The war on advertising fraud is not yet lost. According to new research from the Association of National Advertisers (ANA) and White Ops, ad fraud was down 10% from last year’s $7.2 billion to an estimated $6.5 billion in 2017. As digital ad spend continues to climb by 10% this year, the decrease is even more impressive.
The third annual Bot Baseline Report also found that the top 20 ANA members had strategies in place that could reduce losses to $700 million globally in 2017. That means: ad fraud can be dramatically reduced.
Bob Liodice, CEO, ANA, says:
“Marketers worldwide are successfully adopting strategies and tactics to fight digital ad fraud. This is a powerful indicator that the war on digital ad fraud is winnable for those who establish proper controls and protocols. And that is exceptionally good news for the advertising, marketing, and media communities worldwide.”
The research found that traffic sourcing, ie. purchasing traffic from inorganic sources, still a major risk factor for ad fraud. ANA observed 3.6 times as much fraud from sourced compared to non-sourced traffic.
Part of the problem is that bot traffic is significantly cheaper than paying for legitimate search traffic and viewability measures are scoring these sources all the same.
Fraud losses within desktop display will amount to 9%, compared to 11% last year. Desktop video spend will see just a minor reduction in fraud losses from 23% in 2016 to 22% in 2017. Given the rapid rise of desktop video, it continues to be a key target for fraudulent traffic.
Fraud levels are also considerably higher during certain times across the year. The graph below shows a clear trend towards higher fraud levels across the Holiday season in November and December.
Great news comes for mobile advertisers, where ad fraud across apps and mobile web display buys is significantly lower than feared. Study participants saw just 2% of ad fraud across these formats. This is driven by lower CPMs and decreased advertising units on mobile. In-app ad fraud is further limited to an app’s audience and counterfeit inventory on programmatic platforms hasn’t achieved a very high price.
However, none of this means that the battle is over. Bots are becoming more sophisticated and growingly resemble humans. 75% of this fraud observed in the study came from computers that contained a human and a bot on the same machine.
In addition, fraud prevention is not as easy as simply detecting a bot. Feedback loops with verification companies and advertisers have become a growing issue in having fraudulent traffic go unnoticed.
There is also a false sense of security regarding the places where fraud would happen. Private marketplaces for example saw just as much bot traffic as others, despite the overall assumption that these are clean and fraud-immune spaces.
So what can be done to combat ad fraud by bot? Among the best practices recommendations are:
- Demand Transparency for Sourced Traffic: Buyers should insist upon transparency from all vendors and publishers about traffic sources, and build language into RFPs and insertion orders requiring publishers to identify all third-party traffic sources. When such transparency does not occur, buyers should reconsider relationships.
- Refuse Payment on Non-Human Traffic in Media Contracts: Insertion orders should include specific language indicating buyers will only pay for non-bot impressions.
- Avoid Excessive Restrictions: In situations where supply does not meet demand for a target audience, fraud often follows. Marketers should avoid any actions that may restrict potential scale — e.g., too many simultaneous targeting parameters.
- Encourage MRC (Media Rating Council)-Accredited Third-Party Fraud Detection on Walled Gardens: Large digital media companies referred to as “walled gardens” should work with MRC-accredited third-party fraud detection companies to support Sophisticated Invalid Traffic (SIVT) detection. Some large digital media companies have taken steps toward seeking MRC accreditation while others have not.
Michael Tiffany, CEO at White Ops, concludes:
“Despite the proliferation of fraud detection mechanisms, bots continue to become more sophisticated and evasive. With this in mind, we’re pleased to see ad fraud rates decline in the ANA 2016–2017 Bot Baseline Report. But, as these declines are relatively modest, it’s critical that those affected by this threat remain vigilant. We look forward to collaborating as an industry to continue fighting the war on fraud and advise others to follow the industry best practices of many ANA members.”