¶ Duplicate User Detection
Duplicate User Detection is vital for ensuring data integrity and combating fraud in digital marketing. This filter identifies and blocks repeated clicks from the same user, enhancing tracking accuracy, reporting reliability, and optimizing ad spend.
The filter is available for both clicks and conversions
¶ Key Features of Duplicate User Detection
Allowed Duplicates
Sets the maximum allowable clicks from a single user within a defined timeframe.
Lookback Time
Specifies the monitoring window for duplicate clicks, typically from minutes to hours.
Detection Levels
- Offer Level: Limits duplicates for specific offers.
- Publisher Level: Controls duplicates across publishers.
¶ Detection Mechanisms and Data Points
The system identifies duplicates using robust data points, including:
- IP Addresses: Tracks IPv4 and IPv6.
- UUIDs: Monitors unique user identifiers.
- Device IDs: Ensures clicks originate from distinct devices.
- User-Agent: Detects browser and OS details.
- Fingerprinting: Combines multiple data points into a unique identifier.
¶ Configuration Rules
Customizable rules allow granular control over duplicate detection:
- Rule Creation: Tailored rules based on data points.
- Application Levels: Apply rules to offers or publishers.
¶ Setting Up Duplicate User Detection
- Navigate to filter settings.
- Define allowed duplicates and lookback period.
- Select data points for detection, such as IP or UUID.
¶ Duplicate IP - Case Study
A Duplicate IP occurs when two or more clicks/conversions on the same Offer or traffic Partner within a predefined time interval have the same user IP Address. Our settings are tolerant by default (24h / same game / same partner).
Duplicate IPs are considered fraudulent when an IP Address repeats itself excessively on the same product, as this indicates that the same user is performing multiple conversions. This can often relate to Bots or users on Rewarded traffic trying to get their rewards multiple times.
Argument
Duplicate IPs are duplicate installs / conversions as they relate to the same device/user. We only consider IP duplication on the same Offer / Publisher within 24 hours. The likelihood of a false positive is extremely low, as all of these factors would have to come together to make this a false positive.
In addition, for example,the U.S. has about 1.6 billion registered IP addresses and a population of about 300 million people. IP collisions are, therefore, extremely rare, as we also consider the other factors listed above.
¶ Conclusion
Duplicate User Detection is indispensable for digital marketing campaigns, particularly CPC models. By employing advanced detection mechanisms and configurable rules, businesses can ensure accurate data, combat fraud, and maximize ROI.