How bot-related frauds are robbing advertisers of their marketing budget.
By Sumeet Batra
The digital ecosystem is plagued by bot traffic, and it is not slowing down. For starters, a bot (short for robot) is an automated program that runs over the internet. In December, 2016, a major revelation by White Ops, a cyber-security company, was an eye-opener for many advertisers. It ousted the Russian ‘Methbot’ indicatively making as high as $5 million a day with bot traffic operation.
Multiple studies from IAS, Tune Mobile, DoubleVerify (bodies and companies that run or monitor digital advertising) and the likes indicate that the extent of bot-related fraud can be as high as 30% of the total advertising spends on digital. That’s money which has zero ROI and, in some cases, even negative ROI.
Over a period of time, bots have become increasingly complicated in their operations. While machine learning techniques are applied to contain the extent of the spread, they continue to be prevalent. Domain spoofing and click bot are the leading types of ad frauds; there can be up to 60 types of fraud implementations that are among the known ones. And then there are the unknowns. Besides generating clicks, views and impressions, that are the most common reasons why fraudsters use bots, they have also been used to infect other machines using a virus or fake IP addresses. The good news is that the tools available currently can detect many of these, but not all.
Domain spoofing operates in a programmatic environment (when ads are placed programmatically and not manually). Simply put, domain spoofing is a situation where unscrupulous publishers, ad networks or exchanges obscure the nature of their traffic to resemble legitimate websites. When a website requests a bid for impression, the request is sent to the participating ad exchanges and multiple networks working in the middle. This involvement of multiple platforms makes it hard to pinpoint the source. Finally, when an ad impression is delivered, the advertiser might think of buying into a genuine website, but it actually would get placed on
Jess Barrett, global head of programmatic for The Financial Times, said that the brand had discovered 15 different ad exchanges offering video ads on FT.com, despite the fact that the media brand does not even sell its video ads programmatically. The scale of the fraud was so large — across ad networks —that the number of fake ads being offered on the FT platform on a single day was higher than the actual monthly number.
The implication of getting fake traffic does not end with just the media cost of the click, but also extends beyond to distorted analytics data, thereby giving a skewed interpretation of the audience. This has far-reaching consequences. It becomes a self-feeding mechanism whereby the advertiser continues to re-target the bot with more premium pricing. A wrong interpretation from site analytics leads to wrong TG identification and, in return, advertisers begin to target the wrong TG.
It sounds intimidating at first to tackle such a huge technological problem that is eating into media money. However, with efficient tracking mechanisms, the problem can be mitigated substantially. For advertisers, it is imperative to employ accredited tools to ensure that media is bought only on transparent platforms with preset KPIs and appropriate clauses for payment based on the same. Ignorance is not an option anymore. It is about time we make that leap from following the media benchmarks to questioning and dictating those benchmarks.
The author is associate VP, head
(digital advisory), Spatial Access