They’re everywhere: advertisements embedded into sites or apps such as Google and Facebook.
The ads that blend most easily into digital content are examples of “native advertising” — those that match closely the style and layout of the surrounding media.
Such ads raise the question of whether consumers are aware of them in the first place. Indeed, regulators like the Federal Trade Commission are concerned these ads might fail to disclose sponsorship sufficiently, thus deceiving consumers into making misguided purchase decisions.
Can we differentiate native ads from content — and do they influence our purchase behavior?
Stanford Graduate School of Business professors Navdeep Sahni and Harikesh Nair designed a recent experiment to answer these questions. They found that we’re very good at distinguishing native ads from digital content — but the ads still exert significant influence on our shopping behavior.
“Native advertising is a relatively new form of advertising,” Sahni says. “Advertisers and publishers have embraced this because of the rise in mobile browsing behavior, and because banner ads are hard to implement on mobile screens, and are known to be not very effective.”
Do these native ads, especially more subtle ones, trick us into clicking? Sometimes, but they’re no more effective than ads that are prominently labeled as such.
Sahni and Nair developed a field experiment in which they manipulated how native advertising for specific restaurants appeared on a restaurant search mobile app, creating two “extreme” ad presentation conditions to compare to a more typical native ad.
In the “no-disclosure” condition, nothing identified the native ad as an advertisement, making it difficult to differentiate from unsponsored content. In the “prominent-disclosure” condition, ads were presented within the same content as no-disclosure ads, but highlighted prominently as ads (with a bold highlighted label), so consumers could easily differentiate them from surrounding content. The typical presentation of native ads is somewhere between these extremes, with moderate identifying information (such as the small “Ad” label as seen in Google search results).
The study examined differences in how over 200,000 users responded to the varied presentations and found that responses to typical native ads were similar to those in the full-disclosure condition. In other words, in many cases we see native ads for the sponsored messages they are, rather than confusing them with so-called “organic” content. “We found that people who respond to the ad can spot this kind of advertising in its typical format,” Sahni says. That is, people more likely to matter to an advertiser — their target population — seem generally able to differentiate native ads from the content surrounding them.
But even when we’re aware of them, such ads exert influence on us. In the study, people who didn’t click on the ads were still much more likely to call the featured restaurant than those who saw no advertising at all. “The effect of advertising seems to happen through direct exposure and can result in conversion even if people don’t click on the ad itself,” Sahni says.
Overall, the findings suggest that because consumers who are more likely to be affected by ads can identify typical native ads easily, making the ads more prominent is unlikely to change people’s behavior. That may diminish concern about deceptiveness on the part of regulators like the FTC. At the same time, advertisers hoping for subtler ad placement should be aware that people are “on” to them, at least in the context of typical advertising practices.
Sahni points out another important implication of the work, this one for consumers: “The mechanism through which such advertising works is not clicks — the consumers we observe don’t inadvertently click on an ad and buy a product without knowing they saw the ad. Instead, they internalize the ad and may later search for the product in question or visit the advertiser’s page.” Meaning that even in a time of advanced analytics, ad exposure continues to have a deeply subtle, and thus harder-to-quantify, effect.