Click fraud
Clicks made deliberately to drain a budget or distort a report, separated from ordinary bot traffic by intent rather than by technique.
Updated August 14, 2026
Click fraud is clicking a link on purpose to produce a result the click was never meant to produce: draining a budget that pays per click, inflating the numbers a paid placement is judged on, or exhausting a campaign so that it stops being shown. What separates it from ordinary bot traffic is intent, not method. The same script that harmlessly indexes a page can be pointed at a paid link, and some fraudulent clicks are made by people who were paid to make them, with no automation involved at all.
Intent cannot be read from a single request, so detection works on patterns across many: repeated clicks on one link from one address within a minute, bursts from data-centre ranges that no ordinary phone or home connection uses, headers a real browser always sends and this one did not, traffic that generates clicks but never a conversion. Keeping those signals as a score with its reasons attached is more useful than a verdict, because the score explains what was seen while the rules you write decide what happens.