Research firms and vendors have varying methodologies and definitions for ad fraud, which creates divergent forecasts. Estimates of recent annual losses to digital ad fraud range from $6.5 billion to $19 billion. Some of the most definitive statistics come from anti-fraud vendor White Ops and advertising trade group the Association of National Advertisers (ANA).
After analyzing 27 billion ad impressions across 50 brand marketers, White Ops and the ANA projected that $5.8 billion will be lost to fraud globally this year, down from $6.5 billion in 2017.
It goes without saying that advertisers are annoyed by the persistent fraud in their industry. In a poll of 317 US marketers by research firm Advertiser Perceptions, 37% of respondents said that fraud was one of the worst aspects of programmatic ad buying.
The prevalence of ad fraud can sway advertisers away from spending more money. Ad measurement firm Integral Ad Science (IAS) found that 69.0% of US agency execs said that fraud was the biggest hindrance to ad budget growth; 52.6% of brand professionals said the same.
In May 2017, the Interactive Advertising Bureau (IAB) Tech Lab launched ads.txt, which is a text file on publishers’ sites that lists all the vendors authorized to sell their inventory. Because domain spoofing and arbitrage have plagued programmatic advertising, IAB created ads.txt so that ad buyers could have a tool to check whether a vendor’s claim to a piece of inventory was legitimate.
According to ad verification company Pixalate, the 1,000 most-trafficked websites that sell advertising programmatically have significant increased their ads.txt usage. In Q4 2018, 78.3% of these sites used it, up from 57.5% in Q4 2017.
But ads.txt is not an ad fraud panacea. Though it can help advertisers and publishers fight domain spoofing, it does not block bot traffic.
Publishers also make mistakes in their ads.txt files, which makes it difficult for demand-side platforms (DSPs) to properly filter unauthorized inventory. FirstImpression.io estimates that among the top 1,000 publishers according to Alexa ranks, 27% have errors in their ads.txt files.
Per Q4 2018 Pixalate estimates, websites that had adopted ads.txt had display ad fraud rates that were 3.5 percentage points lower than websites that hadn't implemented ads.txt. While the initiative can help lower fraud rates by keeping ads away from spoofed domains and low-quality arbitraged inventory, it’s possible the difference in these fraud rates is partly driven by self-selection bias. Websites with ads.txt adoption might have taken ad fraud seriously from the beginning and have other guards in place that drive their fraud rates down.