Downstream harm
Fake conversions and lead fraud
Bots or paid workers fill in forms and trigger "sales" signals after the click, so bad traffic looks like good traffic.
How it works
Search engines judge a partner's traffic partly by what happens after the click: did the visitor sign up, request a quote, buy? A fake conversion forges that evidence. A bot or low-paid worker completes the advertiser's form with made-up or stolen details. Lead fraud is the same act seen from the advertiser's side: they pay for a "lead" who never asked to be contacted.
This is fraud aimed at the measuring instrument. It also poisons automated bidding. An advertiser using Target CPA bidding tells the system to find more people like those who converted; if the converters are fake, the system dutifully buys more of the fake source.
Who pays for it
Advertisers pay twice: for the click, and for sales staff chasing people who do not exist or never enquired. Real people whose details were used receive unwanted calls. Honest publishers are out-ranked by sources with forged quality.
Who does it, and why
Traffic sellers protecting bad traffic from detection, and lead sellers paid per form. In competitive verticals a single insurance or legal lead can be worth many dollars.
Warning signs
- Conversion rates far above the vertical norm from one source.
- Leads with disposable emails, mismatched names and phone numbers, or repeated details.
- Forms completed faster than a person can type.
- Leads that never answer, or say they never enquired.
- Conversions arriving in batches at fixed times.
Defences
- Score leads on contact rate and sales outcome, not form submissions.
- Validate phone and email at capture and watch for duplicates across sources.
- Feed back only verified conversions to bidding systems (value-based bidding).
- Advertisers can add click-level protection; ClearTrust's ClickTrust is one such product for invalid clicks.
An example
Real case: in 2009 Microsoft sued three people it accused of clicking rivals' car-insurance search ads to exhaust their budgets, then collecting contact details from the visitors their own ads captured and selling them as insurance leads. Illustrative arithmetic: an advertiser buys 200 leads at $20, or $4,000. If 60 are fabricated, the real cost per genuine lead is $4,000 ÷ 140 = $28.57, 43% higher than it appears, before counting wasted sales time.
Documented cases
- Microsoft v. Lam (competitor click fraud) (2009): Competitor click fraud on search ads to drain rivals' budgets, combined with lead selling