Why form completion rate statistics disagree

Search for an average form completion rate and you will find several answers. Some are a few points apart. Some are twenty points apart. Each one is usually right about the forms it measured. The trouble starts when two of them are put side by side as if they measured the same thing.

What does a completion rate count?#

A completion rate is a simple fraction. It is the number of people who finished a form, divided by the number in a starting group. Every published figure makes three choices, and most headlines leave them out.

1. What counts as a start. It could be everyone who saw the form, everyone who opened it, or only people who typed a first answer.
2. What counts as a finish. Usually it is pressing the submit button. Some tools count reaching the last screen instead.
3. Whose forms. A figure from one tool's customers, one industry, or one kind of form describes those forms only. It does not describe forms in general.

Change any one of these and the number changes, even when the forms stay exactly the same. So a figure that does not say how it was counted is only a story, not a measure you can compare against.

Which number is which?#

Published averages often look alike, but they do not always measure the same thing. One figure may count from people who viewed a form. Another may count from people who started it. A third may come only from forms built on one tool. Before you put two figures side by side, check each one at its original source: what it counts from, whose forms it covers, and when the data was collected.

Why does the starting point move the number so much?#

Think of a form's visitors as groups that get smaller at each step. Some people see the form. Fewer open it or click into it. Fewer still type a first answer. The smallest group presses submit. Each step drops people who were never likely to finish. So a rate counted from a later step is always higher.

Here is a made-up example. Say 1,000 people see a contact form, 500 type a first answer, and 300 submit it.

The same form, on the same day, has a completion rate of 30% or 60%. It depends on one choice. Neither number is wrong. They answer different questions. The first asks how well the page turns visitors into inquiries. The second asks how well the form keeps people who already chose to begin.

What counts as finished on a conversational form?#

A conversational form has no single submit button. So "finished" needs its own meaning. In Qualiform, a conversation is marked completed when the AI says your sign-off line or when it runs out of turns. A conversation still in progress is only treated as abandoned after thirty minutes with no new messages. A completion rate there counts finished conversations, not button presses.

That makes comparing even harder. A visitor who answers two questions and leaves has still told the business something. A conversation can also end early because the request was not something the business does. Whether that counts as finished is one more choice. Before you put a conversational form's rate next to a regular form's rate, write down how each one decides a conversation or form is done.

Should you compare your form to a benchmark?#

Mostly, no. It is more useful to compare your form with itself. Use the same way of counting, the same kind of visitors, and two periods of the same length. If your rate goes up from one month to the next under the same rules, something you changed probably helped. If you switched tools in between, the change may be in the counting, not in the form.

A published average can still help. It can tell you whether your number is roughly normal, as long as you count the same way it does. If a published figure counts from people who started the form, compare it only with your own rate counted from people who started. If you compare it with your rate counted from everyone who saw the form, your form will look worse than it is. You might then cut a question that was doing useful work.

How do you read a completion rate you find online?#

Before you trust or repeat a figure, try to answer four questions about it.

1. What is the starting group? People who saw the form, opened it, or typed a first answer.
2. What counts as finished? Pressing submit, reaching the last screen, or a finished conversation.
3. Whose forms are in it? One tool's customers, one industry, or a mix.
4. When, and how many? The dates the data was collected, and how many forms or visitors it covers.

If the source does not say, treat the number as a rough story about someone else's forms. The form completion rate entry explains the two most common ways of counting, side by side.

A completion rate only means something if you know how it was counted. Write down how you count before you measure. Then compare next month's number with this month's. If you are measuring contact forms, the guide to qualifying leads covers what to ask once a visitor starts.