As Microsoft founder Bill Gates said this week, “People don’t mind having a little bit of demographic information about themselves used to target ads. That’s value added to the user.” When data is clearly used to improve the relevance of the ads consumers see, it can improve their overall experience.
Unfortunately, the amount of data collected and the myriad ways in which it is used are opaque to most consumers. And, when they get a glimpse behind the curtain, (as with Facebook’s Cambridge Analytica dealings) consumers grow mistrustful of data collection—and even online advertising—as a whole.
However, it is possible that if consumers are more informed on the tracking process and what’s going on behind the scenes, advertising would be more effective and impactful. This is what Tami Kim, Kate Barasz, and Leslie John examine in their study, “Why Am I Seeing This Ad? The Effect of Ad Transparency on Ad Effectiveness” published recently in the Journal of Consumer Research. Their research analyzes the impact of ad transparency on ad effectiveness.
Today, some websites and advertisers are informing users (albeit in a limited fashion) about what they are tracking and their data practices. Some sites may display an adChoices icon that indicates an ad is targeted based on user characteristics. Consumers can find out why the ad is being displayed to them by clicking on the icon. Other websites are alerting visitors, upon first visits, of their tracking user software and practices (e.g., cookies).
The Ad Transparency Effect
Kim, Barasz, and John conducted five studies using the Facebook platform. They analyzed the findings to identify the core dimensions of ad transparency needed to positively impact ad effectiveness.
Study 1 allows consumers to select acceptable tracking methods from a pre-defined list. The analysis identifies two important ad tracking practices consumers find acceptable: 1) the information is obtained from tracking within the site and not outside of it; and 2) the information (attributes) is provide by the consumer and not inferred by the site.
Study 2 tests the effectiveness of revealing data tracking practices within a site versus across-websites. Participants are shown an ad. Then, they are shown either no information or messaging as to why they are seeing this ad (message a: the ad has been generated based on user information obtained within the platform or message b: the ad has been generated either based on information obtained cross-platform). The results from Study 2 confirms that transparency messaging increases ad effectiveness especially when it reveals the information is obtained within the platform and not from cross-website tracking.
Study 3 examines the impact of revealing a targeted ad is based on stated attributes versus inferred attributes. Participants are shown an ad and then given one of two messages (message a: the ad is generated based on information stated by the consumer or message b: the ad is generated based on information inferred by the site). Results for Study 3 indicate that privacy concerns are higher in the inferred attributes and detracts from advertising effectiveness.
Study 4 looks at the role trust plays in a site’s data transparency and its impact on ad effectiveness. The results of this study indicate that users who trust a site are more likely to engage with an ad that offers ad transparent messaging than those sites they distrust.
Study 5A and 5B explore the impact of ad effectiveness when there’s both trust of a site and ad transparent messaging. Study 5A divides consumers into two groups, half are assigned to a loyalty program (higher trust of site) and the other half to a non-loyalty program (no trust of site). Personal information was obtained transparently from the loyalty program consumer group.
This study found consumers are more willing to click on recommended items, spend more time, and purchase more when they both trust the platform. They will also provide their own data (attribute information). Study 5B includes two additional consumer groups and offers messaging regarding shared or inferred user attributes. Message group one: Recommended based on what you’ve shared with us (implies data user provides). And message group two: Recommended (implying the data is inferred about the user). Study 5B reveals that consumers are more likely to have a higher propensity to click on recommended items if they state their attributes than those targeted based on inferred attributes.
Kim, Barasz, and John’s research is an important step to understand how consumers’ readiness to engage with digital ads is affected by their awareness of the data practices used to deliver such ads. Ad transparency messaging is a critical step to increase advertising effectiveness on your site.