Skip to main content

Score investigation

Audience tip

This article is relevant for Fraud Management SaaS users.

The score investigation tool is designed to help you find the right score threshold when writing rules in Rule manager. Select a score and check how it affects the following KPIs:

  • Decline rate: the percentage of transactions declined by the Fraugster AI score from the total number of transactions.

  • Catch rate: the percentage of fraudulent transactions declined by the Fraugster AI score from the total number of transactions tagged as fraudulent.

  • Good user decline rate: the percentage of declined transactions that were actually good from the total number of good transactions.

  • Precision: the percentage of fraudulent transactions from the total number of declined transactions.

To use the score investigation tool, drag the knob on the slider to the score you wish to investigate. The data in the KPIs change accordingly.

Additionally, you can filter your results further by choosing a specific merchant or sub-merchant, and switching between value and count. The score investigation tool works with the live data in production as well as historical data. This is different to the rule simulation tool that runs on recent production data only. Therefore, you might see slight discrepancies in the results.

As the Fraugster Engine improves, the models for the score investigation tool also get updated. You can expect the results in the score investigation tool to be even better than in the rule simulation tool.

Score investigation is available to customers with at least 6 months of production data. This is the minimum required amount of data in order for the tool to provide meaningful insight. If you see that score investigation is disabled for you, it means that Fraugster doesn't have sufficient production data for your account yet. Please get in touch with us if you have any questions or concerns in this regard.

Last updated on