Form drop-off, and what it actually tells you

Drop-off is where people stop. It's the one metric of a form that points at a specific culprit instead of an overall number, which is why it's the first thing to look at when responses are thin.

Updated 27 August 2026

The three numbers, and why they differ

Opens: how many times the form was opened in a browser. Starts: how many people answered at least one question. Completions: how many reached the end. The biggest gap is almost always between the first two, and it's about the intro and the first question — not the questionnaire.

Completion rate calculated on «starts» is a kind number: it excludes exactly the people who opened and left, which is who a badly made form loses. Calculated on opens it's lower and truer.

The three typical drops

A drop at the open: visible when opens are many and starts are few. The cause is nearly always expectation — the form looks longer than promised, or it isn't clear who's sending it. You fix that outside the form, in the invitation.

A drop at one specific question: a required question asking something personal, or an open question placed too early. Fix it by making it optional, moving it later, or removing it.

A slow, steady decline: no single question is guilty, the form is simply too long. Nothing to fix here — something to cut.

How to read it without fooling yourself

With conditional logic the curve stops being monotonic: a question almost everyone skips looks like a cliff and isn't one. Before panicking, check whether that question sits inside a branch — in drop-off, «reached a question» means the path went through it, not that everyone had to answer.

The other trap is volume: across twenty responses, two abandonments on the same question are chance, not a signal. Under a hundred responses, drop-off is for forming a suspicion, not for deciding.

The one fix that nearly always works

Move the hardest question after an easy one. Someone who has already answered has invested something and carries on; someone who meets the biggest effort first closes the tab. It doesn't change the content of the survey, and it's often enough.

Look at where the curve falls: at the open it's the invitation, at one question it's that question, steadily throughout it's the length.

Start from here

Templates whose question order already accounts for this:

Build your formFree, unlimited responses. No card.

Read next