Designing forms people actually finish

We rebuilt our intake form builder three times before shipping the version customers use today. The first version had more features. The third has fewer, and it converts noticeably better — internally we saw completion rates climb from the high 50s into the high 70s across our own template forms after the changes below, though your mileage depends heavily on what you're asking for.

One column, always

Multi-column layouts look tidy on a design file and confuse real users, who read forms top to bottom regardless of how the fields are arranged visually. Every field we moved from a two-column layout into a single vertical column reduced the number of people who skipped it entirely. It sounds too simple to matter. It's one of the highest-leverage changes we made.

Show the finish line early

A progress indicator — even a crude one, "Step 2 of 4" — measurably reduces abandonment on anything longer than about six fields. People will tolerate a long form if they know how long. They abandon short forms that feel open-ended because there's no sense of when it ends.

Autosave changes the psychology, not just the mechanics

Once we added autosave to in-progress forms, something changed beyond the obvious technical benefit of not losing data. Users started forms they weren't sure they'd finish in one sitting, because leaving and coming back stopped feeling risky. That's a bigger behavioral shift than it sounds — it turns a form from a single commitment into something you can chip away at.

Ask for less than you want

Every optional field is a tax on completion, even ones that take five seconds to answer. Our rule now: if a field isn't required to process the submission, it doesn't go on the form — it goes in a follow-up email or a CRM enrichment step instead. Legal and sales both wanted more fields on our own signup form. Completion went up 11% the week we cut it from eleven fields to six.

None of this is exotic. It's the same handful of principles that have shown up in conversion research for over a decade. The hard part isn't knowing them — it's having the discipline to cut a field a stakeholder asked for.