New data has been released on the estimated amount of global advertising dollars that will be wasted on fraudulent traffic in 2017. According to a WPP study, the figure is $16.4 billion. That’s more than twice the (mere) $7.2 billion estimated was wasted the previous year.
There is an telling split in top reasons for concern about digital media buying between brand marketers and media agencies. According to an April 2016 MyersBizNet survey, the 70% of respondents from media agencies are concerned about viewability. This is on par with the 71% of brand marketers who also could count this as a concern. But on the brand marketer side of the table, 78% of respondents are concerned about click fraud and bot traffic compared to 63% and 59% for media agency respondents.
Programmatically traded inventory, particularly inventory sold in open marketplaces, is more likely to be fraudulent. Integral Ad Science recently found that of all US digital inventory, 8.7% of programmatic impressions were classified as inventory with no possibility of being viewed by a human. Compare this with publisher-direct sold ads which came in at 2.4%.
The higher the CPM, the greater the bot rate.
An ANA/White Ops survey saw that inventory priced at $10 or above had a 39% higher bot rate than cheaper inventory. This survey also showed that programmatic video was the most vulnerable format for bot fraud. The bot percentage range for programmatic video based on impressions monitored by White Ops ranged from 1% to an amazing 70%.
If we don’t address this problem of ad fraud, over the next 10 years, the 2017 estimate could be dwarfed by the potential lost revenue. The World Federation of Advertisers estimates that by 2027, the global cost of ad fraud could rise to $50 billion.
Publisher-direct is a safer place to plan a digital advertising campaign but the advertising industry collectively needs to be better at combatting ad fraud. The work that has been done to set guidelines and standards is clearly not effective at stemming the growth of this problem.