Will worked at Branch for five years, coming in as a Senior Product Manager and eventually being promoted to Lead Product Manager. He managed Branch’s Universal Ads and contributed to the company Blog.
May 31, 2018
Digital advertising has now surpassed TV to become the primary channel for advertisers. And as ad spending has increased, so has ad fraud: estimates of the annual losses caused by fraudsters are $6.5 billion in US alone. Mobile is the next frontier, not just for user acquisition, but also for bad actors who are looking to take advantage of the fragmented ecosystem.
Today, we’re rolling out Branch’s mobile ad fraud solution, as a part of our Universal Ads Mobile Attribution product. We believe comprehensive fraud detection should be a core part of any attribution offering, so our best-in-class functionality will be included by default, for all of our partners using Universal Ads, at no additional charge.
Branch automatically blacklists activity from known bad IPs and bots, and our dashboard lets you view and export metrics that flag all common fraud types. Because of our cross-platform reach, Branch is uniquely positioned to provide fraud analytics insights that incorporate data signals no other provider can match. For example, we allow you to compare metrics like click-to-install times from your non-paid channels as a baseline for paid activity.
Using this data, it is easy to work with your ad networks to eliminate sources of low quality traffic.
Branch dashboard showing how a partner would visualize key fraud indicators for install hijacking compared by ad partner.
Our research team uses organic signals from our cross-channel products, as well as web and app signals, to screen for fraudulent and bot activity. We are constantly adding protection as new fraudulent patterns emerge.
Branch dashboard showing how a partner would visualize click to install time distribution for ad conversions compared against baseline metrics from other channels.
Branch’s fraud solution automatically blocks activity from known bad IPs in real time. However, most ad fraud today is sophisticated enough to make it through simple rule-based blacklisting. For activity that makes it through the blacklist, Branch reports on specific metrics to help partners identify patterns of suspicious behavior. For example: if organic traffic takes about 40 seconds to download an app after clicking a Branch link, click-to-install times of under 5 seconds or longer than 24 hours would be abnormal. Branch will highlight these as suspicious activity.
The Branch Fraud dashboard makes it easy to identify common fraud types, including:
|Fraud Type||Description||Detected by|
|Emulators and clicks farms generate clicks from suspicious IPs, as well as show abnormal behavior after installing. Check blocked clicks and your user value metrics in Ads Analytics to identify this fraud.||IP Blacklist (proxies, farms), Many clicks on device IDs (RealScore), SDK auth, Geo or platform mismatches (OS or Geo changes between click and install)|
|When a new app is installed, other apps on the phone can sometimes detect the install, even before the first open. Fraudulent publishers may fire off a click when they detect an organic install, just before the app is opened. This can be identified by suspiciously low click to install times||Short CTI times|
|Click Flooding||Publishers can generate millions of clicks with different device IDs in the hope that one of those users will install later. Generally the click and install actions are disconnected, so you can expect to see long click to install times.||Long CTI times|
|Ad Stacking||Publishers will sometimes put dozens of ads in one ad placement, causing high numbers of impressions and clicks, with low install rates.||Low Impr to click %, loads of clicks on one IP address|
The specific core fraud metrics Branch provides include click-to-install rates, clicks from blacklisted IPs, traffic percentages by short and long click-to-install times, click-to-install times, and multiple plot of click-to-install time distribution that allows customers to visualize baselines of normal behavior between non-paid channels, paid channels, and individual ad networks.
Using our Data Feeds product, customers can export device level records with complete data in batch or real time, for internal use or sharing with ad networks.
This functionality is available today in the new Fraud section of the Universal Ads sidebar in the Branch dashboard. As with all Branch premium products, Branch Universal Ads with our Fraud Solution included is free for mobile brands with fewer than 10,000 MAUs.