GreetAI

Form abandonment statistics 2026, and where voice answers help

11/9/2026 • 4 min read

Every team with a lead form, intake form or application has the same question: why don’t people finish it?

Most “form abandonment statistics” pages recycle numbers nobody can trace. Below are only figures we could follow back to the organisation that published them, followed by what they mean for how you design a form in 2026.

The numbers

65% of people who see a form never complete it. Zuko’s benchmarking (January 2022 to April 2025) puts average view-to-completion at 34.6%. Of the people who actually start, 51.7% finish.

Phones lose more people. Starter-to-completion is 55.5% on desktop and 47.5% on mobile in Zuko’s data, an 8-point gap. Statcounter shows mobile at about 60% of web traffic in October 2026, so the weaker device is now the main one.

Some form types are much worse. In Zuko’s data, contact forms convert only 14.5% of viewers, applications 37.7% and registrations 46.6%.

Autofill doubles completion. Forms completed at 62.2% when autofill kicked in and 30.2% when it didn’t (Zuko). Anything that saves typing pays off.

Length matters less than effort. Zuko found total length barely moves completion rate. What hurts is how hard the form feels in the first minute, and fields that need real thought.

What actually drives drop-off

Put those together and a pattern shows up:

  1. Typing is the cost. Autofill doubles completion because it removes typing. Mobile underperforms because typing on glass is slow.
  2. Open questions are the wall. “Describe your situation” is where effort spikes, and where the most useful answer lives.
  3. First impressions decide. A form that looks like work loses people before they start.

Where voice answers help

If typing is the cost, letting people speak is the obvious experiment. Most people speak several times faster than they type on a phone, and answers that feel like a chore to write (“what brings you in?”, “tell us about your business”) are easy to say.

Research backs part of this. A 2026 study in the Journal of Survey Statistics and Methodology (1,461 US respondents) found spoken answers to open questions were longer and slightly more informative than typed ones on most questions.

Where voice hurts

The same study also found that when people were forced to answer by voice, many more dropped out, and only 47% gave open answers versus 99% when typing. Some people are in an open office, on a bus, or just don’t want to talk to their phone.

So the rule is simple: offer voice, never require it.

A voice form checklist

If you try a voice form, these choices decide whether it lifts completion or hurts it:

  • Speak, type or tap on every question. Multiple choice should stay one tap.
  • Ask for the email first, in a field. Then the conversation. If someone leaves, you can still follow up.
  • Keep it short. Three to six questions. Let follow-ups do the digging.
  • Follow up only when an answer is vague. “Soon” becomes “this month or next quarter?”
  • Return structured fields. Your CRM needs a budget range and a phone number, not a recording.
  • Measure starts and finishes, not just submissions, so you can compare against your old form.

Try it on your longest form

Greet voice forms follow every point above: people talk, type or tap, the AI asks a follow-up when it needs to, and you get clean fields, a summary, the transcript and the recording in one list. It goes on your site with one script tag.

See how a voice form works or create one free.

Sources

Form abandonment statistics 2026, and where voice answers help