Digital ad fraud is costing advertisers
$8.2 billion each year, according to the Interactive Advertising Bureau. So what is the industry doing to address this massive crime?
To date, the collective approach has been largely reactive: A reputable third-party measurement company audits a publisher’s traffic, and if fraud is detected, reports on the percentage that is non-human traffic (NHT). If that percentage is unacceptably high, advertisers are encouraged to block those sources of media from future campaigns, or even refrain from paying for NHT traffic.
Inherent in this approach is the acceptance of fraud as inevitable. It views NHT as a scourge that will inevitably cost the industry some sum of money, and our goal should be to get those costs as low as possible. But is NHT a fact of life we must learn to accept?
Is non-human traffic inevitable?
To find out, The 614 Group, my employer, designed a quantitative and qualitative survey to understand where publishers currently stand in their battles against NHT. Our study, conducted in the fall of 2016, consisted of a mass-market survey of 40 of the largest 250 media companies in the US, including AccuWeather, A&E Networks, Hulu, Thomson Reuters and Univision.
With those results in hand, we discussed the findings in one-on-one interviews and roundtable discussions with industry leaders. As a critical next step, we sought to understand why many publishers are currently not blocking NHT.
One reason we landed on is that companies don’t exactly understand how NHT is affecting their sites and what the campaign-by-campaign impact is. Having this information in hand is critical.
For example, if you knew that NHT was costing your client $1 million in damages, would you spend $100,000 to fix it? The obvious answer is yes, but understanding the real impact of the problem is the first step to solving it.
For the final phase of this study, we designed a test with our technology partners at Distil Networks to learn more about the impact of ad fraud. So what did we learn from our research, and what were the data points?
Is purchasing traffic at fault?
The data turns some long-held notions on their heads. Case in point: It’s generally assumed that publishers more or less usher NHT onto their sites when they purchase traffic. But in fact, 78 percent of publishers report experiencing NHT on their sites, yet only 38.4 percent purchase traffic. Clearly, NHT is getting onto their sites through other means, through no fault of the publisher. For this reason, a lot of marketers are losing trust in the digital ad marketplace.
Adam Moser, head of operations for Hulu, explained in a one-on-one interview during our research how NHT gets onto premium publisher sites:
These bots don’t just sit on low rent, unlit sites; they’re programmed to visit a lot of premium sites so they’ll look like targetable humans. And that’s the fraudster’s goal: To create bots that are ultimately targeted with the most premium quality media, especially video.
The Association of National Advertisers (ANA) has advised its members to no longer pay for fraud, and they appear to be heeding that advice: 74 percent of publishers say that traffic quality issues are part of pre-sales discussions, and 68 percent said they’ve received RFPs that included acceptable NHT thresholds.
At this year’s IAB Annual Leadership Meeting, Procter & Gamble’s Mark Pritchard very publicly criticized what he called the “crappy” the media supply chain, giving agencies a year to “get to a transparent, clean and productive media supply chain” or risk losing its business. I have no doubt that he is the first of many advertisers who will draw a line in the sand, “voting with dollars” to stop working with companies that don’t ensure their ads are seen by humans.
Not even on the same page
Let’s go back to that $8.2 billion stat. That’s a mind-boggling sum that is utterly irresistible incentive for determined and skilled fraudsters. Advertisers feel the weight of that crime, but do publishers?
In our one-on-one conversations, we learned that publishers are failing to draw a connection between the $8.2 billion lost to fraud and campaign damage. To a publisher, the cost of fraud is the revenue that’s lost when post-campaign traffic is assessed, and they must compensate for the small percentage of traffic that’s deemed NHT. No big deal.
“Actually, publishers lose a lot more money than they realize to fraud, because they don’t factor in the revenue they lose when fraudsters send bots to visit their sites to create a fraudulent cookied bot audience, and then sell their cookied NHT as readers on the open exchanges,” Elias Terman, vice president of marketing at Distil Networks, explained to me. “Advertisers allot budget for this traffic, but it’s the fraudsters, and not the publishers, who collect it.”
Advertisers, on the other hand, are left fuming. To them, NHT is an $8.2 billion crime perpetrated against them each year, and they are increasingly wary of the digital channel as a result. To them, publishers need to take the total cost of fraud more seriously.
What constitutes a serious approach to fraud?
Let’s start with resources. Only 5 percent of publishers have a dedicated fraud-control officer, a number that sends a bad signal to advertisers. A fraud officer can advocate for the kinds of investments publishers need to make to seriously combat fraud.
More importantly, many believe the ideal solution is to adopt a proactive, rather than a reactive, approach. The idea is simple: Filter, identify and block NHT before it’s able to get on-site.
Despite the industry’s current ability to proactively block fraud, less than one-third of publishers do so. Of course, getting the capital to invest in this kind of infrastructure requires justification. But what’s the ROI publishers can expect?
That’s the next step in our journey.
Testing ROI of anti-fraud on individual campaigns
In our survey, 70 percent of publishers said they believe it’s possible to calculate the ROI of successful anti-NHT efforts per campaign and per client (even though none of them are yet doing so). That’s when we decided to conduct the testing on real campaigns, using real KPIs.
We hope to create some knowledge and understanding of what it might take to fight it and what the impact is. We’re actively seeking premium publishers to run tests in the next quarter, and we plan to release the results.
We’re excited by this, because we think there is a great opportunity for a proactive approach, and like 65 percent of the publishers surveyed, we believe that certified NHT-free traffic is worthy of a premium. We will share our next set of results right here in Marketing Land.
Comments