Form abandonment: why visitors quit your forms

Form abandonment is when someone starts filling in a form and leaves without sending it. It's common: in Zuko's benchmark of 93 million form visits, only about half the people who started a form went on to submit it — a bit more than half on desktop, a bit less on phones.

Most visitors never get that far. The majority of people who open a page with a form leave without clicking into a single field. Put the two together and out of everyone who sees your form, only around one in three submits it.

You may have seen a scarier number floating around — "81% of people abandon forms." It comes from a 2018 survey that asked whether people had ever quit a form, not how often forms get quit, so it isn't a rate to plan around. And checkout forms are their own, worse story: seven in ten shopping carts are abandoned before payment, for reasons that get their own post.

Survey the abandoners and two reasons dominate: they didn't trust the form, or the form asked too much. In the study roundups, security concerns lead (about 29% cite fear of spam or data misuse), form length is right behind (27%), and a long tail covers ads mid-form, questions with no obvious purpose, and forms that break on phones.

The distrust is field-specific, and one field is the standout offender: the phone number. Around 37% of abandoners quit on a required phone field — everyone knows what happens after a company has your number — and making it optional roughly doubles its completion. Fields asking for addresses or government IDs carry the same smell of future consequences. Email, by contrast, barely registers; people expect to hand it over.

Length works differently than most teams assume. The famous case studies about deleting fields are old and shakier than their fame suggests, and Baymard's checkout research found field type matters as much as count — a dropdown costs more than a text field, a text area more still. The pattern that survives scrutiny: every question is a small cost, required questions cost more, and questions whose purpose the visitor can't guess cost the most.

GA4's enhanced measurement gives you two form events for free: form_start, the first interaction, and form_submit. Divide submits by starts, per form, and you have your abandonment rate — a number worth checking weekly next to your conversion numbers. GA4 won't compute the ratio for you, and it counts a submit even when validation rejected the input, so treat the rate as approximate.

What pageview-level analytics can't tell you is where the form dies. That takes field-level data — which field was last touched before the visitor left, how long each field took, which ones got filled, deleted and refilled. Dedicated form analytics tools measure exactly this, and the rule of thumb in that world is that any field losing 15% or more of the people who reach it has earned an investigation.

The other route to the same answer is watching it happen. Session recordings of visits that fired form_start but not form_submit show you the visitor reaching the phone field, pausing, scrolling to check whether it's required, and closing the tab — the kind of moment no aggregate metric can narrate. In Bigdelta, filtering replays by page and event gets you that shortlist in a couple of clicks.

Start with the fields, not the button color. Cut every question you can't defend, make the phone number optional or drop it, and move nice-to-know questions out of the critical path — you can always ask later, inside the product or the reply email. Where a question looks intrusive, say why you're asking in a line under the label; "for delivery updates only" defuses the phone field for a share of the people it was scaring off.

Then reduce the cost of answering what remains. One column, not two — multi-column forms measurably slow completion. Inline validation that flags a bad email while the visitor is still there beats an error page after submit, and an error message that says what's wrong beats a red border. On mobile, the right keyboard per field (numeric for numbers, email layout for emails) is a five-minute fix for the platform where abandonment runs highest.

Trust is the last lever: a form on a page with a privacy line, a real company footer and an SSL padlock loses fewer of the security-anxious 29% than the same form floating in a void.

Then re-measure the start-to-submit rate and see whether the needle moved. Forms are one of the few places on a website where a fix shows up in the numbers within days.

Half the people who start your form giving up is normal; it is also where signups, leads and sales quietly leak. Compute starts-to-submits per form, find the field where the drop concentrates — with field-level analytics or a few recordings of abandoners — and fix in that order: fewer questions, optional phone number, clear labels, inline errors, mobile keyboards. Every field you delete is a conversation you can have after the signup instead of a reason not to get one.