Detection strategy
Detection strategy is a set of filters the 24metrics use to define conversion fraud probability. You can select the needed strategy while creating an Anti-fraud rule.
There are four strategies:
Mobile CPI
Mobile CPA
Lead Offers
Ecommerce
Each of them contains a set of filters by 24metrics applied to the selected conversions. Here you can learn more about each filter.
⚠️ Affise doesn't define the filters set within a detection strategy. 24metrics do it on their side.
Mobile CPI
The best for mobile-app traffic (cost per install).
Filter name | Setting details |
Duplicate IP | Last 24h / 1 IP (same offer / same affiliate) |
Duplicate Device ID | Same offer / All affiliates |
Fraud Database | Global IP / Device ID based |
ClickSpam | 50% > 1h CTIT |
Low Session Time Anomaly | 50% < 1min CTIT |
Min Click To Install Time | 20s |
Bot | No configuration options |
VPN / Proxies | No configuration options |
Third Party Rejection | Sync with other fraud system |
Max Click to Conversion Time | 24h |
Sub ID Validity Check | More than 700 SubIDs (per offer + affiliate) |
Referrer URL | No configuration options |
OLD OS / Browser | 60% > Old devices |
Mobile CPA
The best for mobile-app traffic (cost per action).
Filter name | Setting details |
Duplicate IP | Last 24h / 3 IP (same offer / same affiliate) |
Duplicate Device ID | Same offer / All affiliates |
Fraud Database | Global IP / Device ID based |
ClickSpam | 60% > 2h CTIT |
Low Session Time Anomaly | 60% < 1min CTIT |
Min Click To Conversion Time | 15s |
Bot | No configuration options |
VPN / Proxies | No configuration options |
Third Party Rejection | Sync with other fraud system |
Max Click to Conversion Time | 24h |
Sub ID Validity Check | More than 700 SubIDs (per offer + affiliate) |
Referrer URL Blacklist | No configuration options |
OLD OS / Browser | 70% > Old devices |
Lead Offers
The best setting for lead offers.
Filter name | Setting details |
Duplicate IP | Last 24h / 1 IP (same offer / same affiliate) |
Duplicate Device ID | No duplicates (same offer / all affiliates) |
Fraud Database | Global IP / Device ID based |
Low Session Time Anomaly | 50% < 1min CTIT |
Min Click To Conversion Time | 25s |
Bot | No configuration options |
VPN / Proxies | No configuration options |
Third Party Rejection | Sync with other fraud system |
Max Click to Conversions Time | 24h |
Sub ID Validity Check | More than 500 SubIDs (per offer + affiliate) |
Referrer URL Blacklist | No configuration options |
OLD OS / Browser | 70% > Old devices |
eCommerce
The best setting for the eCommerce.
Filter name | Setting details |
Duplicate IP | Last 24h / 3 IP (same offer / same affiliate) |
Fraud Database | Global IP / Device ID based |
Click Spam | 50% > 24h CTCT |
Low Session Time Anomaly | 50% < 1min CTIT |
Min Click To Conversion Time | 40s |
Bot | No configuration options |
VPN / Proxies | Below 10% |
Third Party Rejection | Sync with other fraud systems |
Max Click to Conversions Time | 14 days |
Referrer URL Blacklist | No configuration options |
OLD OS / Browser | 70% > Old devices |
⚠️ The No configuration options record means the filter is active, but no specific parameters are indicated for it.
Fraud probability
Fraud probability allows defining the conversions to be declined automatically according to the selected options (see below). You can select the needed strategy while creating an Anti-fraud rule.
24metrics calculates scores depending on which type of abnormality the system detected for each conversion. Each type has a certain number of points. As a result, every conversion has its score and the fraud probability. Here you can learn more about what every level means.
⚠️ Affise doesn't influence the scoring process and fraud probability determination and calculation. 24metrics perform it on their side.
There are four options:
Do not decline: the system doesn't decline conversions.
High: the system declines conversions with the high risk level.
Medium and higher: the system declines conversions with the medium and higher risk level.
Low and higher: the system declines conversions with the low or higher risk level.
Please contact the Affise Customer Support team regarding all raised questions via the e-mail: [email protected].