Learn about Singular’s new mechanism for protecting customers from a common type of attribution fraud.
For more anti-fraud tools offered by Singular, see Fraud Protection Methods. Note that Click Flooding Protection is different from the normal "protection methods": it cannot be configured in the Fraud Global Settings dashboard, and the clicks that it blocks are not included in fraud reports.
Click flooding is a form of fraud in which fraudsters send extremely large volumes of clicks to an MMP in an attempt to hijack the attribution credit for legitimate, often organic, installs. The goal is for the MMP to misattribute installs to a publisher that is controlled by the fraudsters and from which they can get a payout.
Since fraudsters do not know when installs of the app might occur, they send millions of clicks every day for the app that is under attack.
While click flooding is a basic form of fraud and can often be spotted simply by looking at click numbers or conversion rates, it is also one of the most effective and common fraud methods.
Singular's Fraud Prevention Engine detects click flooding by looking at click volumes, click volume changes (and pace of changes), conversion, and clickthrough rates broken down by ad network, publisher, and sub-publisher, as well as other parameters indicative of fraud.
When the Engine detects a click flooding attack, it blocks the campaign (the tracking link) from firing any additional clicks for the next 24 hours, after which the status of the campaign is reassessed.
In extreme cases, when there are multiple campaigns being attacked and the advertiser and publisher are both at high risk of losing money to fraudster payouts, the Fraud Prevention Engine may block an entire publisher.
Singular's high precision and low false-positive rates are unmatched thanks to the flexibility of the fraud rule engine. Singular is the only MMP that can detect and reject clicks at the campaign level, while other MMPs are forced to block publishers entirely due to infrastructure risks.
Singular uses multiple, separate tools to protect you from fake clicks.
Click Flooding Protection is an additional level of defense that operates differently from the normal protection methods. Click Flooding Protection is activated automatically, it cannot be configured in the Fraud Global Settings dashboard, and the clicks that it blocks are not included in fraud reports.
The blocking of a campaign or publisher does not affect real users. A real person clicking the ad would still get to the store normally and be able to install the app. The "blocking" only affects the collection of data for attribution. Clicks are rejected, not blocked. This means that real users will still get to the store and install the app. Any other clicks for that install will still get considered according to normal attribution logic. If no clicks are found, the install will be marked as Organic.
Singular cannot disclose the exact thresholds or parameters it uses to determine that a campaign is being attacked via click flooding.
However, click flooding volumes are extreme - as large as millions and tens of millions of clicks per day and billions per week. In conjunction with extremely low CVR rates, which means that users don't actually install the app following these clicks, the detection of click flooding is often clear-cut.
Campaigns that were not flagged as containing high volumes of click flooding are not affected.
A campaign (which corresponds to a link) gets blocked for 24 hours. After that, the campaign is reassessed, and if the click flooding attack has stopped, normal attribution is resumed.
Singular only rejects campaigns that exhibit an extreme number of clicks or that are trending at an extremely fast pace. We use machine learning to learn the expected behaviors of publishers and campaigns. If your campaign genuinely reaches high volumes of legitimate clicks at a fast pace, you will not be affected.