As marketing budgets shift increasingly online, fraudsters are following the money.
In the same year that digital ad spend is set to overtake traditional, pay-per-click campaigns are routinely riddled with organised click fraud scams that waste media budgets and undermine advertisers’ ability to drive sales.
Botnets are the main culprit. By infecting thousands of home personal and business computing devices with malware and turning them into (ro)bots – quietly doing the bidding of a remote command & control server – fraudsters can manufacture millions of bogus ad clicks.
But as bad as botnets are, even more sneaky are click farms, where low-paid workers are hired to manually click paid ads. While botnets are usually exposed by their behaviour and the ‘signature’ of the data traffic they create, click farms are much harder to detect.
What Are Click Farms And How Do They Work?
Click farms represent an extreme and exploitative manifestation of online fraud. Basically undercover operations, they typically employ hundreds of underpaid workers and set them up with banks of smart phones to visit and fraudulently drive up traffic to publisher websites. Once there they click selected ads or paid links to generate clickthroughs, creating the impression that the ads are ‘working’.
For new media giants like Facebook and Twitter, click farms are a major source of frustration, as they are often used to game their algorithms by artificially boosting the follower count and engagement for social media profiles.
Click farms can also inflate app downloads, app usage statistics, and traffic to publisher websites, raising the costs to advertisers.
Click farms are usually located in developing countries like China or Indonesia where cheap labour is plentiful and many people are struggling to find work.
Using humans to create bogus clicks is still more expensive than using bots, but click farms have one over-arching advantage: people can be trained to click links in a more natural way, and the clicks they create come from real mobile devices. That makes the fraud much harder to detect.
Lifting The Lid On Advertising’s Dirtiest Secret
Click farms are online advertising’s skeleton in the closet, sometimes used by otherwise legitimate internet publishers to increase their income, or employed by companies to deplete the advertising budgets and effectiveness of competitors.
In both cases the main victims are advertisers, forced to sit by helplessly, wondering if their media budgets are being eaten up by invalid clicks.
Advertisers may initially take comfort when they see clickthroughs and traffic start to rise, but soon realise that the conversion rate from traffic to sales has dropped significantly.
Likewise, when app installs are the objective, advertisers will initially see a spike in conversions and installs, but without the expected rise in-app purchases.
By skewing data and eroding revenues, click farms make it impossible for advertisers to know which campaigns are working and which aren’t. That effectively destroys the core benefit of online marketing – its measurability – and leaves advertisers with bad data that can lead to bad decision making.
How To Stop Illicit Fake-Click Operators From Destroying Your Ads
Fighting click farms isn’t easy. The accounts they use exhibit human behaviour patterns without the tell-tale signs of bot-driven traffic, and that makes them hard to detect.
—One thing advertisers can do is be on the lookout for disconnects in their ad performance metrics, for example, a high number of clicks without a subsequent rise in conversions.
—Comparing the average conversion rate between publishers is another effective tactic. If one partner’s conversion numbers are significantly out of sync with others, that needs to be investigated.
—Adjusting ad targeting can also help weed out bogus clicks. As click farms are typically based in poorer countries, it may be worthwhile to exclude some geographic locations and their respective languages.
Of course great care has to be taken not to exclude good traffic. Restricting ad targeting to combat click fraud should only be done if there’s evidence that the majority of clicks from one area or location are fraudulent.