Do Consumers Respond to Web Ads - SSRN

3 downloads 4774 Views 294KB Size Report
Web Advertising ” Internet Marketing Research: Theory and Practice, ed. Ook Lee,. Hershey, PA: ... Newark, NJ 07102-1897 email: [email protected].
Please cite as: Chatterjee, Patrali (2001), “Beyond CPMs and Clickthroughs: Consumer Interaction with Web Advertising ” Internet Marketing Research: Theory and Practice, ed. Ook Lee, Hershey, PA: Idea Group Publishing, 209-216.

BEYOND CPMs AND CLICKTHROUGHS: UNDERSTANDING CONSUMER INTERACTION WITH WEB ADVERTISING

Patrali Chatterjee Department of Marketing Rutgers University Newark, NJ 07102-1897 email: [email protected] http://www.rci.rutgers.edu/~patrali/

Book chapter for Internet Marketing Research: Theory and Practice.

Submitted to Dr. Ook Lee

July 2000.

The author thanks the Office of Research and Sponsored Programs, Rutgers University for their generous support of this project.

Electronic copy available at: http://ssrn.com/abstract=1861333

BEYOND CPMs AND CLICKTHROUGHS: UNDERSTANDING CONSUMER INTERACTION WITH WEB ADVERTISING

Abstract Measuring effectiveness of online ads is of critical importance to the survival of the Web as an advertiser-supported medium. Present efforts in measuring performance of Web ads are concentrated on adapting measures used in traditional media to the online medium (Novak and Hoffman 1996). However these measures do not take into account the unique interactive characteristics of the medium and hence differences in how consumers process advertising stimuli on the Web. Further ad processing and performance measurement capabilities differ across various advertising formats on the Web. This research proposes a framework to investigate how consumers interact with different advertising formats on the Web and identify metrics that reflect how effectively the ad is processed.

Electronic copy available at: http://ssrn.com/abstract=1861333

BEYOND CPMs AND CLICKTHROUGHS: UNDERSTANDING CONSUMER INTERACTION WITH WEB ADVERTISING

Introduction Advertising revenues are critical to survival of many commercial Web sites on the Web. Since advertiser-supported Web sites bear important parallels to traditional media in the physical world, established advertising practices can be borrowed from traditional media environments to assist initial commercial efforts. However, the complexity of the medium makes direct adoption of performance metrics problematic. As advertisers and marketers debate the best ways to measure response to Web ads, rate cards currently use aggregate ad reach (CPM impressions or pageviews) and ad click (clickthroughs) measures to value advertising space for most ad formats. A static banner ad impression and an interstitial impression lead to different processing outcomes – that rate cards ignore. Moreover declining clickthrough rates, and uncertainty concerning how ad exposure and clickthrough metrics relate to communication outcomes for the wide variety of ad formats used online are contributing to increasing skepticism regarding the value of advertising in this medium. In this chapter we discuss the various ad formats used by practitioners in the Web advertising industry. We use the information processing framework to investigate how consumer interaction with ads relate to communication outcomes that are of interest to marketers. Further, we identify the minimum

(and predictable) measurable outcomes that the various ad formats are able deliver. The role of tracking technology in measurement of interactive outcomes and the implications for use of ad formats will be discussed. Figure 1 presents an outline of our approach for deriving response measures for ad formats at advertiser-supported Web sites.

Figure 1. Framework for Relating Ad Processing to Performance Measures. AD FORMAT Nature of Exposure Ad Characteristics -Creative Execution -Message -Placement Interactive technology

Ad Processing

Assured minimum level of ad processing

Exposure metrics

Highest level of ad processing possible

Interactivity metrics

Advertising Formats on the Web The Web with its audio, video, and interactive capabilities offers tremendous opportunity for effective and creative development of a variety of advertising formats. In this section we discuss ten ad formats commonly used by advertiser-supported sites on the Web. Together they account for 99% of ads posted at advertiser-supported Web sites at this time. Descriptions of ad formats

can be found in Zeff and Aronson 1999. In this research we discuss the ad formats in terms of their execution characteristics. Since research on information processing implications of Web ad formats is lacking (exception: banners ads) we classify them in terms of basic characteristics – nature of exposure, coverage of visual field and scope of interaction. Despite the rapid pace with which advertising formats on the Web are evolving, response measurement of most existing and future ad formats can be investigated in this framework. Table 1 shows how ad formats differ on the classification criteria and the metrics used to measure performance by the Web advertising industry. Exposures to ads on the Web may be incidental (similar to ads in traditional media) or on demand. Ad formats that are incidentally exposed (or passive advertising exposure, Chatterjee, Hoffman and Novak 2000) are typically encountered while the consumer is browsing through content at the Web site and hence under the control of the publisher. For these ad formats exposure (while predictable by the publisher) does not necessarily imply that the consumer has noticed the ad or is aware of it unless it covers a significant portion of the computer screen (or the consumer‟s visual field). If a consumer clicks on a passive ad, the consumer is either transferred to the advertiser‟s site or is exposed to a page, what we term the “target ad,” dedicated to the advertiser‟s message (Chatterjee 1996). A clickthrough is then recorded in the server access log.

Table 1: Advertising Formats on the Web Ad Format

Nature of Exposure Incidental

Coverage of visual field Partial

Scope of interaction Limited to click

Incidental

Partial

Rich media banners

Incidental

Partial

Interstitials

Incidental

Complete

Sponsorships and Advertorials Text links or listings Ad navigation toolbars

Incidental Incidental

Partial or complete Partial

Incidental

Partial

Radio buttons, Pull-down menu Radio buttons, Pull-down menus, fill-out forms, search, complete transactions Limited to click Limited to click Limited to click Limited to click

Browser bookmarks Ad cursors

On demand

Partial

Incidental

Partial

Target ads

On demand

Complete

Buttons & Static or animated banners HTML banners

Measurement unit used CPM impressions, clickthroughs CPM impressions, clickthroughs CPM impressions, how long consumers interacted with it

CPM impressions CPM impressions CPM impressions Number of downloads, usage time Limited to CPM click impressions Limited to Number of click downloads Click to Number of advertiser site, unique visitors, search and full number of clicks range of to advertiser site, interactive time spent, functions purchase and other behavioral outcomes.

Exposure to ads that are only displayed on consumers‟ demand (or active advertising exposure Chatterjee, Hoffman and Novak 2000) by clicking on them are under consumer control and hence performance numbers cannot be guaranteed by the publisher. Since exposure is solicited by the consumer, exposure necessarily implies that the consumer is aware of or has noticed the ad and hence has attended to it. Some ad formats occupy a relatively small portion of the screen (or consumer‟s visual field) and compete with other information on the Web page for the consumer‟s attention. Consumers can easily avoid seeing or noticing these ad formats (termed “ad blindness” in Web advertising industry) compared to more intrusive formats that occupy the entire screen thus ensuring awareness (and possibly irritation). Interactivity emerges as a unique characteristic distinguishing the Web from other traditional media. Steuer (1992) defines interactivity as “the extent to which users can participate in modifying the form and content of a mediated environment in real time”. Ad formats differ in the scope of interaction they allow. This may range from a limited click response in static banner ads to a full range of interactive features in rich media ads. Interactive technologies like InterVU (www.intervu.com), Narrative Communications‟ Enliven Technology (www.enliven.com), and Macromedia‟s Flash (www.macromedia.com/software/flash) allow users to interact by providing search facilities, audio and video capabilities, ability to play games and enter

contests, send e-mail and complete purchase transactions without ever leaving the publisher‟s site.

A Model of Interaction with Web Advertising Formats Research in advertising indicates that in order to produce a response, all advertising stimuli must proceed through three stages: exposure, attention and processing (cf. Mitchell 1983) in order. The hierarchical model of ad processing (Greenwald and Levitt 1984) suggests that if processing at one level fails to evoke the next highest level, processing of the ad is terminated and the capacity is allocated to some other task (i.e., navigating editorial information at the publisher Web site). Chatterjee, Hoffman and Novak (2000) propose a model of a consumer‟s interaction with banner (passive) ad stimuli during network navigation at an adsupported publisher Web site. A session at a Web site starts with the first access of any page at the site and ends with last page accessed before the consumer exits the site. Figure 2 depicts our proposed model of interaction with Web advertisement. Interaction with an ad stimulus starts when a consumer is exposed to a Web page that has an ad on it representing the opportunity to view the ad.

Figure 2. Model of Consumer Interaction with Web Advertisement

Information Processing Stages

Enduring Effects when Ad Processing is terminated at this stage

OTS (Ad Exposure) Exposure Metric: Impressions/ Pviews

Incidental exposure

No A

Preattentive Processing: Subconscious mental representation, no recognition or awareness

Awareness (Attention) Yes

Ad Format Characteristics*

Image formation Recognition Awareness

B

No click Ad processing

Higher-level communication (e.g. comprehension) outcomes

Click

Jump to advertiser's Web site

Solicited exposure or Incidental exposure with complete coverage of visual field Publisher Site Medium or high scope of interaction

Higher level communication and behavioral outcomes

Return to navigate editorial content

*Ad format characteristics that predict that a particular information processing outcome will necessarily occur. Note: Objects with solid lines represent behavioral stages that may be tracked with clickstream data

Advertiser Site Ad

If exposure to the ad is incidental and the ad occupies a relatively small portion of the screen (as in banner ads, text links, ad cursors or ad navigation toolbars see Table 1) processing of ad information is terminated if the consumer ignores the ad and continues navigating editorial content (A in Fig.2). In this case, research in preattentive processing would indicate that information that is present but is secondary is processed at a subconscious level without awareness or recognition (Janiszewski 1993). Eye-tracking studies investigating awareness of banner ads have found that consumers do not see as many as half of the banner ads they have been exposed to (Dreze and Hussherr 1999), termed “banner blindness” by the online advertising industry (Benway 1998). On the other hand, if the ad covers the entire screen (e.g. an interstitial) the consumer has no option but to notice the ad. If exposure to the ad is on demand, for example if the consumer clicks on a banner or text link to view a target ad awareness and attention is certain to occur. Hence awareness and recognition of an ad is assured if the ad is solicited or it covers the entire visual field of the consumer. Further processing of the ad is done at the divided or focal attention level. Attentional capacity is required to determine what the ad is about and to decide if further processing or clicking on the ad will contribute to achieving the consumer‟s goals. In this stage, attending to the passive ad becomes the primary task, with editorial information relegated to the background. Ad information processing will be terminated at this level, if the consumer does not click and

moves attentional focus back to navigating editorial information (B in Fig. 2). Hence attending to the ad is a necessary, but not sufficient condition for further processing or click to occur. Clicking is possible only if possible scope of interaction for the ad format is medium or high. Higher level processing of ad information leading to communication and behavioral outcomes is possible if sufficient information content is available (as in interstitials and target ads) and/or the consumer is able to interact with the ad (as in rich media banners and text links). Table 2 discusses the information processing stages that are predictable and outcomes possible for the various ad formats on the Web. The consumer may evaluate brand information (brand processing strategy) or merely satisfy her/his curiosity or perhaps enter a sweepstakes (non-brand processing strategy). Higher level communication outcomes are not possible for static banners, ad cursors, ad navigation toolbars and browser bookmarks because of their limited informational content and scope of interaction. Behavioral outcomes are only possible for ad formats offering medium to high scope of interaction. Behavioral outcomes can be tracked with clickstream data and can be interpreted unambiguously, however value of communication outcomes is subject to speculation. This lack of unobtrusive and objective measures of communication outcomes make effectiveness of ad formats that implicitly do not have behavioral outcomes especially hard to gauge and quantify.

Table 2. Ad Information Processing Stages for Web Ad Formats Ad Format

Static or animated banners HTML banners

Rich media banners

Interstitials

Sponsorships and Advertorials Text links or listings

Ad navigation toolbars Browser bookmarks

Ad cursors Target ads

Ad Information Processing Stages OTS Awareness Communication (exposure) (attention) processing outcomes Predictable Consumer Minimal controlled, Awareness and not Recognition predictable Predictable Consumer Awareness, controlled, recognition, not comprehension predictable Predictable Consumer Awareness, controlled, recognition, not comprehension, predictable elaboration. Predictable Predictable Awareness, recognition, comprehension, elaboration. Predictable Predictable Awareness, recognition, comprehension, elaboration. Consumer Predictable Awareness, controlled, recognition, not comprehension, predictable elaboration. Predictable Predictable Awareness, and recognition Consumer controlled, not predictable Predictable

Predictable

Consumer controlled, not predictable

Predictable

Predictable

Awareness, recognition, comprehension, elaboration. Awareness and Recognition Awareness, recognition, comprehension, elaboration

Behavioral outcomes None

Search and limited information exchange Purchase and search outcomes

None

None

Purchase and other behavioral outcomes None

Purchase, search and other behavioral outcomes None Purchase, search and other behavioral outcomes

Web Advertising Measurement Models Consumer response to online advertisements can be tracked by analyzing the “log file” from which clickstream data is derived. In some ad formats (see Table 1) interactive technologies1 allow marketers to track behavioral and transactional outcomes within the ad itself (e.g., Thinking Media‟s ActiveAds (www.thethinkingmedia.com/activeads/activeads.htm). Outcome based metrics must be developed by specifying exactly what the marketer would like the ad to do. Examples of outcomes that can be tracked include purchase, software or document downloads, sending e-mail, filling out forms, or answering surveys. However communication outcomes present considerable challenge in terms of quantifying the value of an ad exposure. Novak and Hoffman (1996) offer a detailed review of advertising metrics in the Web medium. They propose exposure metrics and interactivity metrics at ad, page and vehicle level. Exposure metrics of reach and frequency are based on the one-to-many communication model underlying traditional media, and indicate that a visitor has had the opportunity to view an advertisement (page or site). Currently, exposure models namely CPM banner impressions are the dominant approach to Web media pricing. While over 86% of Web publishers (IAB 1999) use CPM impressions as one basis for pricing advertising space, the

1

Examples: (www.macromedia.com/software/flash).

value of an impression varies based on the ad format for which it is measured. For example, a static banner ad impression represents an OTS with very low probability that a consumer noticed it, however an interstitial impression guarantees awareness of the ad by the consumer. Interactivity metrics of duration time, interaction depth and width are based on the many-to-many communication model underlying the Web, and indicate the extent to which the visitor actively engages with the Web content or ad. Unlike exposure metrics, interactivity metrics seek to quantify the quality of a consumer‟s navigation experience in terms of „flow‟ (Novak, Hoffman and Yung 2000) and represent a more accountable measure of communication and learning outcomes of the ad information. Collaboration between marketers, consumer psychologists and Web technology developers is required to identify ways in which interactivity metrics can be derived from clickstream data.

Implications for Practice This research provides a descriptive overview of consumer interaction with ads in the Web medium. We propose a typology to classify the wide variety of ad formats online in terms of their nature of exposure, coverage of visual field and scope of interaction. Ad formats differ in the way consumers interact with them. Hence developing metrics based on the ease with which they can be measured rather than how consumers respond to and process ads undervalues the performance of these ads. The same metric used for two ad formats offer

differing values in terms of achieved communication outcomes for the advertiser. For a large number of advertisers who seek communication outcomes from their advertising efforts this research offers a framework to evaluate their advertising strategies.

References:

Benway, J. P. (1998), "Banner Blindness: The Irony of Attention Grabbing on the World Wide Web," Proceedings of the Human Factors and Ergonomics Society 42nd Annual Meeting, 1, 463-467. Chatterjee, P. , D. L. Hoffman and T. P. Novak (2000), “Modeling the Clickstream: Implications for Web-Based Advertising Efforts,” working paper. Dreze, Xavier and Francois Hussherr (1999), "Internet Advertising: Is Anybody Watching," working paper. Greenwald, Anthony G. and Clark Leavitt (1984), “Audience Involvement in advertising: Four Levels,” Journal of Consumer Research, 11 (June), 581-592. IAB/CASIE (1998), Guidelines for Interactive Advertising Measurement. [www.iab.net/] Internet Advertising Bureau (1999), “Net Ad Revenues Adding Up,” August 23. [www.iab.net/news/content/new%20/q2.html] Janiszewski, Chris (1993), “Preattentive Mere Exposure Effects,” Journal of Consumer Research, 20 (December), 376-392. Mitchell, Andrew A. (1983), “Cognitive Processes Initiated by Exposure to Advertising,” Information Processing Research in Advertising, R. Harris (ed.), 1342. Hillsdale, NJ: Lawrence Erlbaum Associates. Novak and Hoffman (1996), “New Metrics for New Media: Towards the Development of Web Measurement Standards,” Project 2000 white paper. Novak, T.P., D.L. Hoffman, and Y.F. Yung (2000), “Measuring the Customer Experience in Online Environments: A Structural Modeling Approach,” forthcoming, Marketing Science. Steuer, J. (1992), “Virtual Reality: Dimensions of Telepresence,” Journal of Communication, 42 (4), 73-93. Zeff, R and B. Aronson (1999), Advertising on the Internet.