Drop-off rate, the metric that decides whether your KYC flow works

Solidus measures drop-off in its own verification flow. This is one of the operational metrics we actually track, and this page defines it for a buyer who will ask.

What it is

The proportion of people who start a verification and do not finish it.

Not failures. A drop-off is someone who abandoned: closed the tab, gave up on a document capture, declined to continue at a consent screen, or ran out of patience. They are not rejected: they are gone, and in most businesses they are gone as customers too.

That is why this is the metric that decides whether a KYC flow works commercially. A verification process with excellent accuracy and severe drop-off is a bad process. It is also the metric a vendor is least likely to volunteer, because it measures the vendor's own friction rather than the end user's honesty.

What actually moves it

From building the flow, the drivers are consistent and mostly unglamorous:

• Capture difficulty. Glare, focus, and document framing. Most abandonment happens with a camera open, not at a decision point.
• Number of steps, and whether their purpose is legible. People tolerate steps they understand. An unexplained selfie request after a document scan loses people who would have completed either one alone.
• Device and context. Mobile, one-handed, poor light, moving. The verification is never the reason the person opened their phone.
• Retry behaviour. Whether a failed capture is recoverable in-place or restarts the flow.
• Language. A flow in the user's second language drops off harder, which is why the hosted flow's copy is a product decision and not a translation afterthought.

Where a portable credential changes the shape

For a returning verified person, the drop-off surface largely disappears, presenting a held credential removes the camera, the capture retries and the multi-step flow that generate most abandonment.

That is the same structural point as re-KYC and remediation: reuse is worth most exactly where the repeated flow is the thing costing you customers. The argument is good; what is unverified is our implementation, not the shape of the idea.

The bound: it only applies to people who already hold a credential. First-time verification has the same drop-off surface as anyone else's, and we do not claim otherwise.

Why there is no number on this page

We track drop-off. We will not publish a figure, and you should be suspicious of vendors who do.

Publishing it would be a real measurement used misleadingly, which is worse than no measurement.

When there is volume worth measuring, the number goes here with the conditions attached. Until then, ask any vendor what their drop-off figure was measured on before you compare it to anything.