At a glance: Fraud protection studio (FPS) provides greater control over how fraud detection is applied than usual methods. With FPS, you can choose whether fraud detection data is blocked and reported as fraud or tagged, empowering tailored fraud management.
About fraud protection studio
Fraud protection studio introduces a tagged detection status.
This status enables advertisers to detect fraud without automatically implementing actions, which provides you:
- Evaluation: Evaluate the impact of detection updates before they apply.
- Control: Tailor fraud detection to your business model and needs.
- Transparency: Tagged data provides insights on additional fraud protection layers.
Important!
Only the Admin, Marketing lead, Team manager, and Security permissions roles are able to access the Fraud protection Studio.
How the fraud protection studio works
Key concepts
FPS modes
FPS has two modes, implemented and tagged. Advertisers can switch between these modes based on their goals, ensuring that detection is aligned with their specific needs.
Current detection mechanism: Installs are either considered:
- Valid attribution
-
Detected
- Blocked In real-time
- Reported as post-attribution fraud
Mechanism with FPS: Installs are either considered:
- Valid attribution
-
Detected and
- Tagged but remain valid for attribution
-
Implemented
- Blocked In real-time
- Reported as post-attribution fraud
Note
In tagged mode, all fraud detections that are tagged but remain valid are considered valid according to the client configuration, and as such, we won’t credit back to the consumption plan
Where is tagged detection reflected?
Tagged detections aren't reflected in fraud reports. Tagged detections are available in regular attribution reports and the Protect360 dashboard as follows:
Reports:
- Real-time tagged installs
- Post-attribution tagged installs
- Real-time tagged in-app events
- Post-attribution tagged in-app events
Protect360 dashboard:
- Total tagged count in the upper page components (1)
- Overtime trend of “Detection Status” (2)
- Identified fraud by “Tag type”, “Tag reason,” “Tag Sub Reason” (3)
How to use fraud protection studio
General
FPS is a configuration page allowing you to specify when you want detections to be implemented or tagged. The platform helps define tagging requirements by providing:
- Detection status data.
- Option to switch between implemented and tagged modes.
To get started:
- Log in to AppsFlyer.
- From the left-hand menu, go to Settings > Protect360 > Fraud Protection Studio.
- Use the app dropdown at the top of the page to select the app you want to configure.
- Select the card from the left menu. the current options are:
Click flood detection
FPS makes managing click flood detection simple and customizable. Allowing tagging of click flood detections on the app level (all partners) and showing tagged click flood detection data in dedicated reports.
On the click flood card page you can see and adjust the detection status between implemented and tagged, and view the current detection status data.
Download the Click flood report data sample
Important!
- When tagged detection is selected they are reported in raw data reports, meaning, not in fraud reports.
- Tagging the click flood is on the app level and will prevent blocking or post-attribution implementation. Therefore, by tagging it, you are exposed to abuse on your non-organic and organic installs.
How to access Click Flood
- Log in to AppsFlyer.
- From the left-hand menu, go to Settings > Protect360 > Fraud Protection Studio.
- Use the app dropdown at the top of the page to select the app you want to configure.
- From the left menu, select the Click Flood card
Click flood page
The click flood page is broken into three sections.
- Detection status - Allows you to choose the detection mode. you can choose between implemented and tagged
-
Current detection status - Shows how enabling Click Flood would have affected your installs and relative events based on the last 7 days prior to last week. This section is further broken up into:
- Current detection widget
- Click Flood detection breakdown widget
- Click Flood by media source widget
How to adjust the detection status
The detection status allows you to choose the detection mode. you can choose between implemented and tagged.
- In the Click Flood window, choose the detection mode.
-
Click the Save button at the bottom of the page.
Current detection status
The current detection status section shows an estimation based on the last 7 days prior to last week.
The current detection status section shows the following:
- Current detection widget
- Click Flood detection breakdown widget
- Click Flood by media source widget
Current detection widget
- Total installs - The total amount of app installs
- Click flood detection - The total detected click floods
- Total detection (without click flood) - The total amount of fraud detections not including click flood
- Total detection ( with click flood) - The total amount of fraud including click flood. Calculated by: (Click flood detection + Total detection (without click flood) )
Click flood detection breakdown widget
This shows the breakdown of the Click Flood detection by organic and non-organic installs.
Click Flood by media source widget
Shows the detected click Flood installs broken down by media source.
Note: for a more detailed report, click the Export button to download the detection metrics as a CSV file for offline analysis.
New releases
With New Releases, you can evaluate the impact of new fraud detection updates before implementation. By default, each new fraud detection package starts in tagged mode, giving you time and control to evaluate its impact.
Download the New Release report data sample
Note
A detection package is one or multiple detection updates that are grouped and released together.
How this works
- Fraud detection updates are grouped into packages.
- Fraud detection packages start in tagged mode.
- New packages can be set to tagged mode for up to 14 days.
- The current active package can be evaluated and implemented.
- You will be notified of new package releases through the notification bell.
How to access New Releases
- Log in to AppsFlyer.
- From the left-hand menu, go to Settings > Protect360 > Fraud Protection Studio.
- Use the app dropdown at the top of the page to select the app you want to configure.
- From the left menu, select the New Release card.
New releases page
The New Releases page is broken up into two sections.
- Future Releases - Allows you to evaluate the impact of new fraud detection packages by marking detections as tagged for up to 14 days before they come into effect.
- New Releases - Allows you to evaluate the current active package by monitoring its detection performance and implementing it now.
How to set the tagged mode time period for future packages
Future Releases allow you to set the number of days you want future detection packages to be marked as tagged before they go into effect. To set this:
- In the New Release window, from the dropdown box, select the number of days you want to mark new packages as tagged before they get implemented (default: 14 days, adjustable up to 14 days)
- Click the Save.
Note
- This only affects future packages. Detection packages that are currently in effect will not get delayed or changed.
- Selecting 0 days from the dropdown box is equivalent to selecting Implement now, and overrides the the Future Release settings for this package.
How to evaluate and implement current active packages
The New Current Release section allows you to see:
- The last package released
- Released date
- Status - when this package will go into effect (if this package is in tagged mode).
- Metrics on how this detection package will affect your data
- Export detail reports
- Bypass the Future Releases settings and implement this detection package now
To implement the current package if in tagged status.
- Open the New Release Card in the FPS tab.
- Monitor detection performance using the table, and exportable reports.
- When ready, click Implement Now to apply the changes.
Note
This overrides the Future Release settings for this package.
App detection evaluation
The App Detection Evaluation Card allows you to change a detection’s status based on your business objectives. Decisions made here are applied at the app level and affect all of the advertiser’s partners.
Tagging a detection on the App Detection Evaluation Card impacts all Protect360 logic. However, detections triggered by your defined validation rules are not affected. These detections remain in Implemented status unless you manually deactivate the corresponding validation rule.
On the App detection evaluation page you can see and adjust the detection status between implemented and tagged, and view the current detection status data.
Important!
- When tagged detection is selected, they are reported in raw data reports, meaning not in fraud reports.
- Tagging the app detection evaluation is on the app level and will prevent blocking or post-attribution implementation of Protect360 logics. Therefore, by tagging it, you are exposed to abuse on your non-organic and organic installs.
- Any validation rule defined and activated will always remain in implemented mode. To pause the implementation of any validation rules, go to the validation rules page and deactivate it there.
How to access App detection evaluation
- Log in to AppsFlyer.
- From the left-hand menu, go to Settings > Protect360 > Fraud Protection Studio.
- Use the app dropdown at the top of the page to select the app you want to configure.
- From the left menu, select the App Detection Evaluation card.
App detection evaluation page
The app detection evaluation page is broken into two sections.
- Detection status - Allows you to choose the detection mode. you can choose between implemented and tagged.
-
Current detection status - Shows the current detection affecting your installs and relative events based on the last 7 days prior to last week. This section is further broken up into:
-
Protect360 logics
Implemented detection mode:- Implementation rate - the rate represents (the amount of logic implemented by Protect360 logics) divided by (the app’s total attributions).
-
App-level
Tagged detection mode:- Tagged rate - the rate represents (the amount of logic tagged by the app detection evaluation tagging) divided by (the app’s total attributions)
- Total installs
- Total detections
-
Implemented detection - Used to assist you before making the decision to change the detection status from implementation to tagging
- Protect360 logics - This shows how much of the detection is by Protect360 logics. When changing to tagged - this is the expected amount of detection which will be converted to tagged detection
- Validation rules—This shows how much of the detection is based on validation rules you have defined and activated. If you change the detection mode to tagging, validation rules-based detection will remain in the implemented mode.
-
Tagged detections - Used to assist before making the decision to change the detection status from tagging to implementation. You will have a reflection of all the tagging reasons to reflect the expected outcome before changing the detection status from tagged to implemented
- App level - This KPI allows you to understand how much of the tagged is due to tagging made by the app detection evaluation card. The number reflected in this KPI is the expected increase if you implement the detection status.
- Other tagging - You can tag the detection in multiple capabilities (cards) in the fraud protection studio.
-
Protect360 logics
How to adjust the detection status
The detection status allows you to choose the detection mode. You can choose between implemented and tagged (the default is implemented)
- In the app detection evaluation page, choose the detection mode.
- Click the Save button at the bottom of the page.
Re-engagement hijacking detection
The Re-engagement hijacking detection in Protect360 expands the standard fraud detection coverage specifically for re-engagement activities. This detection enhances your control over fraud management.
By default, this detection is released in tagged mode, allowing you to evaluate its effectiveness and potential impact before implementation.
How this works
- The re-engagement hijacking detection is set to tagged mode by default. You can evaluate its detection performance without impacting your live attribution data.
- When implemented, detections occur post-attribution, meaning fraudulent re-engagements are identified and removed from attribution records.
- When implemented, all in-app events following a detection removal will be re-attributed to the original User Acquisition (UA) provider or marked as organic if no valid UA provider exists.
How to access
- Log in to AppsFlyer.
- From the left-hand menu, go to Settings > Protect360 > Fraud Protection Studio.
- Select the app you wish to configure using the app dropdown at the top.
- From the left menu, choose the Re-engagement Hijacking card.
Re-engagement hijacking page
The Re-engagement hijacking page is broken into three sections.
- Detection status - Allows you to choose the detection mode. You can choose between implemented and tagged
-
Current detection status - Shows how enabling Re-engagement hijacking would have affected your installs and relative events based on the last 14 days (excluding the current day).
Note: Since Re-engagement Hijacking Detection operates post-attribution, data from the last 7 days may be partial and continue to update retroactively.
This section is further broken up into:- Re-engagement hijacking detection widget
- Detections - the number of detections by the re-engagement hijacking logic alone
- Detection rate - percentage of re-engagement hijacking detections out of the re-engagement detection
- Total Re-engagements detection widget - the amount of overall re-engagement detection from all logics
- Re-engagement attribution volume widget
- Attributions - The amount of re-engagement attribution in the timeframe
- Rate out of total attributions - the portion of re-engagement attributions out of all app’s attributions
- Re-engagement detection by media source - The number of re-engagement hijacking detections for the top detected media sources
- Re-engagement hijacking detection widget
FAQs
Can I configure tagged detection for partners or media sources?
This depends on the detection type. For example, partner tagging is unavailable for hijacking detection to prevent biased detection.
What happens when the tagging period ends?
At the end of the evaluation period, tagged detection automatically converts to implemented mode.
Where can I see tagged installs?
Tagged data is displayed in four dedicated reports.
Can tagged detection be adjusted for app-specific settings?
Yes, click flood tagging is configurable at the app level.
Can I see tagged in-app events?
Yes. In-app events that are connected to a tagged install by the same AppsFlyer ID will also be tagged.
Who can change, set up, or view FPS?
Only the admin can change the set up of FPS. FPS is visible according to the permissions page.
Are tagged installs credited back to the consumption plan?
No. tagged detection will not be credited back to the consumption plan.
Are tagged installs considered as fraud?
Tagged installs are detected by Protect360. However, they are tagged per the customer configuration. Therefore, both in the Overview and in the Protect360 dashboards they will be reflected under the attributed metrics.
Can I see how many installs are tagged only by the click flood tagging?
Yes. This data is available in the new dedicated reports on the export page and on the Protect360 Dashboard.