AOL's New Video Audience May Consist Largely of Robots

Photograph by Yagi Studio

AOL’s purchase of video-advertising company for $405 million last week may come with a serious problem—that no one sees many ads purchased through the company’s exchange. According to a report on Thursday in AdWeek, multiple sources with first-hand knowledge of the exchange say a significant portion of the network’s activity is phony.

Online advertisers typically pay based on how many people see their ads, but juices its numbers by sending ads to invisible windows on high-traffic sites, displaying them underneath the visible ads or in impossibly small windows, according to the AdWeek report. (This kind of practice is called impression fraud; one impression equals one viewer.)

Such deception has been a long-running problem with display ads. It is working its way into video ads whose rates are higher, security is less established, and advertisers aren’t as worried about click-through. In a report presented this week at the Usenix conference in Washington, two researchers from the firm Broadcast Interactive Media estimated that advertisers are losing $180 million annually by paying for traffic that never sees human eyes.

Impression fraud pervades’s network, according to AdWeek:

One top exchange buyer said in a recent report that 32 percent of’s sites were flagged as suspected of being bogus (nearly 500 sites), compared to 3 percent for other video exchanges. Separately, a Web fraud researcher found at least 30 percent of’s traffic to be bot-driven. The percentage of suspect inventory grows when nonviewable ads are included.

Another buyer said: “ has historically been the worst inventory we see in terms of suspicious activity. At one point it was close to 80 percent of their traffic. This month it was 40 percent suspicious.” That buyer added that a large amount of’s suspicious traffic probably is driven by bots, since they consistently see an extremely high overlap between sites.’s chief executive denied that the network’s traffic is overrun with bots, and says it bans five to 10 sites a week.

While advertising companies generally don’t endorse fraud, the incentives to engage in it are far higher than those to oppose or prevent it, writes John Battelle:

The bad actors are currently far ahead of the good guys, and worse, many in our industry are turning a blind eye, hoping the problem goes away in time, without too much publicity. Why? Well, nearly everyone gets paid from fraud—the publishers, the exchanges, the data providers, and the agencies. Even the marketers, who are footing the bill, feel like they are getting value—because the success metrics they’ve set up are being met.

There could be some perverse game theory at work here. Sure, businesses may be profiting from the fraud of others, but the practice weighs on the advertising industry. It’s certainly one factor pushing down ad rates because it artificially inflates viewership to near-infinite proportions. Weakened rates could be one reason why investors on Wall Street haven’t warmed to video advertising companies such as Tremor Video and YuMe, both of which went public earlier this year.

Before it's here, it's on the Bloomberg Terminal.