Identify and report on suspicious activity
Block based on known fraud patterns
Analyze signals to find and block new fraud
Scalable, data-driven action
Implementing hard-coded blocking rules (that quickly become stale) allows fraudsters to respond quickly to ever-evolving fraudulent patterns.
Branch’s unique, cross-platform scale provides the data depth and breadth required for a resilient, scalable detection model based on machine learning. We automate the detection of statistical variations across a vast pool of data signals to spot and block fraudulent patterns as they emerge.
Quality insights
Most attribution providers mark fraudulent attribution data as “organic.” This distorts healthy analytics and prevents customers from comparing the quality of traffic across their ad sources.
Customers can clearly view suspicious patterns in data, separate from healthy analytics, right in the Branch dashboard. While attribution data is preserved, Branch blocks attribution payouts and takes action to remove bad actors in real time, so customers do not have to worry about recouping losses.
Stay ahead of common fraud techniques
Fake Devices/Click Farms
Fake Devices/Click Farms
Fraud Method: use emulators and device farms to generate fraudulent clicks.
Detection Strategy: blacklisted IPs, because many conversions will come from the same IP address; high click-to-install rates, because this fraud method is most cost-effective when clicks lead to installs.
Install Hijacking/Click Injection
Install Hijacking/Click Injection
Fraud Method: use a compromised app already on the device to detect installs, and fire off a click just before the new app is opened.
Detection Strategy: short click-to-install times, because users typically open apps right after downloading them.
Click Flooding
Click Flooding
Fraud Method: generate millions of clicks with random device IDs (IDFA or GAID), and hope that some of those users will install later.
Detection Strategy:  long click-to-install times, because there is no legitimate connection between click and install.
Ad Stacking
Ad Stacking
Fraud Method:  layer dozens of ads on top of each other, leading to fraudulent clicks on ads that aren't even visible.
Detection Strategy:  low click-to-install rates, because this fraud method generates excessive numbers of clicks.
Fake Devices/Click Farms
Fake Devices/Click Farms
Fraud Method: use emulators and device farms to generate fraudulent clicks.
Detection Strategy: blacklisted IPs, because many conversions will come from the same IP address; high click-to-install rates, because this fraud method is most cost-effective when clicks lead to installs.
Install Hijacking/Click Injection
Install Hijacking/Click Injection
Fraud Method: use a compromised app already on the device to detect installs, and fire off a click just before the new app is opened.
Detection Strategy: short click-to-install times, because users typically open apps right after downloading them.
Click Flooding
Click Flooding
Fraud Method: generate millions of clicks with random device IDs (IDFA or GAID), and hope that some of those users will install later.
Detection Strategy:  long click-to-install times, because there is no legitimate connection between click and install.
Ad Stacking
Ad Stacking
Fraud Method:  layer dozens of ads on top of each other, leading to fraudulent clicks on ads that aren't even visible.
Detection Strategy:  low click-to-install rates, because this fraud method generates excessive numbers of clicks.
Create custom rules that block at the event level based on any attribute.
Enable AI-generated rules to stay protected as fraud patterns evolve.
We use constantly advancing proprietary algorithms to dynamically block attributions on browsers and devices showing suspicious behavior across our platform. Our advanced machine learning gives you the highest level of attribution while focusing on user privacy.
You can learn more about Branch recommended fraud rules by speaking with your customer support team.