Thanks, But No Thanks: Irrelevant Ads Online

In newspapers and magazines, on radio and television, advertising has always struggled with the question of precision. Particular outlets or shows might be associated with a general audience size and demographic, but the measures were coarse, the impacts uncertain.

It’s different today. Digital advertising promises to replace the “spray-and-pray” tactics of the past with a hyper-refined consumer insight and segmentation of consumers. “This relevance is the core promise, the bright future, of the digital advertising industry,” writes Jiwoong Shin, professor of marketing at Yale School of Management, in a working paper coauthored with Woochoel Shin from the University of Florida. “But the reality is not playing out so well.”

A 2019 survey of 287,000 people found that 90% of them consider targeted mobile phone ads annoying. This was up from 79% in 2018. How could this be? If the technology is getting better, why are the outcomes getting worse?

It turns out the problem of irrelevant – if not downright annoying – advertising is widespread. A 2019 survey of 287,000 people found that 90% of them consider targeted mobile phone ads annoying. This was up from 79% in 2018. How could this be? If the technology is getting better, why are the outcomes getting worse?

Shin and his coauthor explore this paradox using a model that replicates online ad purchase process. They find that this inefficiency arises not from failures in technology, but from fundamental incentive structures in the marketing industry that lead to irrelevant ads.

The model presents a world in which two companies hire the same marketing agency. The agency is both contractually committed to placing a certain number of ads on each company’s behalf and limited (by the companies’ marketing budgets) in the number of ads it can place. Given this, when an advertising slot opens up in an online auction – for, say, a banner at the top of a newspaper, or a spot in a social network feed – the agency has to decide whether to place an ad for either one of the companies in light of the consumers they’ll be reaching. Well-placed ads provide a triple benefit: the company reaches the right consumers; consumers learn about relevant products; and the agency makes more profit.

The researchers find that as one company becomes increasingly niche compared to a second mainstream company, the agency is likely to push a number of irrelevant ads from the niche company on a completely mismatched set of consumers.

Why? “In essence, to save a future and potentially more relevant impression for the mainstream advertiser, the ad agency ‘dumps’ the current impression to a less relevant (or even completely irrelevant) advertiser,” the researchers write. In this way, the agency “can spare the mainstream advertiser for the future impression, which may have a higher chance of matching,” while also meeting its contractual obligations. It is, in short, the future prospect of agency profit that creates irrelevant online ads.

But what happens, the researchers ask, if the agency does not need to place a specified number of advertisements? In this case, oddly enough, the agency places no ads at all, as it waits perpetually for a better match to come along. “Without being contractually bound, a strategic agency may choose to allocate zero impressions by forgoing the current mediocre opportunity for a better future opportunity,” the researchers write. Future possibility again becomes the enemy of present choice, though with different outcomes.

In the end, the promise of modern advertising technologies, which tout nearly perfect knowledge of the consumer, can only go so far. After all, marketing agencies “serve multiple advertisers, and they strategically allocate impressions to maximize profits,” write Shin and his coauthor. “Irrelevant advertising is an inescapable reality arising not from technological imperfections, but from an incentive problem inherent in the advertising industry’s institutional arrangement.”