The cost of bot fraud in 2016 is likely to rise to a staggering $7.2 billion, according to a new report by the Association of National Advertisers (ANA).

2015 Bot Baseline study places the cost of bot fraud at over $7 billion

The study, conducted in conjunction with WhiteOps, shows that despite efforts to reduce the impact of bot fraud, criminal gangs are still managing to game the online advertising industry. Advertisers are being tricked into thinking that real visitors are viewing their adverts and are paying for those visits, when in actual fact a substantial percentage come from bots.

For some companies the losses were shocking. The highest losses were reported to have cost one company $42 million over the course of the year. However, even smaller companies did not escape unscathed. The cost of bot fraud for the least affected advertiser was $250,000.

ANA studied 1,300 advertising campaigns conducted by 49 major companies over a period of two months from August 1, 2015., to September 30, 2015. The results of the study were then extrapolated to provide the cost of bot fraud for 2016.

The study examined more than 10 billion ad impressions to determine the percentage that were real visitors. To distinguish bot visits from the human visits, ANA/WhiteOps added detection tags to the advertising campaigns under study.

The same study was conducted back in 2014 and this year’s results show that virtually nothing has changed, with just a fall in bot fraud of just 0.2% registered. The level of bot fraud has remained constant, although the cost to companies has increased.

In 2014, online advertisers were estimated to have lost around $5 billion to bot fraud, with the rise in cost of bot fraud due to an expected increase in advertising investment over the course of the next 12 months.

Last year, brands suffered an average of $10 million in losses to bot fraud. That’s an average of $10 billion paid to advertise to bots. For 25% of companies, 9% of impressions go to non-human traffic.

Methods of bot detection have improved, but they are clearly not having much of an effect on the cost of bot fraud for advertisers. As detection methods improve, bot operators have improved their ability to obfuscate their bot visits.

Unfortunately, it is difficult to distinguish bot traffic from real traffic as more residential IP addresses are being used, and the bots are becoming better at mimicking real browsing habits.