Protect your affiliate program from fraudulent activity with built-in fraud detection. Trackdesk automatically monitors clicks, conversions, and affiliate accounts for suspicious patterns — helping you catch fraud early and keep your data clean.
How It Works
Once enabled, Trackdesk runs a set of fraud detection rules against your traffic in real time. When a rule is triggered, the affected click, conversion, or affiliate is flagged as suspected of fraud. Fraud indicators appear throughout the platform — in your affiliate list, click reports, and conversion reports — so you can quickly spot and investigate suspicious activity.
All rules are enabled by default. You can turn individual rules on or off from Settings > Fraud Prevention to match your program's needs.
Fraud Detection Rules
Rules are organized into three categories based on what they monitor: Clicks, Conversions, and Affiliates.
Click Rules
These rules analyze incoming click traffic to detect non-human or suspicious behavior.
Rule | What it detects |
IP address tied to data centers | Clicks originating from known cloud hosting providers (e.g. AWS, Google Cloud, Azure). Real users rarely browse from data center IPs — this traffic usually indicates bots or automated scripts. |
Clicks too frequent | The same IP address generating an unusually high number of clicks within a short time period. This pattern often indicates automated click fraud or a single user artificially inflating click counts. |
Library robots | Clicks made by known web crawlers and bots (e.g. Googlebot, Bingbot). These automated programs are not real users and their clicks should not count as legitimate traffic. |
Unknown device | Clicks where the device type cannot be identified. Legitimate users typically browse from recognizable devices — unknown devices may indicate spoofed or manipulated traffic. |
Unrecognized user agent | Clicks with user agent strings that don't match any known browser or application. This often indicates custom scripts or bots that don't properly identify themselves. |
Unsupported OS version | Clicks from outdated or unsupported operating system versions. Traffic claiming to come from very old OS versions is often spoofed by bots. |
Conversion Rules
These rules check whether conversions show signs of being fabricated or manipulated.
Rule | What it detects |
Conversion occurred too soon after click | Conversions that happen suspiciously fast after the originating click. Legitimate users typically need some time to complete an action — instant conversions often indicate automated fraud. |
Click is suspicious | Conversions that originated from a click already flagged by any click-level fraud rule. If the initial click was fraudulent, the resulting conversion is likely fraudulent too. |
Conversion IP matches affiliate registered IP | Conversions where the converting user's IP address matches the affiliate's registered IP. This may indicate that the affiliate is completing conversions themselves rather than driving real customers. |
Conversion contains advS param matching an affiliate email | Conversions where the advertiser's sub-parameter (advS) matches an affiliate's registered email. This may indicate that an affiliate is generating fake conversions using their own identity. |
Conversion contains external CID matching an affiliate email | Conversions where the external click ID matches an affiliate's registered email. This suggests an affiliate may be injecting their own identifier to claim credit for conversions they didn't generate. |
Affiliate Rules
These rules monitor affiliate accounts for patterns that suggest duplicate or fraudulent registrations.
Rule | What it detects |
Affiliate registered from duplicate IP | Multiple affiliate accounts registered from the same IP address. This may indicate a single person running multiple accounts to game the system. |
Affiliate registered with duplicate billing details | Multiple affiliate accounts sharing the same billing information. This may indicate a single person operating multiple accounts to manipulate payouts. |
Customizable Rule Parameters
Most rules work automatically with no configuration needed. Two rules allow you to fine-tune their sensitivity:
Clicks too frequent
Control how many clicks from the same IP address trigger a fraud flag:
Clicks at least — the number of clicks from the same IP that triggers the rule (default: 4 clicks)
Period — the time window to count clicks within (default: 86,400 seconds / 24 hours)
For example, with default settings, the 4th click from the same IP address within a 24-hour window will be flagged as suspicious.
Conversion occurred too soon after click
Control the minimum time between a click and a conversion before it's considered suspicious:
Duration — the minimum number of seconds between click and conversion (default: 30 seconds)
Conversions that occur faster than this threshold after the originating click will be flagged.
To customize these parameters, open the rule in Settings > Fraud Prevention and select Set custom parameters. You can return to the defaults at any time by selecting Use default parameters.
Enabling and Disabling Rules
Each rule can be individually toggled on or off. To manage your rules:
Go to Settings > Fraud Prevention
Use the toggle next to each rule to enable or disable it
Disabling a rule means new traffic will no longer be checked against that rule. Previously flagged items are not affected.
Viewing Fraud Suspicions
When fraud detection flags suspicious activity, you'll see fraud indicators across the platform. You can control whether these indicators are visible by toggling the fraud suspicions visibility setting.
When enabled, flagged items display the specific rule(s) that were triggered, so you can understand the reason behind each fraud flag and decide how to act on it.