Leading the fight-back against bot fraud
In the age of digital advertising, there is a concern that the online adverts which companies have paid for aren’t reaching the eyeballs they’ve been promised.
The problem is partly down to botnets, or “bots”, which get into users’ systems, interact with adverts and create fake impressions, making it seem as if the advert has been viewed by a human.
“Botnets are applications which run over the internet and which can mimic human views,” explains m/SIX’s Jess Burley. “So if a client is paying on a click-through rate or per-view basis, bots make it seem like their messaging, advertising or content has reached a larger audience than it really has.”
The extent to which bot fraud may be affecting the industry emerged in December last year, when a report by the Association of National Advertisers and WhiteOps estimated that almost 25 per cent of video views are fraudulent, and put the likely cost to the industry at $6.3bn for 2015.
There have been loud calls from within the industry to band together to tackle the issue and boost client confidence. But what weapons do advertisers have in their arsenal?
“Until quite recently, we could only assess bot-fraud after a campaign,” says Burley. “Now, technology can see bots and observe them as being a piece of technical application rather than a human being, and judge the attractiveness of an impression or keyword before we make a real-time bid on it. At m/SIX, we insist that these technologies run over the top of our clients’ campaigns.”
m/SIX is using a number of methods to detect bot-fraud, including Google Active View and Integral Ad Science, which labels infected computers as bots, and alerts partners of compromised machines.
ForensiQ is a specialist firm of ad-fraud detection experts seeking to protect the advertising and e-commerce ecosystem. Its systems score billions of impressions daily, and employ sophisticated machine-learning algorithms which allow Forensiq to “accurately identify traffic from botnets, hijacked devices, malicious script injection and other automated means,” says international managing director Erol Soyer. “This means better detection and the ability to stay steps ahead of the bad actors.”
Some in the industry feel that bot fraud has been exaggerated as a threat, however. Ben Walmsley, Sizmek’s UK managing director told Wallblog that “from the advertiser’s perspective, any impression that isn’t shown to a prospect is a wasted delivery – whether fraudulent or not.”