Pass rate

How many genuine users get through on the first try.

If 100 real customers start verification and 92 are approved without needing a retry, the pass rate is 92 percent.

The proportion of legitimate attempts resulting in approval, measured per cohort and document type. A cohort is a group being measured together, such as customers from one country or one acquisition channel, and a legitimate attempt is one made by a genuine person.

Pass rate is calculated by dividing approvals of genuine users by all attempts from genuine users. Measuring it takes care:

• Who counts as genuine, since the measure needs a way to separate real customers from fraudsters, often after the fact.
• First try or eventual, since some measure approval on the first attempt and others include retries.
• Segments, such as document type, device, and country, since an average can hide weak spots.

A falling pass rate usually has a cause that teams can investigate:

1. Poor capture, such as blur, glare, or bad lighting.
2. Unsupported or unusual documents.
3. Thresholds set too strictly.
4. Demographic or device differences in how well the checks work.

Pass rate sits next to terms that are easy to mix up:

• Conversion rate is the share of people who finish the flow, which includes people who give up.
• False rejection rate (FRR) is the share of genuine people wrongly rejected, which is close to the opposite of a high pass rate.
• Fraud rate is the share of events that are fraudulent, which tells the other side of the story.