Singular's Fraud Prevention Engine identifies various types of ad fraud - from click hijacking and fake installs to custom, user-defined fraud symptoms - to ensure that Singular attribution decisions are not influenced by fraud attempts.

The Fraud Prevention Engine ensures that your data reflects your real marketing performance and that you are not billed excessively for CPI/CPA campaigns.

Guide for	Management, UA teams
Prerequisites	Singular attribution is set up for your account

Step by Step

Understand the Different Protection Methods

Singular offers a rich menu of pre-existing fraud protection methods that detect different types of fraud. For example, the "Blacklisted IPs" method detects clicks and installs coming from IPs that are associated with fraud (such as VPNs).

If you enable a fraud method, it will be used to evaluate all the clicks and impressions that are candidates for attribution before a final attribution decision is made. However, not all methods are useful for all verticals, geographic locations, etc. This is why we give you the freedom to enable or disable each method separately.

To learn more about each method, see List of Fraud Protection Methods.

Note: Singular also protects you against click-flooding attacks. Click flooding protection is enabled automatically and you don't need to set it up.

Learn more in the Click Flooding Protection FAQ.

Set Up Your Protection Methods

To manage the methods used in your account, go to Fraud Prevention > Settings, go over the methods, and for each one choose the relevant action:

Screen_Shot_2022-03-01_at_19.33.04.png

Action	Description
Do Nothing	This method is disabled and will not affect the attribution decision or your reporting.
Mark as Suspicious	If this method applies to an ad click or ad view, the click/view will be flagged as suspicious but the attribution decision will not be affected. You can then run a Suspicious Report to look at which of your attributions were marked as suspicious in the.
Reject	If this method applies to an ad click or ad view, the click/view will be rejected from consideration by Singular's attribution tracker. This means that the install will be attributed to the next click/view in line, or if there aren't any relevant ones, the install will be marked as organic.

Best Practices: Use “Mark as Suspicious” Before Enabling “Reject”

Singular's fraud prevention tools - including both the default method and any custom rules - may run into edge cases and produce unwanted behavior in the form of false positives or false negatives. When you activate new methods or rules without testing them, they may affect your ongoing marketing campaigns or impact your volume unexpectedly.

This is why we recommend testing new methods and rules by setting them to "Mark as Suspicious" rather than "Reject". Monitor the performance of the methods/rules for a while, using the suspicious report, and make sure they work as expected.

Once you're satisfied with the methods/rules, promote them to "Reject" so that they prevent Singular from sending billing postbacks and help keep down your CPI/CPA charges.

Set Up Custom Rules (Optional)

In addition to the fraud protection methods developed by Singular, you can also create your own custom rules to prevent types of fraud that are specific to your apps.

For example, you may want to whitelist partners that are guaranteed to not fraud you, e.g., internal cross-promo. Or you may reject installs that don’t match your campaign’s targeting options.

To create a user-defined rule:

Go to Fraud Prevention > Rules.
Select Create New Rule.
Rule Conditions: Each condition is composed of three parts:
- Field: the field that is being evaluated based on the touchpoint and install. For example, if you choose the field "Source," the touchpoint will be evaluated based on its source (the ad network).
  
  For details, see Fields that Can Be Used in Custom Fraud Rules.
- Operator: the operator defines the relation between the field and the values.
- Values: the values used on the right side of the operator for the condition.

Actions: For each rule, you can choose the type of action to be taken when the rule is found to apply to a certain click/view:

Whitelist	Any touchpoint that matches a whitelisting rule will be considered valid. Whitelisting rules override other rules, so even if other rules happen to apply to the same touchpoint, the touchpoint will not be rejected or marked as suspicious.
Mark as Suspicious	If this method applies to an ad click or ad view, the click/view will be flagged as suspicious but the attribution decision will not be affected. You can then run a Suspicious Report to look at which of your attributions were marked as suspicious in the.
Reject	If this method applies to an ad click or ad view, the click/view will be rejected from consideration by Singular's attribution tracker. This means that the install will be attributed to the next click/view in line, or if there aren't any relevant ones, the install will be marked as organic.

Ensure You are Using In-App Purchase Validation

In-App Purchase (IAP) validation is a process in which Singular evaluates revenue events in your apps (both Android and iOS) and detects invalid/fraudulent user purchases so that they do not skew your revenue metrics.

Learn more in the In-App Purchase Validation FAQ.

Set Up Fraud Postbacks (Optional)

Singular can send automated postbacks about fraud decisions to the relevant ad network and/or to the customer's internal BI platform.

There are two types of fraud postbacks:

Rejection postback: Sent when an install would be attributed to a certain touchpoint, but the touchpoint is rejected by Singular’s fraud prevention system.
Notification about a suspicious install: This isn't a separate postback but rather a notification added to the postback triggered by the actual install. This notification is added to the postback only if it's sent to your internal BI system, and its purpose is to help you evaluate how well your "mark as suspicious" methods/rules are working before you set the method/rule to "reject".

See the Fraud Postbacks FAQ for more information and troubleshooting tips.

Leverage Singular Alerts (Recommended)

Singular lets you set up custom alerts on different KPIs so you can be automatically notified about sudden and/or significant changes in your data.

You can leverage alerts to track fraud in two ways:

Set up custom alerts on fraud data to be notified about significant changes in the number of rejections or installs marked as suspicious.
Set up custom alerts on any of your data to be notified when a source or a publisher suddenly performs extremely well or extremely poorly. These abnormalities may indicate fraud which you can then fight using custom rules.

Singular creates a default alert to get you started. This alert emails all admin users when there is an increase of rejected installs by 10% in the last 7 days. You may safely edit or discard the alert.

How to Set Up Singular Fraud Protection

Step by Step

Understand the Different Protection Methods

Set Up Your Protection Methods

Set Up Custom Rules (Optional)

Ensure You are Using In-App Purchase Validation

Set Up Fraud Postbacks (Optional)

Leverage Singular Alerts (Recommended)

Use Fraud Reports to Track Fraud Attempts