ud Le an ba no Jo n rd a Tu n ni si a S yr ia U M AE or oc co Ira q A lg er ia S au d E i gy pt

an

ou

S

rit

M

au

jib

or

D

om

C

ti

os

percentage points above Bahrain, the next country on the rank in terms of penetration rate. Overall, the eight countries of Qatar, UAE, Bahrain, Kuwait, Jordan, Saudi, Lebanon and Tunisia have more than 50% penetration rates of Facebook accounts in each country. Meanwhile, the four countries of Comoros, Yemen, Sudan and Somalia have less than 10% penetration rates. The pace of change in penetration rates of Facebook accounts is another measure indicating growth. Qatar, UAE, Algeria and Libya each has seen an increase of more than 20 percentage points in penetration rates of Facebook accounts during a period of 30 months between June 2014 and January 2017. Change in Number of Facebook Users Between May 2014 and Oct 2016 (Thousands)

Penetration of Facebook Accounts (and number of accounts) in the Arab States in 2017 Millions

Facebook Penetration in the Arab Region (2017)

35

Iraq, 7,291 Saudi, 5,782 Morocco, 5,537

30

90%

UAE, 2,717 Jordan, 1,685 Tunisia, 1,498 71% 73% Libya, 1,182 66% Lebanon, 1,084 Qatar, 928 58% Sudan, 827 55% 56% Kuwait, 782 Oman, 671 49% Somalia, 558 43% Yemen, 421 41% 38% 40% Palestine, 314 37% 37% Bahrain, 280 34% Mauritania, 185 Djibouti, 74 Comoros, 32

25 20 15 10

80% 70% 60% 50% 40% 30%

20%

Syria, -374

-2,000 5

Egypt, 14,060

94% 95% 100%

Algeria, 9,380

0 2,000 11% 8% 8% 8% 9%

4,000

6,000

8,000

10,000

12,000

14,000

20% 16,000

Thousands 10%

0 ba a no n S au d Jo i rd a K n uw ai B ah t ra in U A E Q at ar

a

Le

Tu

ni

si

ia

by

Li

er

A

lg

q

an

O

m

co

Number of Users

Ira

pt M

or

oc

ia

E

gy

e

S

yr

ti P

al

es

tin

ia

ou

jib

rit

D

om

au

C

M

Y

an

os

or

en

an

em

ud

S

S

om

al

ia

0%

Penetration

Average

Change in FacebookRates Penetration Rates (May 2014 andArab Oct 2016) Change in Country-level Penetration of Facebook in the States (between 2014 and 2017)

30%

19.0% 18.5% 16.0% 14.9%15.2% Kuwait, 2% 13.5% Qatar, 2% 11.5% Libya, 2% Remaining, 4% 8.8% Sudan, 2% 7.7% 5.6% Lebanon, 2% 4.5% 4.7% 5.0% 3.4% 4.0% 4.0% Egypt, 23% Jordan, 4% 3.1% 1.5% Tunisia, 4%

20% 15% 10% 5% 0%

Ira q ah ra in Li by a A lg er ia U A E Q at ar

om ia or os O m P al an es ti S ne om a Le lia ba no K n uw D ait jib ou Tu ti ni si a E gy M p or t oc c Jo o rd an S au di

C

au M

B

ia

an

UAE, 6%

rit

yr

an

S

ud

S

Y

em

en

Syria, 5%

Morocco, 9%

Saudi, 13%

36

24.1% 23.1% 21.9% 20.0%

DISTRIBUTION OF FACEBOOK USERS IN THE ARAB STATES (2017)

25%

Algeria, 12%

Iraq, 10%

Number of Facebook Users in the Arab States (Jan 2017)


Change in Number of Facebook Users Between May 2014 and Oct 2016 (Thousands)

Changes in Numbers of Facebook Users in the Arab States (between 2014 and 2017) Egypt, 14,060

Millions

Facebook Penetration in the Arab RegionAlgeria, (2017) 9,380 Iraq, 7,291 Saudi, 5,782 Morocco, 5,537

35

UAE, 2,717 Jordan, 1,685 Tunisia, 1,498 Libya, 1,182 Lebanon, 1,084 Qatar, 928 71% 73% Sudan, 827 Kuwait, 782 66% Oman, 671 Somalia, 558 58% 55% 56% Yemen, 421 Palestine, 314 49% Bahrain, 280 Mauritania, 185 41% 43% Djibouti, 74 38% 40% Comoros, 32 37% 37%

30 25 20 15

94% 95% 100% 90% 80% 70% 60% 50% 40%

34%

Syria, -374

-2,000 10

0

2,000

20%

4,000

6,000

8,000

10,000

12,000

14,000

11% 8% 8% 8% 9%

5

30% 16,000

Thousands 20% 10%

ba a no n S au d Jo i rd a K n uw ai B ah t ra in U A E Q at ar

a

Le

Tu

ni

si

ia

by

er

Li

A

lg

q

an

m

O

Ira

co

pt M

or

oc

ia

E

gy

e

S

yr

ti

P

al

es

tin

ia

D

jib

ou

os

an

M

C

au

om

Y

rit

or

en

an

em

ud

S

S

om

al

ia

Egypt gained more than 14 million new users during this 30-month period, followed by Algeria, 0% gaining 09.3 million new users, then Iraq (7.2 m), Saudi (5.7 m) and Morocco (5.5 m). During this period, all Arab states, excluding Syria, saw an increase of number of Facebook users in the country. Syria presents a curious case, where the number of Facebook users in the country has decreased by around 375 thousand during this period, while penetration rates have increased. This is due to the Number of Users Penetration mass population exodus and forced migration of millions of people out Average of the country, pushed by the intensifying military conflict between 2014 and 2016, which has resulted in a cumulative drop of at 11 least 11 percent of theChange population during the length of (May the conflict . 2016) in Facebook Penetration Rates 2014 and Oct As of early 2017, around 23% of Facebook users in the Arab region were based in Egypt, a percentage that has decreased slightly from 25% over the past five years. Saudi Arabia continues to increase its share 30%of Arab users, rising by 3% from 2014. However, the rising star in terms of largest increase of share was Algeria, which gained 4% of the total share of Facebook users in the region. By 2017, 24.1% 23.1% 12% of the total users in the region are now based in Algeria. The remaining countries in the region 25% 21.9% largely maintained their share of the total Facebook users in the Arab region with 20.0% slight fluctuations. 19.0% 18.5%(2017) DISTRIBUTION OF FACEBOOK USERS IN THE ARAB STATES

20%

Distribution of Facebook Users in Arab Region (2017)

16.0% 14.9%15.2% 13.5% 11.5%

15%

8.8% 7.7% Kuwait, 2% Qatar, 2% 5.6% 5.0% 4.5% 4.7% Remaining, 4% 3.4% 4.0% 4.0% 1.5% 3.1%

10% 5%

Libya, 2%

Sudan, 2% Lebanon, 2%

0%

Jordan, 4%

ia

Ira q ah ra in Li by a A lg er ia U A E Q at ar

an om ia or os O m P al an es ti S ne om a Le lia ba no K n uw D ait jib ou Tu ti ni si a E gy M p or t oc c Jo o rd an S au di

Tunisia, 4%

C

au M

B

Syria, 5%

rit

yr

an

S

ud

S

Y

em

en

Egypt, 23%

UAE, 6%

Morocco, 9%

Saudi, 13%

Number of Facebook Users in the ArabIraq, States 10% (Jan 2017)

Millions

Algeria, 12%

34.4

35

1111 Population and immigration estimates are based on data from the United Nations Department of Economic and Social Affairs: Population Division and UNESCO Institute for Statistics 2016

30

Social Media25 and the Internet of Things - Towards Data-Driven Policymaking in the Arab World: Potential, Limits and Concerns

20

37 17.2

18.7

Qatar, 2%

Libya, 2%

Remaining, 4%

Sudan, 2% Lebanon, 2%

Egypt, 23%

Jordan, 4%

Tunisia, 4% 2.1.2. Gender and Age Breakdowns of Facebook Users in the Arab Syria, 5% Region UAE, 6%

Social media users in the region remain largely youthful; this is a consistent finding over the past six years. On average, 64.3% During the past six of Facebook users in the Saudi, region 13% are now under 30 years old.Morocco, 9% years, there was a slow, However, this percentage has consistently witnessed slight but steady, uptake of dips over the past years. It went down from 67% two years Iraq, 10% Facebook by users over ago, and from 70% in 2012. WhileAlgeria, Facebook users in the Arab 12% the age of 30. region remain mostly youthful, the trend is moving towards a more balanced age breakdown in the region. During the past six years, there was a slow, but steady, uptake of Facebook by users over the age of 30. Kuwait remains the most ‘mature’ Facebook population in the region, where the percentage of those above 30 years old, reached up to more than 57% of all users, rising up from 54% two years ago. Somalia also remains the most youthful with 82% of the population under 30 years old.

Age Breakdown of Facebook Users in Arab Region (2017) BREAKDOWN OF FACEBOOK USERS BY AGE IN THE ARAB STATES KUWAIT

42.8%

BAHRAIN

57.2%

47.6%

UAE

52.4%

49.2%

QATAR

50.8%

51.2%

LEBANON SAUDI

48.8%

55.1%

44.9%

55.7%

44.3%

LIBYA

OMAN

57.6%

42.4%

58.2%

41.8%

JORDAN

TUNISIA

61.4%

38.6%

62.3%

37.7%

EGYPT

65.8%

IRAQ

COMOROS

PALESTINE

31.4%

69.1%

30.9%

69.2%

30.8%

70.2%

29.8%

70.7%

29.3%

MAURITANIA MOROCCO

YEMEN

DJIBOUTI ALGERIA

72.1%

27.9%

72.2%

27.8%

74.2%

SOMALIA 10%

20%

30%

15-29 Years (%)

TUNISIA

LEBANON

38

JORDAN

ALGERIA

COMOROS

25.8%

82.8% 0%

PALESTINE

34.2%

68.6%

55.4% 57.4% 58.4% 59.4% 61.8% 63.3%

40%

17.2% 50%

60%

70%

80%

90%

100%

Over 30 Years (%)

44.6% 42.6% 41.6% 40.6% 38.2% 36.7%


Breakdown 2017 (Total Levant Population)

67.1%

67.7%

Users over the age of 30 now constitute more than half of Facebook users in the three countries of Kuwait, Bahrain and the UAE, and almost half of all users in Qatar. There 24.6% seems to be continued GCC Facebook Gender maturity of social media users in these four Gulf Breakdown countries. Overall, compared to the rest of the Arab 2017 (Total GCC Population) region, the six GCC countries largely have more balanced breakdowns of young users who are under 30 and those who are overFemale 30 years old, indicating maturity of usage across age groups75.4% in society. Male This also may be aligned with the actual age breakdown of the population in these countries, where a large percentage of the labor force are expatriates, in contrast with the large number of young people in other countries in the region, usually young students or unemployed youth.

Facebook Age Breakdown 2017 in Arab Regions

(TotalUsers Regionalin Populations) Regional Age Breakdowns of Facebook the Gulf, North Africa and the Levant (2017)

Age Breakdown in the Arab Region 2017 (Total Regional Population) GCC’s Facebook Age Breakdown 2017 (Total GCC Population)

45.2% 54.8%

35.7%

64.3%

13-29 Years

Levant's Facebook Age Breakdown 2017 (Total Levant Population)

33.2%

North Africa's Facebook Age Breakdown 2017 (Total North African Population)

32.4%

66.8%

67.6%

Over 30 Years

Millions

In terms of gender balance, the social media gender gap in the Arab region continues to be persistent over the past six years. By 2017, on average, females make up 32.3% of the total Facebook users in the region, almost unchanged from 31.8% in 2014, and 33.4% in 2013. This continues to be lower than the global average which is much more balanced in terms of gender. This finding highlights existing societal constraints that women face in general in the region, including the barriers female face in terms of access to education and technology. This is also consistent with the findings highlighted by our research in the 3rd edition of the Arab Social Media report12. 12

However, there are noticeable changes in terms of gender breakdown in individual countries. 10 Palestine has become the country with most gender-balanced Facebook users in the Arab region with almost 45% of the users being female. It is followed by Tunisia (43% females), Lebanon (42%) 8 and Jordan (41%). Lebanon has dropped from the 1st to the 3rd rank over the past two years, while Yemen impressively jumped from the last spot in the region two years ago, to the 13th in 2017, 6 surpassing all Gulf countries, but Kuwait. 4 2 1212

The Arab Social Media Report (Vol. 3), The Role of Social Media in Arab Women’s Empowerment. Dubai School of Government. Dubai

0

Mar. ‘12

Mar. ‘13

Mar. ‘14

Mar. ‘15

Mar. ‘16


Mar. ‘17

39

15-29 Years (%) Total Facebook Users

30 Users Years (DAU) (%) DailyOver Active

Gender Breakdown of Facebook Users in Arab States (2017) PALESTINE

55.4%

TUNISIA

LEBANON

COMOROS LIBYA

42.6%

58.4%

JORDAN

ALGERIA

44.6%

57.4%

41.6%

59.4%

40.6%

DISTRIBUTION OF FACEBOOK USERS IN THE GCC 61.8%

38.2%

63.3%

36.7%

64.4%

35.6%

DJIBOUTI

65.6%

EGYPT

65.8%

Bahrain, 65.7%

SOMALIA

MOROCCO

34.4%

3%

34.3%

Oman, 5%

34.2%

66.0%

KUWAIT YEMEN

34.0%

Qatar, 7%

71.4%

28.6%

72.1%

MAURITANIA

27.9%

72.8%

27.2%

73.2%

26.8% Kuwait, 8%

BAHRAIN UAE

74.5%

IRAQ

Saudi, 52%

SAUDI

23.0%

77.6%

22.4%

QATAR

OMAN 0%

10%

20%

25.5%

77.0%

79.1%

20.9%

79.4%

20.6%

30%

40%

50%

Male

60%

70%

80%

90%

100%

UAE, 25%

Female

On a regional level, the Gulf region has the worst gender balance on Facebook compared to the other regions in the Arab World. Only 1 out of 4 users in the Gulf region on average is a woman (24.6%), compared to 1 out of 3 in the rest of the region on average. Indeed, Oman has dropped to the last spot in terms of gender balance of Facebook users in the Arab region, while five out of the bottom six Percentage of Facebook users who post in Arabic in the Arab States (% of total users) spots are occupied by GCC countries. North African countries have slightly better gender balance on Facebook compared to the regional average, with 36% of users being women. Meanwhile, the gender balance in the Levant region matches the regional average. 100%

10.0

4.2

9.2 18.7

8.3

Facebook Gender Breakdown 2017 in14.8 Arab Regions Regional Gender Breakdown of Facebook Users in the Gulf, North Africa and the Levant (2017) 90% (Total Regional Populations)

80% 70%

13.5

Gender Breakdown in the Arab Region

60% 2017 (Total Regional Population) 50%

-4.8

-15.8

40% 30%

-3.6 -6.3

20%

-1.9

-5.3

50.6 37.4

32.5

42.6

36.2%

North Africa's Facebook Gender Breakdown 2017 (Total North African Population)

22.0

2014 2016 63.8%

6.2

5.9 32.3% 10%

40 100%

en

pt

em

Y

E

gy

e

a

tin

GCC Facebook Gender Breakdown 2017 (Total GCC Population)

Male

67.1%

P

al

es

by

q

32.9%

Li

an

Ira

ia

er

rd

Jo

co

A

lg

a

M

or

oc

n

si ni

Tu

ia

Le

ba

no

di

an

rit

au

S

Levant's Facebook Gender Breakdown 2017 (Total Levant Population)

M

67.7%

au

ia

t

al

ai

uw

om

S

K

ti

an

m

O

in

D

jib

ou

ar

B

ah

ra

E A

at Q

U

C

om

or

os

0%

24.6%

75.4%

Female

Percentage of Facebook users who post in English in the Arab States (% of total users)


Facebook Age Breakdown 2017 in Arab Regions (Total Regional Populations)

-4.2

15-29 Years (%)

Over 30 Years (%)

2.1.3. Language Use: Arabic, English and French Language Use on PALESTINE TUNISIA Social Media in the Arab Region 55.4%

44.6%

57.4%

LEBANON

42.6%

58.4%

41.6%

59.4% ArabicJORDAN language use has grown across the region, excluding in theALGERIA six Gulf countries, each of 61.8% which has seen a drop in 63.3% COMOROSof people using the platform percentage in Arabic. In general, 64.4% LIBYA Arabic is becoming the dominant language on Facebook within 65.6% DJIBOUTI the Arab region while many people65.7% continue to use multiple SOMALIA languages when posting or reading65.8% social media material13. EGYPT

MOROCCO

40.6% 38.2%

The 36.7% North African region saw 35.6% the largest increase 34.4% in levels of usage of 34.3% Arabic language on 34.2% Facebook, while the Gulf 34.0% region saw a decrease 28.6% in the use of Arabic 27.9% language 27.2%

66.0%

YemenKUWAIT continues to be the top country71.4% in terms of Arabic use on Facebook (with 95% of users using Arabic), followed by 72.1% YEMEN Egypt (94%), Palestine (93.8%), Libya (93.3), Iraq (92.9) and 72.8% MAURITANIA 73.2% 26.8% BAHRAIN Jordan (90%). During the past two years, the largest increase 74.5% 25.5% UAE of percentage of users interacting in Arabic on Facebook was 77.0% 23.0% in TunisiaIRAQ (50%), followed by Algeria (43%), Morocco (37%), Lebanon (33%) and Somalia (22%). 77.6% 22.4% SAUDI Meanwhile, Oman saw the largest decrease in Arabic language use (16%) followed by the rest of the 79.1% 20.9% QATAR Gulf countries, Qatar (6%), Bahrain (5.3%), Saudi (4.8%), UAE (3.6%) and Kuwait (2%). 79.4% OMAN

20.6%

On a regional 0% level, the African region40% saw the of 90% usage100% of Arabic 10%North 20% 30% 50%largest 60%increase 70% in levels 80% language on Facebook, while the Gulf region saw a decrease in the use of Arabic language. English Male Female remains the dominant language on Facebook for users in all Gulf countries, with an increasing numbers of people who use English on the site in these countries. North African countries; Tunisia, Algeria, Morocco, Mauritania as well as Djibouti and Comoros, continue to prefer the use of French over Arabic and English.

Percentage of Facebook users who post in Arabic in theArab Arab States (% Who of totalUse users)Arabic Out of Arabic Language: Percentage of Facebook Users in the States Total Users 100%

14.8

90% 80% 70%

13.5

60% 50%

30%

-3.6 -6.3

20%

-1.9

-5.3

4.2

9.2 18.7

8.3

42.6

-4.8

-15.8

40%

32.5

50.6 37.4

10.0

2014

22.0

2016

6.2

5.9

10%

en

pt

em

Y

E

gy

e

a P

al

es

tin

q

by

Li

an

Ira

ia er

rd

Jo

co

A

lg

a M

or

oc

n

si ni

Tu

ia

Le

ba

no

di

an

rit

M

au

S

au

ia

t

al

ai

uw

om

K

S

ti

an

m

O

in

D

jib

ou

ar

ah

ra

E

at

A

U

Q

B

C

om

or

os

0%

1313 The same user may be accounted for more than once while calculating the breakdowns of Facebook users per language (once for each language they use).

Percentage of Facebook users who post in English in the Arab States (% of total users)

Social Media and the Internet of Things - Towards Data-Driven Policymaking in the Arab World: Potential, Limits and Concerns -4.2 100% 90%

11.0

3.9

2.4

4.0

41

90% 80%

3.9

0.2 11.0

-10.3 7.6

70%

60% Language: Percentage of Facebook Users in the Arab States Who Use English Out English English in the Arab States (% of total users) 21.3 of Total 50% Users Percentage of Facebook users who post in 2014 40%

-8.9

-9.5

30% 100% 20% 90%

0.04

10% 80%

2.3

2.6

2.6 -2.2

-4.3 -5.4

4.5

-5.2

2016

40%

0.04

10%

ia

ar

Q

om

al

E

at

A

in

U

ra

2014

S

B

ah

t

m O

uw

K

ba

Le

ai

n

di

no

ti

au

S

an

ou

21.3

2016

-4.3 -5.4

4.5

-5.2

2.3

2.6

2.6 -2.2

jib

rd

-8.9

-9.5

30% 20%

D

Jo

gy

E

tin

e

pt

q

M

P

al

Tu

es

Ira

si

a

a

ni

by

co

Li

ia

oc

or

M

au

rit

an

en

os

Y

em

ia

or

er

lg

om

A

C

50%

an

7.6

0% 70% 60%

3.9

0.2 11.0

-10.3

-4.2

4.0

2.4

ia

ar

om

at

al

E A

Q

U

in ra

S

B

ah

an

t

m O

uw

K

ba

Le

ai

n

di

no

ti

au

S

ou

D

jib

an

rd

Jo

gy

E

tin

e

pt

q

M

P

al

Tu

es

Ira

ni

si

a

a by

co

oc

Li

ia

or

M

au

rit

an

en

os

em

Y

or

om

C

A

lg

er

ia

0%

Percentage of Facebook users who post in French in the Arab States (% of total users)

French Language: Percentage of Facebook Users in the Arab States Who Use French Out of Total Users 100%

2.14

90%

-2.78

-13.30 -6.54 -7.72

80%

Percentage of Facebook users who post in French in the Arab States (% of total users)

70%

-0.03

60%

2014

50% 40% 30% 100%

2.14

os

a si

C

om

ni

or

ti ou

jib

D

Tu

ia

co

er

lg

M

M

A

oc

ia

an

n

-0.03

au

Le

ba

no

pt

a

gy

by

E

Li

ar

E

at Q

ia

A

al

U

an

om

S

rd

e

Jo

tin

t ai P

al

es

di

au

S

K

50%

uw

in ra

an

m

B

ah

Ira

O

en

Y

em

60%

q

-0.13 -0.22 -0.20 -0.31 0.05 -0.24 -0.35 -0.77 -0.65 -0.27 -0.20 0.03 -1.60

0% 70%

2016

-13.30 -6.54 -7.72

-0.97

rit

10% 80%

or

20% 90%

-2.78

2014

40%

2016

30%

os

si

C

om

ni

or

a

ti ou

D

jib

er

lg

A

oc

an

M

or

no

rit

au

M

Le

ba

gy

by

E

Li

at

Q

A

al

U

om

S

rd

tin

Jo

ai

es

P

al

au

uw

K

S

ra

m

ah

B

Ira

O

Y

em

2.1.4. Active Facebook Users in the Arab Region

Tu

ia

-0.97

n

pt

a

ar

E

ia

an

e

t

di

in

an

en

0%

q

-0.13 -0.22 -0.20 -0.31 0.05 -0.24 -0.35 -0.77 -0.65 -0.27 -0.20 0.03 -1.60

co

10%

ia

20%

For the first time in this series, we compare the percentages of actual active users of Facebook versus the total number of accounts. Monthly active users (MAU) is a measure of accounts that are actually accessed or used at least once a month, meanwhile, a daily active user (DAU) is someone that uses the account at least once a day. This measure provides additional insight on the activity of social media which may be important in different developmental, engagement or policymaking decisions. Percentage of Facebook Daily Active Users (DAUs) out of Total Users Qatar, dominates in both overall penetration rates and percentage of daily active Facebook users in the country out of the total population, followed closely by the UAE. Meanwhile, Jordan has the 3rd most active population on Facebook in the region. On average, the percentage of daily active users on the regional level is 53% of total. 100% 90%

4280% 70%


Percentage of Facebook Daily Active Users (DAUs) out of Total Users 62.8% 63.6% 64.1% 61.8% 62.7%

67.6%

Penetration of Facebook’s Total Users and Daily Active Users out of Total Population in the Arab States (Sorted by percentage of most active) Daily Active Facebook Users out of Total Facebook Users in Arab States (Penetration out of total population - sorted by most active)

Percentage of Facebook users who post in French in the Arab States (% of total users) 100% 90% 80% 100% 70% 90% 60% 50% 80% 40% 70% 30% 60% 20% 8.0% 10% 50% 2.9% 0%

94.3% 95.3%

36.5%

-0.03

C

ar

E

at

os

a

C

om

or

ti

si ni

Tu

ia

ou

D

A

lg

jib

er

co

ia

M

M

or

oc

n no

an

au

ba

Le

Daily Active Users (DAU)

rit

pt

a

gy

by

E

ar

Li

E

at Q

ia

A

al

U

an

S

om

e

Jo

rd

t

es

P

al

K

B

tin

ai

di

Total Facebook Users

uw

in

au

ra

S

an

m

O

ah

q Ira

en em Y

A

U

rit

au

S

M

-0.97

-0.13 -0.22 -0.20 -0.31 0.05 -0.24 -0.35 -0.77 -0.65 -0.27 -0.20 0.03 -1.60

0%

2016

Q

om ia or os D jib ou ti S yr ia Ira M q or oc co E gy pt O m an A lg er P ia al es tin e Li by a S au d Le ba i no Tu n ni si a K uw ai B ah t ra i Jo n rd an

an

an

ia

ud

al

en

om

em 10%

33.7%

7.6% 7.7% 10.6% 9.3% 2014 3.7% 4.1% 4.3% 4.3% 10.7% 13.5% 19.4% 20.1% 20.3% 20.4% 20.8% 20.8% 23.7% 28.0% 35.1% 35.2% 37.4% 40.1% 41.3% 60.4% 64.5%

S

Y

20%

58.1%

55.9% 55.3% -13.30 -6.54 -7.72

49.0%

20.1%

40% 30%

40.9% 42.7% 39.9% 38.4% 36.8%

-2.78 72.7% 2.1471.0% 65.8%

These measures and the contrasts they provide are useful when assessing feasibility of engagement mechanisms (by governments or businesses) across social media platforms. For example, while Saudi ranks 6th in the region in terms of penetration of total number of Facebook accounts, it ranks DailyofActive Users out of Total Users in the GCC 17th in terms of Facebook the percentage daily active users. In another example, while Tunisia ranks 8th in (Penetration out of total population) the region in terms of penetration rates of Facebook users in the country, it is almost tied as 2nd in the region alongside the UAE on the percentage of most active users out of total users. Take a third example, around 73% of the population of Bahrain has Facebook accounts, only 40% of the total 100% population is active at least once a day, representing 55% of the total Facebook users in the country. 90%

95.3%

94.3%

80%

Percentage of Facebook Daily Active Users out of Total Facebook Users in the Arab States Percentage of Facebook Daily Active Users (DAUs) out of Total Users 70% 72.7%

71.0%

60% 50%

58.1%

40%

40.9%

30% 100% 20% 90% 10% 80%0%

Oman

Saudi

60% 50% 40% 37.1%

40.9%

37.4%

40.1%

Kuwait

Bahrain

28.0%

20.4%

70%

64.5%

60.4%

46.8%

UAE

Qatar 67.6%

62.8% 63.6% 64.1% 61.8% 62.7%

53.1% 55.2% 55.3% 48.8% 49.9% 52.3% 52.7% Daily Active Users 48.6% 48.7% Users 48.3% Facebook 48.2% Total

(DAU)

30% 20% 10%

ar at

E

Q

A

a

U

si

ni

an

Tu

Jo

rd

n no

ba

es

tin

e

pt

al

E

gy

in

ah

ra

ti

t

ou

D

jib

ai

K

uw

co

oc

an

or

O

m

ia er

q A

lg

Ira

om

al

ia

a

Li

by

di

S

au

os

or

ia

Bahrain, 3%

Le

P

B

M

S

an

om

DISTRIBUTION OF FACEBOOK USERS IN THE GCC

C

rit

au

M

Y

em

en

0%

Oman, 5%


Qatar, 7%

Daily Active Facebook Users out of Total Facebook Users in Arab States

43

2.2. Mapping Twitter in the Arab World Highlights – Twitter in the Arab Region •

Total Users: The total number of monthly active Twitter users in the Arab region is estimated to reach 11.1 million in March 2017, almost doubling up from 5.8 million three years ago.

•

Largest Share: 29% of all active Twitter users in the Arab region are in Saudi Arabia, a share which has dropped from 40% in 2014. It is followed by Egypt with 18% share and Algeria and the UAE with 9% each.

•

Top Countries: In terms of penetration rates, Bahrain and Kuwait have the highest penetration rates of active Twitter users in the Arab region.

•

New Users: Over the past two years, Algeria witnessed the largest growth in terms of number of new active users on Twitter in the region, followed by Egypt and the UAE.

•

Monthly Tweets Volume: The estimated number of tweets produced by active Twitter users in the Arab region in March 2016 was around 849.1 million tweets, an increase of more than 59% in volume over two years.

•

Tweets per day: Collectively, the Arab world generates 27.4 million tweets per day up from 17.2 million tweets per day two years earlier.

•

Most Tweets: More than half of all tweets generated in the Arab region are coming from two countries: Saudi Arabia generates 33% of all tweets in the Arab region and Egypt generates another 20% of all tweets in the region.

•

Most Active: In terms of tweets per day, Kuwait has the most active Twitter users, producing an average of 4.2 tweets per day. It is followed by Saudi Arabia and Palestine.

•

Gender Breakdown: The percentage of female Twitter users in the Arab region was 32.4% in 2016, dropping from 36.6% two years earlier.

2.2.1. Twitter Growth and Penetration Rates in the Arab Region The number of monthly active Twitter users in the Arab region users according to our research is around 10.8 million in early 2017, a number we estimate to reach 11.1 million in March 2017, almost doubling up from 5.8 million three years earlier. As officially defined by Twitter, a monthly “active user” is someone who logs in (but does not necessarily tweet) once a month. These estimations are based on actual current and historical data collected by the Arab Social Media Report series between 2012 and 2016. The methodology used is detailed in Annex 1. Briefly, it consists of randomly sampling a very large number of Twitter data in each country captured across a certain period, and using this sample to extrapolate and estimate the active Twitter population. The volume of tweets, language, gender breakdown and top trends throughout March 2016 were also estimated using the same dataset. The number of total Twitter accounts in the region is of course higher than the number presented above, which represents only the active users. Based on a regional survey of social media users, our estimation of accounts that are not classified as monthly active users is around 51% on Twitter. This suggests that the number of total Twitter accounts in the region could be as high 16.3 million by 2017. This includes accounts that are totally inactive or dormant, accounts active on less than monthly basis, accounts which are only used for reading tweets or communicating through direct messaging, as well as those accounts that tweet at least once a month which are analyzed here.

44


Twitter Growth Rates in the Arab Region between 2012 and 2017 (Monthly Active Users) 3 14% 12% Millions

12 10%

2.5 2

9.1%

8.7%

8.1%

8% 10

1.5

6.4%

6%

4.6%

8 4%

2%6

2.4%

0.5

om ia or o M or s oc co D jib ou ti Ira q Li by Tu a ni si a S yr ia E gy pt A lg er i Jo a rd Le an ba no n O m P al an es tin e S au di U A E Q at ar K uw ai B ah t ra in

0

an

en

2.0% 1.6% 1.6% 1.6% 1.8%

1

2.8%

C

rit

M

2

au

Y

em

ia

ud

al

S

om

an

0.9% 1.0% 0.6% 0.6% 0.6% 0.6% 0.2% 0.3%

0% 4 S

12.9%

12.6%

Millions

States (March 2016)

Active Twitter Users

0

Mar. ‘12

Mar. ‘13

Twitter Penetration

Mar. ‘14

Mar. ‘15

Mar. ‘16

Mar. ‘17

DISTRIBUTION OF ACTIVE TWITTER USERS IN THE ARAB REGION

Distribution of Active Twitter Users in the Arab States

Number of Active Twitter Users in the Arab States (Mar 2016) Lebanon, 2%

Bahrain, 2%

Jordan, 2% Tunisia, 2%

Remaining, 5%

2,620,000

Oman, 2%

Qatar, 2% Morocco, 2%

Palestine, 3%

Saudi, 29%

1,680,000

Syria, 3% Iraq, 4%

di

au

gy

pt

802,000

er

ia

E

503,000

U

A

t

363,000

uw

ai

q

306,000

ia

Ira

305,000

yr

223,000

220,000

212,000

187,000

187,000

170,000

152,000

114,000

103,000

23,800

17,000

8,510

os or

Algeria, 9%

ou t om i a M au lia rit an ia Li by a S ud an Y em Le en ba no n B ah ra in Jo rd a Tu n ni si a O m an Q a M tar or oc co P al es tin e

4,730

180,000

UAE, 9% Egypt, 18%

811,000

Kuwait, 5%

S

E

lg

A

K

S

S

jib

D

C

om

Saudi Arabia continues to lead with the highest number of active Twitter users in the region by a wide margin. With more than 2.6 million active users, Twitter users in Saudi Arabia account for 29% of all active users in the Arab region. Saudi Arabia is followed by Egypt; the country with the second highest number of active users with around 1.7 million. Algeria, however, has gained the largest number of new users since 2014, with around 774,000 new users, followed by Egypt (590,000) and the UAE (300,000). Overall, all countries, other than Lebanon, gained new active users on Twitter Change in Twitter Users between compared to two years earlier. Penetration Rates of 2014-2016 Active Twitter Users in the Arab States (March 2016)

Algeria

2063%; 773,500

Egypt UAE

60%; 300,000

14%Saudi

9%; 206,000

179,000 Policymaking in the Arab World: Potential, Limits and Concerns Social Media and the Internet of Things - Towards97%; Data-Driven Iraq

12%Syria Kuwait

10%

12.6%

125%; 170,000 46%; 159,000

8.7%

9.1%

12.9%

3 2.5 2

Millions

54%; 590,000

45


Twitter Penetration

Penetration Rates of Active Twitter Users in the Arab States (March 2016)

DISTRIBUTION OF ACTIVE TWITTER USERS IN THE ARAB REGION

14% 12% 10%

Lebanon, 2%

8%

8.1%

Bahrain, 2%

4% 2%

0.9% 1.0% 0.6% 0.6% 0.6% 0.6% 0.2% 0.3%

2.0% 1.6% 1.6% 1.6% 1.8%

2.4%

2.8%

2

Qatar, 2% Morocco, 2%

0.5 0

om ia or o M or s oc co D jib ou ti Ira q Li by Tu a ni si a S yr ia E gy pt A lg er i Jo a rd Le an ba no n O m P al an es tin e S au di U A E Q at ar K uw ai B ah t ra in

en

2.5

1

Oman, 2%

Syria, 3%

an

Iraq, 4%

C

au

M

3

1.5

Palestine, 3%

rit

an

Y

em

ia

ud

al

S

om S

9.1%

4.6%

Saudi, 29%

0%

8.7%

12.9%

Jordan, 2%6.4% Tunisia, 2%

Remaining, 5%

6%

12.6%

Millions

Penetration Rates of Active Twitter Users in the Arab States (March 2016)


Twitter Penetration Kuwait, 5%

UAE, 9%

In terms of penetration rates, Bahrain now leads the region, barely Egypt, 18%

overtaking Kuwait, the country that had the highest penetration The Gulf countries 9% rates in the region betweenDISTRIBUTION 2012 and 2014. The Algeria, penetration OF ACTIVE TWITTER USERS IN continue to have the rates of both countries of active Twitter THE users are REGION around 13%. ARAB highest penetration rates In general, the Gulf countries continue to have the highest of active Twitter users in penetration rates of active Twitter users in the Arab region, the Arab region, holding holding the top five ranks. PalestineLebanon, is the only non-GCC Bahrain, 2%country 2% the top five ranks Jordan, 2% in the top 7 ranked countries in terms of penetration rates. Seven Remaining, 5% countries in the Arab region still have Twitter penetration ratesTunisia, 2% Oman, 2% below 1% of the population. Qatar, 2% Morocco, 2%

Palestine, 3%

Saudi, 29%

New Active Twitter Users in the

Change in Twitter Users between Syria, 3% 2014-2016 Arab States (Mar 2014 and Mar 2016) Iraq, 4%

Algeria

Kuwait, 5% 54%; 590,000

Egypt UAE

60%; 300,000

Saudi Iraq

97%; 179,000

Syria

125%; 170,000

102%; 154,000

Tunisia

404%; 149,900

Algeria, 9%

162%; 131,200

Oman

189%; 117,800

Bahrain

96%; 108,000

Qatar

207%; 76,900

Sudan

134%; 58,900 16%; 26,000

Yemen

19%; 24,000

Mauritania

380%; 18,840

Djibouti

UAE, 9%

191%; 146,300

Morocco

Libya

9%; 206,000

46%; 159,000 Egypt, 18%

Kuwait Palestine

Jordan

688%; 7,430

Comoros

346; 3,670

Somalia

14%; 2,100

Lebanon

-4%; -8,000

Change in Twitter Users between 2014-2016

Algeria

46

2063%; 773,500

Egypt UAE Saudi

2063%; 773,500

54%; 590,000 60%; 300,000 9%; 206,000


Palestine

102%; 154,000

Tunisia

404%; 149,900 191%; 146,300

Morocco

162%; 131,200

Oman

189%; 117,800

Bahrain

108,000 2.2.2.QatarDemographic96%;Breakdown of Twitter Users in the Arab Region 207%; 76,900

Sudan

134%; 58,900 Libya With regards to the gender breakdowns, our findings indicate that roughly a third (32.4%) of Arab 16%; 26,000 Jordan Twitter users are females, which is lower than the global average, and in line with what we see 19%; 24,000 Yemen regarding gender 380%; breakdown on other social media platforms. Bahrain now has the most gender 18,840 Mauritania balanced active Twitter users in the Arab region, followed by Algeria and Lebanon, which dropped 688%; 7,430 Djibouti fromComoros the top spot two years earlier. Yemen continues to have the smallest portion of female out of the 346; 3,670 14%; 2,100 total Somalia active Twitter users. -4%; -8,000

Lebanon

GENDER BREAKDOWN OF TWITTER IN THE ARAB STATES Gender Breakdown of Active Twitter Users in ArabUSERS Region

44% 43% 42% 40% 39% 38% 38% 37% 33% 33% 33% 32% 31% 29% 29% 28% 27% 26% 25% 23% 18% 17%

56% 57% 58% 60% 61% 62% 62% 63% 67% 67% 67% 68% 69% 71% 71% 72% 73% 74% 75% 77% 82% 83%

BAHRAIN ALGERIA LEBANON COMOROS MOROCCO UAE TUNISIA PALESTINE KUWAIT JORDAN DJIBOUTI QATAR SAUDI OMAN EGYPT SOMALIA SYRIA LIBYA MAURITANIA SUDAN IRAQ YEMEN 0%

10%

20%

30%

40%

Male

50%

60%

70%

80%

90%

100%

Female

Twitter Usage by Device Across the Arab States

Desktop 9.57%

Other 10.80%

Mobile 79.62%


47

16%

Arabic English 2.2.3. Tweets Volume across the Arab Region

Other Arabic 72%

The estimated number of tweets produced monthly by active Twitter users in the Arab region was around 849.1 million tweets More than half of all in March 2016, an increase of more than 59% in volume since two tweets generated in the years earlier. Collectively, the Arab world generates an average Arab region are coming of 27.4 million tweets a day, up Arabic from 17.2 million tweets Other from two countries: English per day two years TWEETS earlier. MoreVOLUME than half ofACROSS all tweets generated in the and16) Egypt. THE ARAB REGIONSaudi (MAR Arab region are coming from two countries: Saudi and Egypt. Saudi generates more than 32% of all tweets in the Arab region while Egypt generates another 18%. Expectedly, countries with the highest numbers of active Twitter users account for the majority of tweets in the region. Saudi Arabia, alone, produced around 274 million tweets per month, while Egypt produced around 152 million,30followed by Algeria (71 million) and the UAE (70 million). Millions

TWEETS VOLUME ACROSS THE ARAB REGION (MAR 16)

29 28

Tweets Volume Across the Arab Region (Mar 2016 as an example) 27 26

M 16 ar 1 M 6 a 5 r1 M 6 a 6 r1 M 6 a 7 r1 M 6 a 8 r1 M 6 a 9 r1 M 6 10 ar 1 M 6 11 ar M 16 12 ar M 16 13 ar 1 M 6 14 ar M 16 15 ar 1 M 6 16 ar M 16 17 ar M 16 18 ar M 16 a 19 r 1 M 6 20 ar 1 M 6 21 ar M 16 22 ar M 16 23 ar M 16 24 ar M 16 25 ar 1 M 6 26 ar M 16 27 ar M 16 28 ar M 16 29 ar M 16 30 ar M 16 31 ar M 16 ar 16

ar

4

26

3

M 2

1

M

ar

27

16

Millions

25 30 24 29 23 28

25 24

M 16 a 4 r1 M 6 a 5 r1 M 6 a 6 r1 M 6 a 7 r1 M 6 a 8 r1 M 6 a 9 r1 M 6 10 ar 1 M 6 11 ar M 16 12 ar M 16 13 ar 1 M 6 14 ar M 16 15 ar 1 M 6 16 ar M 16 17 ar M 16 18 ar M 16 a 19 r 1 M 6 20 ar 1 M 6 21 ar M 16 22 ar M 16 23 ar M 16 24 ar M 16 25 ar 1 M 6 26 ar M 16 27 ar M 16 28 ar M 16 29 ar M 16 30 ar M 16 31 ar M 16 ar 16

ar

3

M 2

1

M

ar

16

23

Distribution of Tweets Across the Arab States (Mar 2016)

Mauritania, 0.16% Somalia, 0.15%

Djibouti, 0.1% Comoros, 0.04%

Saudi, 32.3% Mauritania, 0.16% Somalia, 0.15%


Saudi, 32.3%

Egypt, 17.9%

Sudan, 0.8% Libya, 1.1% Yemen, 1.4% Jordan, 1.5% Morocco, 1.6% Tunisia, 1.8% Oman, 1.8% Lebanon, 1.9% Bahrain, 2.0% Qatar, 2.3% Iraq, 2.6% Sudan, 0.8% Syria, 2.8% Libya, 1.1% 3.7% Yemen,Palestine, 1.4% Jordan, 1.5% Morocco, 1.6% Tunisia, Kuwait, 7.7%1.8% Oman, 1.8% Lebanon, 1.9% Bahrain, 2.0% Qatar, 2.3% Iraq, 2.6% UAE, 8.1% Syria, 2.8% Algeria, 8.3%

Palestine, 3.7%

Kuwait, 7.7%

48

UAE, 8.1% Egypt, 17.9% Algeria, 8.3%


Monthly Tweets Volume in the Arab States (march ‘16)

Tweets Volume Generated in the Arab States per Month (Mar 2016)

Comoros Djibouti

Somalia

Mauritania Sudan

Libya

Yemen

Jordan

Morocco

Tunisia Oman

Lebanon

Bahrain

Qatar Iraq

Syria

Palestine Kuwait

UAE

Algeria

Egypt Saudi

350,389 541,654 1,284,667 1,360,371 7,183,102 9,105,048 12,111,426 12,328,526 13,510,572 15,401,976 15,483,768 15,714,161 16,727,258 19,778,519 21,799,048 23,370,967 31,361,315

65,804,053 69,187,285 70,813,468

151,935,760

273,956,200


Millions

2.2.4. Twitter Activity across the Arab Region 10

In terms of country-level activity on Twitter, the 22 Arab 9 countries can be classified into three distinct groups: 1) Only the following six Saudi 8 Countries producing more than 1 million tweets a day, 2) Arab countries produce countries producing between 100,000 and 1 million tweets, more than 1 million 7 Egypt tweets a day: Saudi, and 3) those producing less than 100,000 tweets a day. Only Egypt, Algeria, Kuwait, the following six Arab countries produce more than 1 million 6 Algeria UAE and Palestine. tweets a day: Saudi, Egypt, Algeria, Kuwait, UAE and Palestine. 5 The majority of countries in the region produce between 200 Kuwait and 800 4 thousand tweets a day. Mauritania, Somalia, Djibouti and Comoros each produce less than 100,000 tweets a day. On average, Saudi active Twitter UAE 3 users tweet around 8.8 million tweets a day, followed by Egypt with 4.9 million tweets a day and Algeria2 with around 2.3 million. Palestine 1

2

1

M

ar M 16 3 ar 1 M 6 a 4 r1 M 6 5 ar 1 M 6 a 6 r1 M 6 7 ar 1 M 6 8 ar 1 M 6 a 9 r1 M 6 10 ar M 1 11 ar 6 M 1 12 ar 6 M 16 13 ar M 1 14 ar 6 M 1 15 ar 6 M 1 16 ar 6 M 1 17 ar 6 M 1 18 ar 6 M 16 19 ar M 16 20 ar M 16 21 ar M 1 22 ar 6 M 1 23 ar 6 M 1 24 ar 6 M 16 25 ar M 16 26 ar M 1 27 ar 6 M 16 28 ar M 1 29 ar 6 M 1 30 ar 6 M 1 31 ar 6 M 16 ar 16

0

Thousands

Twitter Activity in Arab States with Less Than 1 Million Avg. Daily Tweet (Mar 2016)

900


800

49 Syria Iraq

1 1 MM a a 2 2 r 1r 1 MM 6 6 3 3ar 1ar 1 MM 6 6 a a 4 4 r 1r 1 MM 6 6 5 5ar 1ar 1 MM 6 6 a a 6 6 r 1r 1 MM 6 6 7 7ar 1ar 1 MM 6 6 8 8ar 1ar 1 MM 6 6 a a 9 9 r 1r 1 MM 6 6 10 10ar ar M M16 16 11 11ar ar M M16 16 12 12ar ar M M16 16 13 13ar ar M M16 16 14 14ar ar M M1 1 15 15ar a6r 6 M M16 16 16 16ar ar M M1 1 17 17ar a6r 6 M M1 1 18 18ar a6r 6 M M16 16 19 19ar ar M M16 16 20 20ar ar M M16 16 21 21ar ar M M1 1 22 22ar a6r 6 M M1 1 23 23ar a6r 6 M M16 16 24 24ar ar M M16 16 25 25ar ar M M16 16 26 26ar ar M M16 16 27 27ar ar M M16 16 28 28ar ar M M16 16 29 29ar ar M M16 16 30 30ar ar M M16 16 31 31ar ar M M16 16 ar ar 16 16

Thousands Thousands 1 1 MM a a 2 2 r 1r 1 MM 6 6 3 3ar 1ar 1 MM 6 6 a a 4 4 r 1r 1 MM 6 6 5 5ar 1ar 1 MM 6 6 a a 6 6 r 1r 1 MM 6 6 7 7ar 1ar 1 MM 6 6 8 8ar 1ar 1 MM 6 6 a a 9 9 r 1r 1 MM 6 6 10 10ar ar M M1 1 11 11ar a6r 6 M M1 1 12 12ar a6r 6 M M16 16 13 13ar ar M M1 1 14 14ar a6r 6 M M16 16 15 15ar ar M M1 1 16 16ar a6r 6 M M1 1 17 17ar a6r 6 M M1 1 18 18ar a6r 6 M M16 16 19 19ar ar M M16 16 20 20ar ar M M16 16 21 21ar ar M M1 1 22 22ar a6r 6 M M1 1 23 23ar a6r 6 M M1 1 24 24ar a6r 6 M M16 16 25 25ar ar M M16 16 26 26ar ar M M1 1 27 27ar a6r 6 M M16 16 28 28ar ar M M1 1 29 29ar a6r 6 M M1 1 30 30ar a6r 6 M M1 1 31 31ar a6r 6 M M16 16 ar ar 16 16

Millions Millions

Monthly Tweets Volume in the Arab States (march ‘16) Twitter Activity in Arab States Producing More than 1 Million Tweets Daily on Average (Mar 2016)


50 10 10

9 9

8 8

Saudi Saudi

7 7

Egypt Egypt

6 6

5 5

Algeria Algeria

4 4

Kuwait Kuwait

3 3

UAE UAE

2 2

Palestine Palestine

1 1

0 0

Twitter Activity in Arab States Producing Less than 1 Million Tweets Daily on Average (Mar 2016) Twitter Activity in Arab States with Less Than 1 Million Avg. Daily Tweet (Mar 2016) Twitter Activity in Arab States with Less Than 1 Million Avg. Daily Tweet (Mar 2016)

900 900

800 800

700 700

600 600

500 500

400 400

300 300

200 200

100 100

Syria Syria Iraq Iraq Qatar Qatar Oman Oman Bahrain Bahrain Lebanon Lebanon Tunisia Tunisia Morocco Morocco Jordan Jordan Yemen Yemen Libya Libya Sudan Sudan Mauritania Mauritania Somalia Somalia Djibouti Djibouti Comoros Comoros

0 0

Top Twitter Country Hashtags in the Arab Twittersphere (Mar 2016) Top Twitter Country Hashtags in the Arab Twittersphere (Mar 2016)


Egypt, 17.9% Algeria, 8.3%

On average, active Twitter users in the Arab region tweet around 2.6 tweets a day. Kuwait has the most active Twitter users, in terms of tweets per day, producing an average of 4.2 tweets per day, constituting the largest average of Tweets per day in the region by a relatively wide margin. It is followed by Saudi (3.4 tweets per user per day) and Palestine (3.3 tweets per user per day). Mauritania has the least active twitter users with 1.8 tweets per day.

On average, active Twitter users in the Arab region tweet around 2.6 tweets a day

Tweets Per User Per Day in the Arab States (Mar 2016)

4.5

4.2

4 3.5 3 2.5 1.8

2

2.4

2.4

2.1

2.1

2.0

2.0

1.9

2.4

2.9

2.9

2.9

2.8

2.8

2.7

2.6

2.5

3.4

3.3

3.0

3.0

1.5 1 0.5 t ai w

i ud

Ku

e in

Sa

n

st

n

ai

Pa

le

hr

Ba

no

Le

ba

yp

t

ar

Eg

at

a

Tweets Per User Per Day

Q

ria

by Li

ge

a

AE

Al

U

si ni

en

Tu

m

ria

Ye

Sy

ia al m

os or

So

m

an

om

O

an

ti ou

rd Jo

n

jib

da

D

co

Su

q

oc or

Ira

M

C

M

au

rit

an

ia

0

Average

2.2.5. Twitter Usage Trends in the Arab Region Across the region, the top country hashtags used in terms of how frequently they are mentioned in tweets, are illustrated in Most discussions on figure below. The top trending country hashtags, were estimated twitter generated in the during the month of March 2016 taking into account all mentions region were taking place of the country names in hashtags in both Arabic and English. in Arabic During this period, hashtags using the different forms of writing Saudi’s name in Arabic (‫ )الســعودية‬were by far the most frequent, with an estimated 261 million tweets from around the region. This is followed with hashtags of Syria in Arabic (‫)ســوريا‬, then Yemen (‫)اليمــن‬, Kuwait (‫ )الكويــت‬and the UAE (‫)اإلمــارات‬. This indicates that most discussions on twitter generated in the region were taking place in Arabic; a growing trend in terms of language use in the region. Across the top ten most frequent hashtags of country names in the region, only two were in English: Yemen and Saudi. This was followed by Syria in English.


51

31

29 30

26 27 28

24 25

22 23

20 21

18 19

16 17

14 15

12 13

9 10 11

8

7

6

5

4

2

3

1

Comoros

M ar M 16 3 ar 1 M 6 a 4 r1 M 6 5 ar 1 M 6 a 6 r1 M 6 7 ar 1 M 6 8 ar 1 M 6 a 9 r1 M 6 10 ar M 16 11 ar M 16 12 ar M 16 13 ar M 16 14 ar M 1 15 ar 6 M 16 16 ar M 1 17 ar 6 M 1 18 ar 6 M 16 19 ar M 16 20 ar M 16 21 ar M 1 22 ar 6 M 1 23 ar 6 M 16 24 ar M 16 25 ar M 16 26 ar M 16 27 ar M 16 28 ar M 16 29 ar M 16 30 ar M 16 31 ar M 16 ar 16

0

2

1

Of a list of all country hashtags included in tweets in Arabic and English combined14, those mentioning Saudi (in either language) constitute 27% of all tweets. This is three times the number of those mentioning Yemen or Syria (9% each), which come second and third. The UAE and Egypt follow next. These findings suggest that discussions on Twitter in the region (indicated by using country hashtags) is partly related to ongoing military in orthe political conflicts in(Mar the 2016) region. Top Twitter Country Hashtags Arab Twittersphere

Top Twitter Country Hashtags in the Arab Twittersphere (Mar 2016)

Millions

Millions

Top Country Hashtags Used Across the Arab Region (Mar 2016) 300 250 200 300 150 250 100 200 50 150

ria #S y

#S au di

#Y em

en

‫اﻟ‬# ‫ﻌﺮا‬ ‫ق‬

‫ﻗ‬# ‫ﻄﺮ‬

‫ﻣ‬# ‫ﺼﺮ‬

# ‫ﻻﻣ‬ ‫ا‬ ‫ﺎرا‬ ‫ت‬

‫ﺖ‬ ‫ﻮﻳ‬

‫ﻟﻜ‬ ‫ا‬#

# ‫ﻤﻦ‬ ‫اﻟﻴ‬

# ‫ﺳﻮ‬ ‫رﻳﺎ‬

‫ﺴﻌ‬

‫ﻳﺔ‬

‫ﻮد‬

50

‫اﻟ‬#

0 100

ia yr #S

di #S

em #Y

au

en

‫اﻟ‬# ‫ﻌﺮا‬ ‫ق‬

‫ﻗ‬# ‫ﻄﺮ‬

‫ﻣ‬# ‫ﺼﺮ‬

‫ﻻﻣ‬ ‫ﺎرا‬ ‫ت‬

‫ﺖ‬

‫ﻮﻳ‬

‫ﻟﻜ‬

‫ا‬#

‫ا‬#

# ‫اﻟﻴ‬ ‫ﻤﻦ‬

# ‫ﺳﻮ‬ ‫رﻳﺎ‬

‫ﻳﺔ‬

‫ﻮد‬

‫ﺴﻌ‬

‫اﻟ‬#

0

Top Country Hashtags inCountry Each of the Arab States 2016) (Mar 2016) TopUsed Twitter Hashtags in the Arab (Mar Twittersphere

Top Twitter Country Hashtags in the Arab Twittersphere (Mar 2016) Oman, 2%

Libya, 2%

Sudan, 1%

Tunisia, 2%

Somalia, 0.37%

Djibouti, 0.18% Mauritania, 0.08% Comoros, 0.03%

Somalia, 0.37%

Djibouti, 0.18% Mauritania, 0.08% Comoros, 0.03%

Morocco, 2% Bahrain, 2%

Palestine, 2% Libya, 2% Oman, 2% Jordan, 3% Morocco, 2% Lebanon,Bahrain, 3% 2%

Sudan, 1%

Tunisia, 2%

Saudi, 27%

Palestine, Algeria, 3% 2% Jordan, 3%

Iraq,3% 5% Lebanon,

Saudi, 27%

Algeria, 3% Qatar, 5% Yemen, 9%

Iraq, 5% Kuwait, 7% Qatar, 5% Kuwait, 7%

Egypt, 7%

Egypt, 7%

Syria, 9% Yemen, 9% UAE, 8% Syria, 9% UAE, 8%

1414 All possible spellings and writing styles of country names in both English and Arabic were analyzed.

52


Mauritania Tunisia Morocco Somalia

12.6

70.1 70

15.9

64.5

15

19.9

64.1

17.3

15

19.6

16

In terms of language use on20Twitter, 30tweets in Arabic50account60 for over 72% of80 all tweets generated 0 10 40 70 90 100 in the Arab world. This percentage has been almost stable (around 2% lower) since March 2013. English comes next with only 16% of tweets, while around 12% used other languages while tweeting. Arabic dominates the tweets produced in Saudi Arabia, Kuwait and Egypt, three of the countries that Desktop/ Laptop Other generate the largest numberMobile of tweets in the region. Meanwhile, Saudi has maintained its position as the generator of the highest percentage of tweets in Arabic in the region (92%), followed by Kuwait (85%) and Oman (81%).

Language Use on Twitter across the Arab Region (Mar 2016)

Other 12%

English 16%

Arabic 72%

Arabic

English

Other

Millions

TWEETS ACROSS THE ARAB At least four out of fiveVOLUME active Twitter users in the Arab regionREGION (MAR 16) tweet using a mobile device according to data collected for this Using mobile devices report. This reflects the growth of mobile connectivity around the to tweet is the dominant region in general as well as the nature of the platform. Only 10% method in every single of tweets are generated through desktop or laptop computers in Arab country the region. On a country level, users in the Gulf region are using 30 Twitter mostly through mobile, with the top five countries in terms 29 of tweeting using mobile being Gulf countries. Oman leads in terms 28 of percentage of mobile users of the platform with around 90% of total users. Meanwhile, the countries ranked lowest in terms of using mobile devices to tweet are dominated by North African 27 countries. This being said, using mobile devices to tweet is the dominant method in every single Arab 26 country. The country that uses mobile devices the least to tweet (Somalia), at least 64% of tweets are 25 generated through mobile devices. actually 24

M 16 ar 1 M 6 a 5 r1 M 6 a 6 r1 M 6 a 7 r1 M 6 a 8 r1 M 6 a 9 r1 M 6 10 ar 1 M 6 11 ar 1 M 6 12 ar M 16 13 ar 1 M 6 14 ar M 16 15 ar 1 M 6 16 ar M 16 17 ar M 16 18 ar M 16 a 19 r 1 M 6 20 ar 1 M 6 21 ar M 16 22 ar M 16 23 ar M 16 24 ar M 16 25 ar 1 M 6 26 ar M 16 27 ar M 16 28 ar M 16 29 ar M 16 30 ar M 16 31 ar M 16 ar 16

ar

4

3

M 2

1

M

ar

16

23


53

Desktop 9.57%

Use of Twitter by Device Across the Arab Region (Mar 2016) Twitter Usage by Device Across the Arab States Other 10.80%

Desktop 9.57%

Other 10.80%

Mobile 79.62%

Mobile 79.62%

Desktop/ Laptop

Other

Mobile

Desktop/ Laptop

Other

Mobile

Twitter Usage Across by Devicethe in the Arab States(%) (Mar 2016) Breakdown of Uses of Twitter by Device Arab States

Oman Bahrain Qatar Saudi Kuwait Comoros Oman Sudan Bahrain Iraq Qatar Palestine Saudi UAE Kuwait Lebanon Comoros Djibouti Sudan Jordan Iraq Egypt Palestine Syria UAE Libya Lebanon Yemen Djibouti Algeria Jordan Mauritania Egypt Tunisia Syria Morocco Libya Somalia Yemen Algeria 0 Mauritania Tunisia Morocco Somalia 0

89.8 86

7.18

85.9

8

Twitter Usage by Device 86 in the Arab States(%)

9.51 6.79 5.69.614.6 7.59 6.82 9.09 7.18 10.31

77.4 86 85.9

9.72

75.9 85.5 75.3 83.7

11.8

74.9 82.8 74.880.6 77.4 73.8

12

12.6

64.5 73.8 64.1 73.6

15.9

72.3

50

60

70

70

Mobile 10

20

64.1 30

Mobile

14.1 19.9 8.05 8018.6

15.9

Desktop/ Laptop 40

15 11.7 12.3

12.6

70.1 64.5

12

50

Desktop/ Laptop

15

19.9 Other

60

70

9.61 13.1 10.31 13.5 12.88 13.9

7.59

13.2 14.1 11.8 8.05 16.2 18.6

75.3 71.8 74.9 70.1 7074.8

8.69 12.3 9.51 8.5

5.81

16.2 6.79

11.7 9.09 9.72 12.3

76.5 73.6 75.9 72.3

71.8 40

7.8712.886.13 6.1 8 10.3

13.2

76.5

30

6.1

8.69

5.81

83.7 89.8 82.8 80.686

20

4.6

6.82

6.13

7.87

85.5

10

5.6

10.3 12.3 12.3 19.65

8.5 9.6 13.1 17.3 13.5 15

19.613.9 1612.3

19.65 90 9.6 100 17.3

15

19.6

16

80

90

100

Other


54 Other

2.3. Mapping LinkedIn in the Arab World Highlights - LinkedIn in the Arab Region •

Twitter Country the Arab (Mar reached 2016) Arab Users: TheTop total number of Hashtags LinkedIninusers in Twittersphere the Arab world 16.6 million as of January 2017, a growth of around 22% over a year earlier.

•

Regional Penetration: By 2017, the actual penetration rate of all LinkedIn accounts in the Arab region reached 4.1%. However, the average Sudan, 1%LinkedIn country-penetration rate across Arab Somalia, 0.37% Djibouti, 0.18% Libya, 2% Tunisia, countries was 6.75%Oman, in October 2016, up 2% from 3.8% two years earlier. 2% Mauritania, 0.08%

•

Top Countries: TheBahrain, UAE 2% has the highest penetration rates of LinkedIn in the region by a wide margin (32.5%), with one2%in every three people in the country owning a LinkedIn account. It is Palestine, followed by Qatar (22%) Jordan, 3% and Bahrain (17%).

•

Comoros, 0.03%

Morocco, 2%

Lebanon, 3% Saudi, 27% 20% of all LinkedIn users in the Arab region are based in the UAE, followed by Largest Share: Saudi (16.2%), Egypt (15.6%), Morocco (9.5%) and Algeria (8.4%). Algeria, 3%

•

Youth: Youth below Iraq,the 5% age of 35 still constitute the majority of LinkedIn users in the region. Two thirds of LinkedIn users in the region (68%) are between the ages of 18-35, a ratio that has been stable over the past three Qatar, 5% years.

•

Gender Breakdown: Female users in the Arab region constitute about 28% of all LinkedIn users, Kuwait, 7% a percentage that has also been largely static over the past two years

Yemen, 9%

Egypt, 7%

Syria, 9% UAE, 8%

2.3.1. LinkedIn Growth in the Arab Region. There were more than 16.6 million LinkedIn users in the Arab region by January 2017, growing by a rate of 22% since a year earlier. LinkedIn penetration has increased across all countries in the region. The average penetration rate among Arab countries is 6.8%. In total, 4.1% of all inhabitants of the Arab region own a LinkedIn account.

LINKEDIN GROWTH IN THE ARAB REGION (2012-2017)

Millions

LinkedIn Growth in the Arab Region (2012 - 2017) 18 16 14 12 10 8 6 4 2 0

Feb. ‘12

Feb. ‘13

Feb. ‘14

Feb. ‘15

Feb. ‘16


Feb. ‘17

55

UAE Saudi Number of Egypt Morocco Algeria UAE Tunisia Saudi Qatar Egypt Jordan Morocco Lebanon Algeria Iraq Tunisia Kuwait Qatar Oman Jordan Bahrain Lebanon Sudan Iraq Libya Kuwait Syria Oman Yemen Bahrain Palestine Sudan Somalia Libya Mauritania Syria Djibouti Yemen Comoros Palestine Somalia Mauritania Djibouti Comoros

Linkedin Users in the Arab States (Oct 2016)

1,480,000 1,300,000

725,000 565,000 556,000 515,000 506,000 725,000 451,000 565,000 340,000 556,000 236,000 515,000 168,048 506,000 158,000 451,000 155,833 340,000 152,000 236,000 143,000 168,048 44,000 158,000 40,000 155,833 25,000 152,000 10,000 143,000 44,000 40,000 25,000 10,000

3,010,000

2,510,000 2,420,000

LinkedIn Users in the Arab States

3,010,000

2,510,000 2,420,000

1,480,000 1,300,000

LINKEDIN PENETRATION IN THE ARAB STATES (OCT 2016)

LinkedIn Penetration Rates across the Arab States

LINKEDIN PENETRATION IN THE ARAB STATES (OCT 2016)

35%

32.5%

30% 35% 25%

32.5% 22.4%

30% 20%

16.9%

Q at

AE

11.3% 11.3%

AE

U

n

ar at

ai

Q

hr

Ba

w

ai t

rd an

Jo

no n

Le

Ku

ba

an

Sa

si a

Tu ni

O m

o

M

or

oc c

ge ria

Al

Regional Penetration

ud i

Average

tin e

le s

Pa

D

jib ou

ti

t yp

a by

Li

q Ira

os or

ni a

om

C

ita

ria M

au r

Sy

m en

Ye

da n

8.6%

16.9%

ar

7.3% 7.8%

8.6%

Sa ud i Le ba no n Ku w ai t Jo rd an Ba hr ai n

a

ni si

Tu

or oc c

M

in

Al

st

le

Pa

jib

yp

D

6.4% O

o

e

ge ria

t

ou ti

a

Eg

q

by

4.3% 2.5% 2.6% 2.8% 3.0% 3.2%

Li

Ira

or

os

ia an

4.3% 2.5% 2.6% 2.8% 3.0% 3.2%

7.3% 7.8%

Regional Penetration Su

m So

C

M

1.2% 1.3% 0.4% 0.4% 0.6% 0.8% 1.0%

al ia

0%

om

ria

rit

au

Sy

en m

Ye

So

5%

Su

m al

ia

10% 0%

da n

1.2% 1.3% 0.4% 0.4% 0.6% 0.8% 1.0%

Eg

15% 5%

6.4%

m an

20% 10%

22.4%

11.3% 11.3%

U

25% 15%

Average

On a regional level, penetration remains highest in the UAE, highlighting the regional status of the country as a primary job market and economic center. Almost 1 out of 3 people in the UAE owns a LINKEDIN ARAB STATES LinkedIn account. Penetration rate USER in theDISTRIBUTION country grew IN byTHE 10 percentage points from 22.7% to 32.5% in less than three years, which was the fastest growth in the region. This is followed by Jordan, which grew by 5 percentage points, almost doubling the penetration rates in the country during the same LINKEDIN USER DISTRIBUTION IN THE ARAB STATES

56

Mauritania, 0.26% Djibouti, 0.16% Comoros, 0.06%

Somalia, 0.28%

Palestine, 0.92%

Yemen, 0.98% Syria, 1.00%

Arab Social Media Report 2017 Libya, 1.02% Sudan, 1.08%

Bahrain, 1.52%

U

a Q

Jo rd Ba hr

n

Ku w

ba

Le

Sa

m O

ni

Tu

o

es ti Al ge M or oc

D jib

Eg

b Li

Ir

rit a C om or

au

Sy

M

Ye m

Su d

So m

Pa l

Average

Regional Penetration

period. The Gulf countries still dominate the top three spots in terms of penetration in the region, a trend that continues since 3 years. In terms of regional shares, the UAE also has the largest share of users in the region. Every 1 in 5 LinkedIn users in all the Arab region is based in the UAE. Saudi and Egypt followed in terms of shares of users in the region. Overall, more than half the users in the Arab region are located in the three countries of the UAE, Saudi and Egypt. LINKEDIN USER DISTRIBUTION IN THE ARAB STATES

Distribution of LinkedIn Users Across the Arab States

Mauritania, 0.26%

Somalia, 0.28%

Palestine, 0.92%

Yemen, 0.98% Syria, 1.00%


Libya, 1.02% Sudan, 1.08%

Bahrain, 1.52%

Oman, 2.19%

UAE, 19.41%

Kuwait, 2.91% Iraq, 3.26% Lebanon, 3.32% Jordan, 3.58% Qatar, 3.64%

Saudi, 16.18%

Tunisia, 4.67%

Algeria, 8.38%

Egypt, 15.60%

Morocco, 9.54%

2.3.2. LinkedIn Gender and Age Breakdowns Adoption of LinkedIn, a career-oriented social media platform, by women in the region remains low, with only 28% of all LinkedIn users being women. This gender imbalance has been largely static over the past two years. On the country level, the 7 top countries in terms of gender balance in the region have remained the Adoption of LinkedIn, a careersame over the past three years. Lebanon continues oriented social media platform, by to lead as the most gender balanced country in the women in the region remains low, region, followed by the three North African countries, with only 28% of all LinkedIn users Tunisia, Morocco and Algeria. Saudi continues to being women. be the least gender balanced in terms of LinkedIn users, with only 15% of all users in the country being women. As the platform is job and career oriented with professional objectives, this low usage by women in Saudi may be related to the nature of the job market in the country and the employment opportunities available to women. Overall, there were limited changes in the ratios of gender breakdowns in most Arab countries. However, slight increases (less than 2%) in terms of women using LinkedIn were observed in Bahrain, UAE, Oman and Saudi. Social Media and the Internet of Things - Towards Data-Driven Policymaking in the Arab World: Potential, Limits and Concerns

57

LinkedIn Gender GENDER Breakdown across the States BREAKDOWN OFArab LINKEDIN USERS IN SELECT ARAB STATES

LEBANON

TUNISIA

MOROCCO ALGERIA

JORDAN LEBANON UAE TUNISIA BAHRAIN

MOROCCO PALESTINE ALGERIA COMOROS JORDAN EGYPT UAE KUWAIT

BAHRAIN DJIBOUTI PALESTINE QATAR COMOROS MAURITANIA EGYPT OMAN KUWAIT LIBYA DJIBOUTI IRAQ

QATAR YEMEN MAURITANIA SOMALIA OMAN SAUDI LIBYA

IRAQ

YEMEN

SOMALIA SAUDI

56.0%

44.0%

62.8% 64.3% 69.4% 56.0% 70.6% 60.6%71.0% 62.8%72.7% 64.3% 75.0% 69.4% 75.7% 70.6% 76.9% 71.0% 76.9% 72.7% 77.0% 75.0% 78.9% 75.7% 79.0% 76.9% 80.0% 76.9% 82.4% 77.0% 83.8% 78.9% 84.2% 79.0% 84.7% 80.0% 82.4% Male 83.8% 84.2% 84.7%

37.2% 35.7% 30.6% 44.0% 29.4% 39.4% 29.0% 37.2%27.3% 35.7% 25.0% 30.6% 24.3% 29.4% 23.1% 29.0% 23.1% 27.3% 23.0% 25.0% 21.1% 24.3% 21.0% 23.1% 20.0% 23.1% 17.6% 23.0% 16.2% 21.1% 15.8% 21.0% 15.3% 20.0% 17.6% 16.2% 15.8% 15.3%

GENDER BREAKDOWN OF LINKEDIN USERS IN SELECT ARAB STATES 60.6% 39.4%

Female

Female

Male

In terms of age breakdown, youth below the age of 35 still constitute the majority of LinkedIn users in the region. Two thirds of LinkedIn users in the region (68%) are between the 18-35 years old. Kuwait has the most ‘mature’ (and balanced) breakdown, with almost 51% under 35 years old. Morocco has the most youthful LinkedIn users, with almost 85% of users being under 35 years old. Overall, the Gulf BREAKDOWN OF LINKEDIN USERSusers IN SELECT ARAB region has the mostAGE balanced breakdowns of LinkedIn in terms of STATES age, with the 5 countries of Kuwait, Qatar, UAE, Oman and Bahrain having the most balanced age breakdowns. On the other hand, users in the North African countries of Morocco, Algeria and Tunisia are mostly dominated by youth under 35 years old. AGE BREAKDOWN OF LINKEDIN USERS IN SELECT ARAB STATES

KUWAIT

48.7%

51.3%

Age breakdown of LinkedIn Users in Select Arab States QATAR 52.5% UAE

OMAN BAHRAIN KUWAIT LIBYA QATAR

SAUDI UAE LEBANON OMAN YEMEN BAHRAIN IRAQ LIBYA JORDAN SAUDI

EGYPT LEBANON PALESTINE YEMEN

ALGERIA IRAQ TUNISIA JORDAN

MOROCCO EGYPT PALESTINE ALGERIA TUNISIA MOROCCO

58

47.5%

44.5%

55.5%

41.9%

58.1%

41.9% 48.7% 47.5% 37.1%

58.1% 51.3% 52.5% 62.9%

37.1% 44.5% 41.9%33.6%

62.9% 55.5% 58.1%66.4%

41.9% 29.7% 37.1%27.9%

58.1% 70.3% 62.9%72.1%

37.1% 26.4% 25.5% 33.6%

62.9% 73.6% 74.5% 66.4%

23.6% 29.7% 20.5% 27.9%

76.4% 70.3% 79.5% 72.1%

26.4%17.5% 25.5%15.5%

73.6%82.5% 74.5%84.5%

23.6%

76.4%

Under 3579.5% 82.5%

35 and Above

Under 35

35 and Above

20.5%

17.5%

15.5%

84.5%

INSTAGRAM USERS IN THE ARAB STATES (OCT 2016)


Female

Male

2.4. Mapping Instagram in the Arab region Highlights – Instagram in the Arab Region •

Arab users: The total number of active Instagram users in the Arab world is around 7.1 million in AGE BREAKDOWN OF LINKEDIN USERS IN SELECT ARAB STATES early 2017.

•

Regional Penetration: By 2017, the penetration rate of active Instagram accounts in the Arab region reached 1.8% overall.

•

•

KUWAIT 48.7% Top Countries: The UAE has51.3% the highest penetration rates of Instagram in the region with 13% of QATAR 47.5% 52.5% the population active on Instagram. It is followed by Bahrain (10%), Lebanon (9.5%), Qatar (9.1%) UAE 44.5% 55.5% andOMAN Kuwait (9%). 41.9% 58.1% BAHRAIN

YEMEN

•

41.9%

58.1%

Largest share of Instagram users in the region,37.1% with 30% of all active LIBYA Share: Saudi has the highest 62.9% It is followed by the UAE (17%), Egypt (11.3%), Lebanon Instagram accounts being in Saudi. SAUDI 37.1% 62.9% (8%) and Kuwait (5%). LEBANON 33.6% 66.4% 29.7%

70.3%

IRAQ 27.9% than 55% of dominant language on Instagram with more Language Breakdown: English is the72.1% JORDAN 26.4% 73.6% Instagram activities taking place in English in the region. Meanwhile, Arabic has a 37% share. EGYPT

25.5%

74.5%

PALESTINE

23.6%

76.4%

ALGERIA

20.5%

79.5%

TUNISIA

17.5%

82.5%

2.4.MOROCCO 1. Instagram Usage in the Arab84.5% Region

15.5%

After merging with Facebook, Instagram has grown to become a leading photo-sharing social media 35 and Above Under 35 platform in the region. According to our research, it now has more than 7.1 million active users with a 1.8% penetration rate among the population of the region. Four countries exceed the regional average in terms of the number of active users in each. The largest number of active Instagram users are in Saudi with 2.1 million active users. This is followed by the UAE with 1.2 million, Egypt with 800 thousand active users and Lebanon with 570 thousand. INSTAGRAM USERS IN THE ARAB STATES (OCT 2016)

Sa ud

i

800,000

AE

pt


U

570,000

Eg y

360,000

t

Le b

an on

340,000

Ku

w ai

290,000

in e st

an

260,000

O m

n

260,000

Jo

rd a

240,000

M

or o

cc o

230,000

at ar Q

140,000

Al

ge ria

100,000

Ba h

ra in

q

66,000

Ira

ia

a

37,000

Sy r

37,000

Tu ni

si

22,000

Ye m

en

15,000

by a Li

an

Su d

Pa le

M

au r

ita

ni a

7,400

1,200,000

2,100,000

Number of Instagram Users in the Arab States (Oct 2016)

59

Average Average

Regional Regional Penetration Penetration

In terms of penetration rates, the UAE leads the region, with around 13% of its population active on Instagram. It is followed by Bahrain, Lebanon, Qatar and Kuwait. The regional average of country penetration rates stands at 4.1%. Meanwhile, almost half of all Instagram users in the region are based in Saudi and the UAE (total of 47%). Saudi Arabia has the largest share of Instagram users in the region at around 30%. Instagram Instagram Users Users Distribution Distribution in in the the Arab Arab States States (Oct (Oct 2016) 2016)

Distribution of Instagram Accounts in the Arab States (Oct 2016)

Sudan, Sudan, 0.21% 0.21% Mauritania, 0.10% 0.10% Mauritania,

Libya, Libya, 0.31% 0.31%

Yemen, Yemen, 0.52% 0.52%

Tunisia, Tunisia, 0.52% 0.52%

Syria, 0.93% 0.93% Syria,

Iraq, Iraq, 1.41% 1.41%

Bahrain, Bahrain, 1.98% 1.98% Algeria, 3.25% Algeria, 3.25% Qatar, Qatar, 3.39% 3.39%

Saudi, Saudi, 29.68% 29.68%

Morocco, Morocco, 3.68% 3.68% Jordan, Jordan, 3.68% 3.68% Oman, 4.10% 4.10% Oman, Palestine, Palestine, 4.81% 4.81%

Kuwait, Kuwait, 5.09% 5.09%

UAE, UAE, 16.96% 16.96%

Lebanon, Lebanon, 8.06% 8.06%

Egypt, 11.31% 11.31% Egypt,

While Arabic language use on social media is growing across most platforms in the region, Instagram usage in the Arab region is still dominated by English, with 55.1% of activities taking place in English and around 37% in Arabic. Instagram Instagram language language use use in in the the Arab Arab Region Region

Language Use on Instagram in the Arab region

8.2% 8.2%

English English

36.8% 36.8%

Arabic Arabic 55.1% 55.1%

Other Other

60


Kuwait, 5.09%

UAE, 16.96%

Lebanon, 8.06%

2.5. Regional Overview: Social Media in the Arab Region

Egypt, 11.31% Social media usage continues to grow The UAE tops the region in terms of both rapidly across the Arab region, impacting LinkedIn and Instagram penetration rates. societies, economies and governments. A Qatar ranks highest in terms of Facebook comparative analysis of the penetration rates penetration, followed closely by the UAE. of different social media networks in the Arab Meanwhile, Bahrain leads the region in region highlights that Facebook remains terms of Twitter penetration rate. the most popular platform, and the fastest growing one. Meanwhile, despite its influence on the media scene, Twitter’s penetration in the region remains relatively small. However, Instagram Instagram language use in the Arab Region is the medium with the smallest penetration rates in the region among the four social media platforms analyzed in this research project.

The UAE tops the region in terms of both LinkedIn and Instagram penetration rates. Qatar ranks highest in terms of Facebook penetration, followed closely by the UAE. Meanwhile, Bahrain leads the 8.2% region in terms of Twitter penetration rate. English Facebook has continued to see growth in all countries in the region, other than Palestine and Syria which saw drops over the past few years in numbers of accounts. The most significant growth on Facebook was The gender gap36.8% on social registered by Egypt, Algeria andArabic Iraq. Yemen and Oman saw media in the region a rebound in growth rates after a slump a year earlier. Qatar, 55.1% remains a chronic issue the UAE and Algeria had the highest growth as percentage of Other over the past six years. the population.

The gender gap on social media in the region remains a chronic issue over the past six years. Female participation in the region on all social media platforms, including Facebook, Twitter and LinkedIn remain well below the global average. On average, only one in three users in the region is a woman. However, social media users are maturing in the region. There is a noticeable and steady increase in the number of users above 30 years old over the past six years.

INSTAGRAM PENETRATION RATES IN THE ARAB STATES (OCT 2016)

Millions

Social Media Growth in the Arab Region between 2012-2016: Facebook, Twitter and LinkedIn as Examples

120 115 100

Linkedin

93 80 71

Twitter

60 51 40

Facebook

37 10.2

20

0

1.5

2.8

Q1 2012

8.1

4.3

5.8

3.8 Q1 2013

Q1 2014

7.5 Q1 2015

13.5 9.2 Q1 2016


61

Qatar UAE

Social Media Penetration in the Arab Countries

Jordan

Penetration Rates of Social Media Platforms in the Arab States in 2016 (as % of population) Bahrain

Kuwait Tunisia Qatar Lebanon UAE Libya Jordan Saudi Bahrain Algeria Kuwait Iraq Tunisia Morocco Lebanon Syria Libya Palestine Saudi Egypt Algeria Oman Iraq Djibouti Morocco Mauritania Syria Comoros Palestine Sudan Egypt

Ordered by Facebook penetration. Countries with highest penetration marked in red.

Yemen Oman Somalia Djibouti Mauritania 0%

10%

Comoros

20%

30%

Facebook

Sudan

40%

50%

Linkedin

60%

70%

Instagram

80%

90%

100%

Twitter

Ordered by Facebook penetration. Countries with highest penetration marked in red.

Yemen Somalia 0%

10%

20%

30%

Facebook

40%

50%

Linkedin

60%

70%

Instagram

80%

90%

100%

Twitter

Facebook in the Arab Region Between 2010-2017: Penetration Rates, Youth and Women

100% 90% 80%

Penetration

70% 60%

Females

100% 50% 90% 40%

Youth (Under 30)

80% 30%

Penetration

70% 20% 60% 10% 50% 0% 40%

Females 2010

2011

2012

2013

2014

2015

2016

2017

Youth (Under 30)

30% 20%

62 10% 0%


The population of the Arab region is estimated at 400 million in 2016 according to the UN Population Prospects data. As such, the region hosts around 5.4% of the total population of the world. In comparison, Facebook users in the Arab region are overrepresented in the global Facebook population, with 8.4% of all Facebook accounts around the world being in the Arab region. On the other hand, Twitter, LinkedIn and Instagram accounts in the region are underrepresented, with 3.5%, 3.3% and 1.2% respectively.

MILLIONS

Users of Social Media Platforms in the Region as Percentage Global Users (2017) Arab Users as Arab Percentage of WorldofUsers

2,000 1,800 1,600 1,400 1,200 1,000 800 600 400 200

8.4%

3.5%

3.3%

1.2%

Twitter

Linkedin

Instagram

0

Facebook

World

Arab

2.5.1. Social Media in the Gulf Within the Arab world, the leading region in terms of social media use and penetration is the Gulf region. The The Top Five spots in terms of six countries of the Gulf Cooperation Council (GCC); penetration rates in the region Bahrain, Kuwait, Oman, Qatar, Saudi and the UAE, have on all four platforms studied maintained their regional lead in terms of penetration of Facebook Instagram here are largely dominated by social media among their populations and their share of Gulf countries. social media population across the region. 85% The Top FiveUAE QATAR 12.9% spots in terms of penetration rates in the region on all four UAE 81.1% BAHRAIN platforms studied here are largely dominated by Gulf countries. Jordan and Lebanon are10.0% the only two countries appearing in the top five 63.1% spots occasionallyLEBANON (Jordan 3rd on Facebook and 4th JORDAN 9.5%on LinkedIn and Lebanon 3rd on Instagram). Every single country occupying Twitter’s top five spots in terms of 58.7% 9.1% QATAR BAHRAIN penetration is a Gulf country. 56.0%

KUWAIT

9.0%

KUWAIT

Linkedin

Twitter

UAE and the Internet of Things - Towards Data-Driven Policymaking Social Media in the BAHRAIN Arab World: Potential, Limits and Concerns 32.5% QATAR

22.4%

KUWAIT

12.9% 63 12.6%

ar

E

at

A

U

C

rit

au

M

Q

an om ia or os D jib ou ti S yr ia Ira M q or oc co E gy pt O m an A lg er P ia al es tin e Li by a S au d Le ba i no Tu n ni si a K uw ai B ah t ra i Jo n rd an

an

ia

ud

al

S

en

om

em

S

Y

Top 5 Countries in the Region in Terms of Penetration Rates in 2016 on Facebook, Twitter, Total Facebook Users Daily Active Users (DAU) LinkedIn and Instagram (as % of population) Facebook

Instagram

QATAR

85%

UAE

UAE

81.1%

BAHRAIN

JORDAN BAHRAIN

63.1%

12.9% 10.0%

LEBANON

Facebook Daily Active Users out of Total Users in the GCC 58.7%

(Penetration out of totalQATAR population)

56.0%

KUWAIT

9.5% 9.1% 9.0%

KUWAIT

100% 90% 70% 50%

58.1%

UAE

QATAR 30%

20% BAHRAIN 10%

JORDAN

0%

KUWAIT

40.9%

20.4%

Oman

32.5%

22.4% 37.4%

11.3%

Twitter

BAHRAIN KUWAIT

16.9% 28.0%

11.3%

72.7%

71.0%

Linkedin

60% 40%

95.3%

94.3%

80%

12.9% 64.5%

60.4%

40.1% QATAR

9.1% 8.7%

UAE

Saudi

Kuwait

Bahrain SAUDI


12.6%

8.1% Qatar

UAE


Within the Gulf, Saudi also slightly increased its majority share of Facebook users. By 2017, around 52% of all Facebook users in the Gulf were based in Saudi, up from 49% in 2014. This increase in Saudi’s share was primarily at the expense of the UAE, whose share of the total users in the Gulf has dropped from 28% in 2014 to 25% in 2017. Meanwhile, very limited change in terms of distribution of Facebook users was observed in the rest of the four remaining GCC countries. Distribution of Facebook Users in Gulf Region (2017)

DISTRIBUTION OF FACEBOOK USERS IN THE GCC Bahrain, 3%

Oman, 5% Qatar, 7%

Kuwait, 8% Saudi, 52%

UAE, 25%

64


Facebook Daily Active Users out of Total Users in the GCC

Penetration of Facebook Daily Active Users and Total Facebook Accounts in the Gulf Region (Penetration out of total population) out of the Total Population (2017) - (Penetration out of total population) 100% 90% 80% 70%

71.0%

60% 50% 40% 30%

0%

95.3%

60.4%

64.5%

UAE

Qatar

72.7%

58.1% 40.9%

20% 10%

94.3%

20.4%

Oman

37.4%

40.1%

Kuwait

Bahrain

28.0%

Saudi



The levels of activity of Facebook users in the Arab region is very similar across the different regions. Users in the Gulf region are slightly more active than the Arab average. Around 54.1% of Facebook users in the Gulf are active daily, compared to 52.8% in the Arab region. However, the penetration of daily active Facebook users in the Gulf region out of the total population of that region stands around 35.7%, which is considerably higher than the Arab regional average of 19.6%. In other words, while 2 out of each 3 persons in the Gulf own a Facebook account (66%), only 1 out of 3 persons (35%) uses that account on daily basis. In contrast, in the Arab region, around 1 out of 3 persons owns a Facebook account (37%). Meanwhile only 1 out of 5 persons (19.6%) use that account on daily basis.

DISTRIBUTION OF FACEBOOK USERS IN THE GCC Bahrain, 3%

Oman, 5%

2.5.2. Concluding Remarks: Localization, Maturity Qatar, and 7% Gender Imbalance Finally, the findings of this edition of the Arab Social Media Report series on the trends of social media growth in the region can be summarized as follows: 1) The Arabic language is increasingly dominating Kuwait, 8% the social interactions online, as these channels are becoming the de facto form of interaction for the masses For almost seven years, Saudi, 52% in the region. This is a departure from the earlier stage of women in the Arab region have growth where this was merely the interaction medium of not gained a representative the “elites” in most societies. 2) Access and utilization of voice online, nor managed these interconnected informational spheres are no more to increase their share of the limited to the youth in Arab societies. Their inhabitants digital space in the region are increasingly maturing in terms of age, as well as in UAE, 25% the ways they utilize these platforms in their daily life. 3) However, as the influence of these mediums grow, the one persistent point of concern is the regional inequality in terms of gender representation online. For almost seven years, women in the Arab region have not gained a representative voice online, nor managed to increase their share of the digital space in the region. They remain underrepresented on social media across the spectrum of channels and around the region. While this series has provided exploratory research on the reasons of this regional imbalance, digging deeper into the barriers and drivers of this critical matter should be viewed as a priority for governments and societies in the region. As decision-making and policy decisions by governments and businesses in the region are increasingly becoming digital and data-driven, the underrepresentation of women in the region in the data-driven economy will undermine the potential economic growth as well as the opportunities for better governance and development. Facebook Gender Breakdown 2017 in Arab Regions (Total Regional Populations)

Gender Breakdown in the Arab Region

(Total Regional Social Media2017 and the Internet of ThingsPopulation) - Towards Data-Driven Policymaking in the Arab World: Potential, Limits and Concerns 36.2% North Africa's Facebook Gender Breakdown 2017 (Total North African Population)

65

66


3. Annex 1: Methodology

Notes on the Population: Population at the Country-level: In calculating penetration rates, the UN population estimates were used. Using the UN’s definitions, the total population of the Arab region used in our research comprises persons of all ages who were living in the country during the reference period, regardless of residency status or citizenship. The population data for the year 2016 is based on projections by population figures used is the latest edition of the UN’s World Population Prospects: The 2015 Revision, issued by the United Nations. Palestine: Based on the population estimates by the UN of the State of Palestine, the population used here includes those living in Gaza, the West Bank and East Jerusalem. Sudan: The population of Sudan used here does not include those living in South Sudan. Syria and Yemen: Over the past few years, three countries in the Arab region witnessed significant population drops according to international organizations data, due to wars, civil strife and political divisions. These countries are: Sudan, Syria and Yemen. Syria witnessed at least a 11% drop in its population in shape of forced migration according to refugee relief organizations under the UN umbrella. Sudan also had around 24% drop in population due to the separation of South Sudan in 2011. Yemen’s population has dropped

by at least 1% as refugees fled the country due to war since 2015. The three countries have seen drops in the numbers of social media users across different platforms during the past few years at different stages. Most likely this is related to decrease in the actual population in these countries. Facebook The demographic data of Facebook users in all 22 Arab countries, was collected periodically between October 2015 and January 2017. This data included language, gender breakdowns, and age breakdowns in the following age brackets: (15-29) and (30 and over). The data collection methods followed two approaches based on grouping the 22 Arab countries into two groups. The raw usage data on all Arab countries was collected and aggregated based on Facebook’s official data in group A, which included all countries in the region excluding Syria and Sudan (both were included in group B), for which data was estimated based on historic unofficial data. This grouping was necessary as persistent technology export sanctions imposed by the US authorities have affected data availability on certain countries in the region, including Syria and Sudan. As such, for all charts in this report, the numbers of Facebook users in Syria and Sudan are estimates, while the numbers for remaining countries were


67

compiled based on actual Facebook usage data collected using different methods. The estimates of the number of Facebook users in Syria are based on a historical baseline of actual usage data first collected in November 2007. All Facebook data on Sudan in this report was estimated using the average regional daily growth rate of Arab users. This was calculated based on regional trends of countries in Group A over the following different phases of research in this series: 1) between January and April, 2011; 2) between January and April, 2012; 3) between January and February 2013; 4) August and November, 2013; 5) January and May 2014, and 6) January 2016 to January 2017). These growth rates were used to calculate the number of users in Group B countries. For Syria, the daily growth rate was adjusted occasionally based on different events, such as blockage of certain platforms, Internet connectivity issues in the country as well as political events related to the Syrian conflict starting March 2011. This baseline growth rate was based on the average daily growth rate in Yemen, a country that has similar ICT and socioeconomic indicators compared to Syria, as well as being a country with similar ongoing popular movements and military conflict influencing— and influenced by—social media activities. In addition, a one-off factor was added to the calculation of the Syria growth rate after lifting the ban on social media websites on February 7, 2011, when social media platforms, including Facebook and Twitter, were no longer blocked in the country, a higher estimate of daily growth rate was used to reflect the surge in growth number of Facebook users. This was estimated based on the surge in number of Facebook users in Egypt after a similar Internet ban was lifted on 2nd February 2011. Similarly, the average daily growth rate for Yemen was re-adjusted at several points in time and applied to generate the data related to users estimates in Syria. Twitter The study was conducted using a customized application based on Twitter’s API. The process included mass scraping of publicly available Twitter data and conducting different rounds of analysis. Estimations of populations, trends, gender breakdowns of Twitter users as well as language usage in each country were then carried out. The Twitter data was collected primarily in the month of March 2016, by collecting 1% of all tweets generated during that period, based on

68

the Twitter “firehose” stream by collecting roughly every 100th tweet. This generated around 31 million tweets as a basis for our data analysis. This provides a dataset as close as it gets to a truly random sample of all tweets generated. The Twitter data stream was then filtered to locate tweets generated in all 22 Arab countries. An unbiased distribution of tweet frequency was obtained by random sampling of the user space. Geo-locating users (estimating the location or country from the data) was done by filtering tweets with location information using a combination of methods based on 1) Yahoo and Google’s geolocation services, 2) a local database, 3) profile information provided by the users, and 4) latitude/longitude coordinates when available. The process included additional clean-up steps. Random sampling of users was applied using Twitter users’ numerical ID. As such, two sampling methods were used: 1) Trends and volume data was collected by sampling 1% of the whole of Twitter traffic, and filtering for location. 2) Users were sampled by randomly inspecting user ID numbers. This allowed for gathering information on both active and inactive users. The population estimates were produced from combining these two data sources: sample (2) provided information on user behavior, which helped assess the fraction of the Twitter population that was collected in sample (1). A correction was then applied for “un-locatable” Twitter users in each country. Population estimation was done by estimating the probability of locating the appearance of a given user in the stream, given the sampling period, tweet-frequency distribution, and the stream behavior as witnessed in the 1% sample of tweets. In some cases, an additional correction was applied for un-locatable users in some countries, to incorporate the number of virtual private network (VPN) users that may have been incorrectly geo-located by passive web analytics. However, as IP addresses were not used as a source of geo-location, the use of VPNs or other methods had limited effect on determining the location of users. Gender data was inferred based on textual analysis of publicly available personal information shared by accounts holders on their accounts. The names and username provided is compared to official census data to identify the gender of each user. Additional correction techniques were applied such as correcting for common names “shortening”, different spellings and styles, uses of titles, and the use of other common gender markers in user-reported information Arab Social Media Report 2017

such as profile descriptions. In cases of “ungendered” accounts or in cases where gender categorization is irrelevant or not applicable (e.g. for accounts held by institutions), these accounts were ignored in the gender analysis. Therefore, only actual accounts held by individuals were analyzed in assessing the gender of users. Language was detected using a languagedetection algorithm developed by Google. Selfreporting of language from Twitter settings, was also included in the analysis when available, however this was seen as unreliable due to limited number of users accurately providing this optional information. Each user was assigned one language. As such, users using multiple languages in their tweets, were assigned to the language they used most. Finally, the methodology followed in this report has been fine-tuned over six years to overcome issues continuously emerging from technological changes made to social media platforms. While all possible measures have been taken to minimize errors, these experimental methods may be subject to unintended sources of error, such as measurement, technological and coverage errors. In certain cases, limited data availability, coupled with abrupt geopolitical events in the region related to Internet governance, access and usage behaviors may play a role in infusing some measurement errors. To ensure validity, the data gathered has been benchmarked and validated against numerous alternative sources, including official and third party source when possible. LinkedIn The number of LinkedIn users in all Arab countries, excluding Syria and Sudan were collected periodically between January 2016 and January 2017. These included gender breakdowns, and age breakdowns in the following age brackets: 18-24, 25-34, 35-54, 55 and above. Usage data for Syria and Sudan were estimated using a baseline of actual breakdowns of regional usage data obtained from comparable professional social media platforms. The usage data and proportions of users in both countries were used as a baseline to estimate the proportion of users LinkedIn for each of the two countries. Instagram Data collection for Instagram were conducted in July 2016 by collecting over 750,000 randomly sampled users on Instagram during that period.

The last 5 Posts of each user were then scraped, including captions, location data when shared by users. Location was further extracted using multiple methods, including GPS location associated with shared photos which then help determine the country. Other information provided by users are assessed for location-related information including user’s other social media accounts, such as Twitter username among others. Language was determined from the posts collected using the same process as for Twitter. Similarly, gender was inferred using the approach followed for Twitter data. Regional Survey The regional online survey was administered in 22 Arab countries and ran from September to December 2016. The survey represents views of those who are actual Internet users in the region. Close to 19 thousand people took the survey. Out of those there were 5,530 successfully completed responses from the targeted region; a response rate of around 30%. More than 74% of respondents chose to complete the survey in Arabic and 26% in English. Around 56% of respondents completed the survey using a desktop or laptop computer, 41% using a smart phone, and 3% using a tablet. The sample was largely representative of the demographic breakdown of Internet users in the region. Around 81% of respondents lived in cities. Around 25% said that they are unemployed, 15% worked in the public sector and 37% in the private sector. In terms of age, 46% of respondents were under 30 years old, and around 31% of respondents were female and 69% were male. In some cases, respondents were divided into country income brackets, as per the World Bank’s classifications. These corresponded to: • High Income: Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, UAE • Upper Middle Income: Algeria, Iraq, Jordan Lebanon, Libya, and Tunisia • Lower Middle Income: Egypt, Mauritania, Morocco, Palestine, Sudan, Syria and Yemen • Low income countries: Comoro, Djibouti, and Somalia.


69

The following charts provide additional information on the demographics of the survey respondents:

SURVEY LANGUAGE

Devices Used %3 English, %26

%41

%56 Arabic, %74

Desktop/Laptop

Smartphones

Arabic

Tablets

Living in..

Respondents Country of Residence (Unwheighted) Remaining (PS, SY, LY, SO, MR, DJ): %3 Egypt %17

English

%6

Yemen %1 Sudan %2 Bahrain %3 Iraq %3

%13

Oman %4 Qatar %4 Kuwait %4

Saudi %11

Lebanon %5

%81

Tunisia %5 Morocco %10 Jordan %8 Algeria %10

Village/Rural area

UAE %9

Town/Semi-urban area

EMPLOYEMENT

City/ Urban area

AGE OF RESPONDENTS Self-employed %7

Non-government / Non-profit %6

Student %10 Private Sector %37

[50-59] Years 6%

60 and Above 2% Under20 2%

[40-49] Years 14% [20-29] Years 44%

Public Sector %15

Unemployed %25

RESPONDENTS' LEVEL OF EDUCATION

[30-39] Years 32%

RESPONDENTS' GENDER

%3 %1 %27

Female %31

%8

Other

%7

Below Secondary Secondary School Vocational College University Bachelor’s Degree University Postgraduate

Male %69

%54

70


4. Annex 2: Additional Data

The content of Annex 3 is available exclusively for members of the ASMR community online on the ASMR website: www.ArabSocialMediaReport.com You can join the Arab Social Media Report community online. Registration (at no cost) will give you access to additional datasets and unpublished charts and enable you to get updates on future editions of the report.


71

72


5. Annex 3: Definitions

The definitions used in this report, such as “social media”, “Internet of Things” (IoT), “Big data”, and others, are created and adapted from dominant definitions used by established scholarly and policy literature. There are numerous definitions of each of these terms in policy, technical and academic sources. Primarily, the following definitions informed the conceptualization of each of these terms used in this report: Social Media: 1. Social media are Internet-based applications that facilitate 1) the creation of structured networks of connected individuals, entities or virtual objects based on social attributes, and 2) facilitate multiple interactions between the nodes of these social networks, including the creation and sharing of content by users. – The Arab Social Media Report 2017, MBRSG 2. Social media are web-based services that allow individuals to construct a public or semi-public profile within a bounded system, articulate a list of other users with whom they share a connection, and view and traverse their list of connections and those made by others. – The World Economic Forum 2016 Internet of Things (IoT): 3. All devices and objects whose state can be read or altered via the Internet, with or without the active involvement of individuals. – The OECD 2015

4. The network of physical objects or ‘things’ embedded with electronics, software, sensors and network connectivity, which enables these objects to collect and exchange data – The UN / ITU 2016 Social Media and the Internet of Things - Towards Data-Driven Policymaking in the Arab World: Potential, Limits and Concerns

73

Big Data: 5. Big data are data sources with a high volume, velocity and variety of data, which require new tools and methods to capture, curate, manage, and process them in an efficient way – The UN DESA Working Group on Big data 2014

6. Big data means large data sets that have different properties from small data sets and requires special data science methods to differentiate signal from noise to extract meaning and requires special computer systems and power – The Data Science Association 2016

7. Big Data is a paradigm for enabling the collection, storage, management, analysis and visualization, potentially under real-time constraints, of extensive datasets with heterogeneous characteristics. – The International Telecommunication Union 2016

74


6. Authors, Contacts and Citation

The Arab Social Media Report series has been produced since 2011 by Governance and Innovation Program at the Mohammed Bin Rashid School of Government (formerly Dubai School of Government).

This Report was Authored by: Fadi Salem Mohammed Bin Rashid School of Government

You can use materials included in this report as long as you follow proper citation and referencing. To cite this report please use the following format:

Salem, F. (2017). The Arab Social Media Report 2017. Arab Social Media Report Series (Vol. 7). Dubai: Governance and Innovation Program, MBR School of Government.

The views expressed in this report are those of the author and do not necessarily reflect those of the trustees, officers and other staff of the Mohammed Bin Rashid School of Government (MBRSG) and its associated entities and initiatives. For questions or media enquiries please direct emails to the authors at: [email protected] or [email protected]. To follow the author on Twitter: @FadiSalem To access additional datasets and charts, not published in this report, join the Arab Social Media Report (ASMR) community and register (at no cost) online at: www.ArabSocialMediaReport.com Nonregistered members can download the full report, and follow the ASMR social networking groups through the website. Social Media and the Internet of Things - Towards Data-Driven Policymaking in the Arab World: Potential, Limits and Concerns

75

76


7. Acknowledgements Appreciation The author wishes to express personal appreciation to the following individuals for their input to the different stages of producing this report and for providing essential contributions, input and assistance into the report and its related materials: Daniel Winterstein Maryam Minhas Ghaith Yagan Tara Fischbach Marwah Alantali Yusra Alagha Sarah Alshaer Engy Osman Salha Bu-Kattara Marouen Ghezal

Special Acknowledgement The author would like to acknowledge the valuable contribution of Bayt.com and its team to the data collection efforts while conducting the survey. Specifically, we would like to thank: Hatem Hannoun Batool Al-Jboor Lama Ataya Rana Nseir Hedel Frank Social Media and the Internet of Things - Towards Data-Driven Policymaking in the Arab World: Potential, Limits and Concerns

77

78


References

CIGI. (2016). Global Commission on Internet Governance, One Internet: Centre for International Governance Innovation and The Royal Institute for International Affairs. Hale, S., Yasseri, T., Margetts, H., & John, P. (2016). Political Turbulence : How Social Media Shape Collective Action. Princeton, New Jersey: Princeton University Press. Howard, P. (2015). Pax Technica : How the Internet of Things May Set Us Free or Lock Us Up. New Haven: Yale University Press. Internet Society. (2015). THE INTERNET OF THINGS: AN OVERVIEW - Understanding the Issues and Challenges of a More Connected World. In K. Rose, S. Eldridge & L. Chapin (Eds.). ITU. (2016). Measuring the Information Society Report 2016. Geneva: International Telecommunication Union. Mayer-Schönberger, V., & Cukier, K. (2013). Big Data: A Revolution that Will Transform how We Live, Work, and Think. Boston: Eamon Dolan/Houghton Mifflin Harcourt. OECD. (2015). OECD Digital Economy Outlook 2015. Paris: OECD Publishing. UNDESA. (2016). UN E-Government Survey 2016: E-Government in Support of Sustainable Development. New York: United Nations Department of Economic and Social Affairs. UNDP. (2016). Arab Human Development Report: Youth and the Prospects for Human Development in a Changing Reality New York: United Nations Development Programme. UNHRC. (2016). The promotion, protection and enjoyment of human rights on the Internet. United Nations. WEF. (2016). The Global Information Technology Report 2016: Innovating in the Digital Economy. In S. Baller, S. Dutta & B. Lanvin (Eds.). Geneva: World Economic Forum. World Bank. (2016). World Development Report 2016: Digital Dividends. World Bank. Washington, DC. Retrieved from Social Media and the Internet of Things - Towards Data-Driven Policymaking in the Arab World: Potential, Limits and Concerns

79

80


8. Copyright Information

Readers are free to copy, distribute, transmit and adapt the work, on the following conditions: You must attribute ownership of the work to the Mohammed Bin Rashid School of Government; you must not use the work for commercial purposes; and, if you share, alter, transform or build upon the work, you must distribute the resulting work only under the same or similar conditions. These conditions may be waived if you obtain written permission from the Mohammed Bin Rashid School of Government. Where the work or any of its elements is in the public domain under applicable law, that status is in no way affected by the license.


81

82


The Governance and Innovation Program

The Governance and Innovation Program at Mohammed Bin Rashid School of Government (MBRSG) conducts research and programmatic activities focusing on policies for government innovation and development through information technologies in the Arab states. The objectives of the program are aligned with regional objectives towards nurturing a culture of innovation in society, promoting participatory, inclusive and transparent government models; and enabling more responsive and efficient governance through effective adoption of information technologies. The program works on three tracks: •

Policy and Scholarly Research: Conducting research focusing on government policies and societal transformation through technological innovation in the Arab region.

•

Policy Advisory: The ultimate objective of the Program is to inform present and future Arab policy makers in assessing the impact of the ongoing transformations in their societies and governments; and to help develop locally fitting policies for future governance initiatives.

•

Regional Development Activities: The Program brings together regional and international networks of practitioners and scholars working in related areas through programmatic and educational activities, in order to encourage proactive regional knowledge sharing and bridge the gap between policy and research.


83

84


The Mohammed Bin Rashid School of Government

The Mohammed Bin Rashid School of Government (formerly Dubai School of Government) is a research and teaching institution focusing on public policy in the Arab world. Established in 2005 under the patronage of HH Sheikh Mohammed bin Rashid Al Maktoum, Vice President and Prime Minister of the United Arab Emirates and Ruler of Dubai, in cooperation with the Harvard Kennedy School, MBRSG aims to promote good governance through enhancing the region’s capacity for effective public policy. Toward this goal, the Mohammed Bin Rashid School of Government also collaborates with regional and global institutions in delivering its research and training programs. In addition, the School organizes policy forums and international conferences to facilitate the exchange of ideas and promote critical debate on public policy in the Arab world. The School is committed to the creation of knowledge, the dissemination of best practice and the training of policy makers in the Arab world. To achieve this mission, the School is developing strong capabilities to support research and teaching programs, including: • applied research in public policy and management; • master’s degrees in public policy and public administration; • executive education for senior officials and executives; and, • knowledge forums for scholars and policy makers. Social Media and the Internet of Things - Towards Data-Driven Policymaking in the Arab World: Potential, Limits and Concerns

85

Mohammed Bin Rashid School of Government The Convention Tower, Level 13 Dubai World Trade Center (DWTC) P.O. Box 72229, Dubai Tel: +971 4 329 3290 Fax: +971 4 329 3291 www.mbrsg.ae


7th Edition