Search ads—those little listings that appear when you search for anything online—might be wasting big retailers $2.36 billion each year, suggested a report last week from eBay Research Labs. And search-engine marketers were not amused. Nor, doubtless, was Google, which sells most of the search ads on the web.
But some commentators, such as online marketing expert Brian Eisenberg and Greg Sterling of the blog Search Engine Land, suggested that the problem wasn’t with search ads themselves, but with the way eBay buys them. Apparently, the company’s scattershot approach to keyword-buying (egregious examples here and here) has earned it a terrible reputation in the SEM industry.
“EBay’s usage of the paid search channel is among the worst in the business,” Larry Kim, founder and CTO of Wordstream tells Quartz. ”So it’s hard to take seriously their attempts to draw broader conclusions about their failures.”
But that view is largely out of date, says Alistair Dent, a marketing head at Periscopix. “You won’t see many examples of those ads any more,” Dent tells Quartz. ”Attention was drawn to them and they fixed the problem.”
So who’s right? Maybe it’s just everyone getting a little too worked up. Dent says he thinks both the report material and the media response “are very dramatic.” It’s partly because of eBay’s reputation for having an occasionally combative relationship with Google. ”There have been some incidents between Google and eBay in the past around ads,” explained Wordstream’s Kim. And muddying that even more is that eBay explicitly views Google as a competitor and a threat to its traffic (p. 28, 43 and 44). Against that backdrop, it’s easy to read ulterior motives into the report’s release. (Ebay did not return Quartz’s request for comment.)
For that to be the case, eBay Research Labs would have to be under the thumb of eBay the company. It’s not clear whether it is. The lab’s site lists research that, like that of other technology company labs, seems more bent on advancing microeconomic scholarship than on developing actionable marketing intelligence.
Plus, the eBay Labs report isn’t as damning as that headline $2.36 billion figure suggests. The study analyzes sales—as opposed to traffic alone—that resulted from a test group of web users who were shown ads against their search results, compared with a control group who weren’t shown ads.
But this is a rather narrow way to study the impact of marketing. As the report itself says, it tracks only the “short-term, immediate revenues,” while “the long term benefits of adding customers or informing customers of new product holdings that leads to future returns are not accounted for” (pdf, p.20). In less jargony terms, the eBay study measured only how likely you are to buy something now if you see an ad for it, not how likely you are to buy it in the future.
It also didn’t take into account the opportunity costs of not buying a given ad, and letting a competitor have it. Nor did it look at the value of ubiquitous branding associated with specific queries. Of course, that’s all much harder to quantify. But just to keep things interesting, maybe eBay Research Labs should tackle those next.