The main affiliate fraud types are cookie stuffing, fake or low-quality leads, incentivised traffic passed off as genuine, bot and click fraud, brand bidding and typosquatting, attribution hijacking (through browser extensions or click injection), and sales made with stolen payment details. Each leaves a recognisable pattern in your data, usually a mismatch between clicks, conversions and what happens after the conversion.
Fraud hurts both sides. Program owners pay for customers who never existed, and honest affiliates lose commission to whoever steals the last click, or see their account frozen when a traffic source they bought turns out to be dirty. This guide lists the common schemes and the detection signals for each, from the affiliate's and the program owner's point of view.
Why affiliate fraud happens
Most affiliate programs pay on an event they can only partly verify: a click, a lead, a sign-up or a first payment. Wherever the payout trigger is easier to fake than the real customer value, someone will try. Last-click attribution adds a second incentive: whoever touches the visitor last gets paid, so fraudsters try to be last rather than first.
The risk rises with:
- Payouts on early events, such as single opt-in leads or free trials (see SOI vs DOI payouts).
- High CPA rates relative to the true customer value.
- Weak approval checks on new affiliates and on sub-affiliates.
- Bought traffic from sources with poor placement controls.
Affiliate fraud types at a glance
| Fraud type | How it works | Main detection signal |
|---|---|---|
| Cookie stuffing | Affiliate cookies are dropped without a real click, via hidden iframes, pixels or forced redirects | Conversions with near-zero click-to-conversion time or no visible referring page |
| Fake leads | Forms filled with invented, stolen or recycled details | Leads that never confirm, bounce on email, or cluster by IP or device |
| Incentivised traffic | Users are paid or rewarded to sign up, against program terms | High conversion rate, almost no later spend or retention |
| Bot and click fraud | Automated clicks or form fills to inflate volume | Data-centre IPs, odd user agents, impossible timing |
| Brand bidding | Bidding on the merchant's brand terms in search ads without permission | Affiliate sales from branded search, ads seen for brand terms |
| Typosquatting | Registering misspelt brand domains that redirect through an affiliate link | Direct-type traffic with an affiliate tag and no content page |
| Attribution hijacking | Extensions, toolbars or click injection overwrite another affiliate's click at checkout | Clicks seconds before purchase, often from one affiliate across many sales |
| Transaction fraud | Purchases with stolen cards, later charged back | Chargeback rate far above the program average |
Cookie stuffing and forced clicks
Cookie stuffing means setting an affiliate cookie on a visitor's browser without them choosing to click. Methods include invisible iframes or images that load an affiliate link, pop-ups that open the merchant in the background, and scripts that redirect visitors through a tracking link on page load. The fraudster then collects commission on purchases the visitor would have made anyway.
Signals for program owners: a large share of an affiliate's conversions arriving days after the "click" with no session on the merchant site at the time of the click; clicks with an empty or unexpected referrer; and click volumes that track a site's page views rather than its content.
Signals for affiliates: your commission falls while your clicks hold steady, especially on programs with long cookie windows. Someone else's stuffed cookie may be overwriting yours. Raise it with the program with dates and examples.
Lead fraud and fake sign-ups
Lead fraud is the most common scheme on pay-per-lead and dating or cam sign-up offers. It ranges from obvious fake data typed by humans to lists of stolen identities submitted by scripts.
Detection signals
- Email addresses that never confirm, use disposable domains or follow patterns (name plus sequential numbers).
- Many leads from one IP range, device fingerprint or user agent within a short window.
- Form completion times measured in a second or two.
- Leads that convert at normal rates on the payout event but almost never at the next stage, such as first purchase.
- Country of IP not matching the country, phone prefix or language of the details entered.
Programs reduce lead fraud by paying on double opt-in or on a later event, holding commission for a validation period and scoring leads before approving them.
Incentivised traffic fraud
Incentivised traffic is not always fraud: some programs explicitly allow reward sites. It becomes fraud when an affiliate sends rewarded sign-ups to an offer whose terms prohibit them, and presents them as organic interest. The users are real, but they joined for the reward, not the product.
The giveaway is downstream behaviour. Conversion rates look excellent, sometimes far better than any content site could achieve, but retention, spend and rebills are close to zero. Program owners should compare each affiliate's lead-to-customer and customer-to-revenue rates against the program average, not just their raw conversion volume.
For affiliates buying traffic: some networks and sub-affiliates resell incentivised or bot traffic. If you resell traffic or use smartlinks, you are responsible for what you send. Check placements, use whitelists and keep evidence of your sources. Our guide to whitelisting and blacklisting placements explains how.
Bot traffic and click fraud
Bots inflate clicks to make a source look busier than it is, or fill forms to trigger lead payouts. Sophisticated bots run in real browsers and rotate residential IP addresses, so simple user-agent filters will not catch them.
Look for data-centre and hosting IP ranges, sessions with no scrolling or mouse movement, identical screen resolutions in bulk, traffic peaks at unusual hours for the GEO, and click-to-conversion times that are too fast or too uniform. Our guide on how to detect bot traffic covers log and analytics checks in detail.
Brand bidding and typosquatting
Many programs ban affiliates from bidding on the merchant's brand name, misspellings or brand plus generic terms in paid search. An affiliate who does it intercepts customers who were already looking for the brand and takes a commission the merchant would not otherwise have paid.
Detection is straightforward: run regular searches for brand terms across your key countries and devices, use an ad-monitoring service, and look for affiliates whose sales come disproportionately from branded search. Typosquatting (registering lookalike domains that redirect through an affiliate link) shows up as affiliate-tagged traffic with no real content page behind it. Both can also raise trademark issues, so program owners should state the rules clearly in their terms.
Attribution hijacking: extensions and click injection
Some browser extensions, coupon tools and toolbars insert their own affiliate click at checkout, overwriting the affiliate who actually referred the customer. On mobile, click injection apps detect when another app is being installed and fire a click just before the install completes.
The pattern to look for is a very short time between click and conversion across a large share of an affiliate's sales, plus an unusually high proportion of sales that also had a different affiliate's click earlier in the journey. If you are a publisher, persistent unexplained gaps between your clicks and approved sales are worth raising, with evidence, using our guide to tracking discrepancies.
An affiliate fraud detection routine
Fraud detection works best as a routine, not a one-off investigation.
For program owners
- Vet new affiliates: ask for traffic sources, sample URLs and, where relevant, company details.
- Set a validation period before commission is approved, long enough to see confirmations, refunds and chargebacks.
- Compare every affiliate weekly on conversion rate, click-to-conversion time, downstream value and chargeback rate. Investigate outliers in both directions.
- Monitor branded search and new lookalike domains.
- Write your rules into the terms, including the right to reverse commission for prohibited traffic.
For affiliates
- Track your own clicks with sub-IDs so you can prove what you sent.
- Audit bought traffic by placement and drop sources with suspicious patterns early.
- Keep records of traffic sources and campaigns in case a program questions your leads.
- Watch for sudden falls in approval rates or rises in reversals, which can point to a hijacker or to dirty traffic in your mix.
A simple dashboard that shows these metrics side by side makes outliers obvious. See how to build an affiliate reporting dashboard.
Frequently asked questions
Is affiliate fraud illegal?
Breaking a program's terms is a contractual matter, but deliberately deceiving a program to obtain commission can also be a criminal offence, for example under the UK's Fraud Act 2006 or equivalent laws in other countries. This is general information, not legal advice; take professional advice if you face a specific case.
What happens if a program decides my traffic was fraudulent?
Typically the program reverses the commission, withholds pending payments and may close your account, as allowed by its terms. Ask for the evidence, supply your own traffic records and, if you bought the traffic, raise a claim with the source.
Can fraud detection tools catch everything?
No. Tools catch known patterns and obvious bots, but determined fraudsters adapt. Combining automated scoring with manual review of outliers and downstream quality is what works in practice.
Next steps
Start with the cheapest check: compare each traffic source or affiliate on conversion rate and downstream value. If you need outside help with verification or traffic scoring, look through the compliance and verification services directory, and read more in our tracking and analytics hub.