One question at a time forms: when they beat a single page, and when they don’t

A one question at a time form puts each question on its own screen, with a button or the Enter key to move to the next. It’s also called a one question per page form, a conversational form or a Typeform-style form, after the tool that made the layout well known. The alternative is the classic single-page form, where every field sits on one scrolling page.

Most tools that work this way let you bend the rule on purpose. A “page” screen can hold a few fields that belong together, such as first name, last name and email. The GOV.UK Design System (opens in a new tab) describes the same pattern: one question per page by default, though “Sometimes it makes sense to group a number of related questions on the same page.”

The strongest case comes from UK government design guidance, and it’s about clarity, not conversion: one question at a time is easier to understand, answer and correct. The GOV.UK Service Manual (opens in a new tab) recommends starting with one thing per page because it helps people:

• “understand what you’re asking them to do”;
• “focus on the specific question and its answer”;
• “find their way through an unfamiliar process”;
• “use the service on a mobile device”;
• “recover easily from form errors”.

It helps the people running the form too. GOV.UK notes that one thing per page makes it easier to “save a user’s answers automatically as they go”, “capture analytics about each question” and “handle branching questions and loops”. Those three are the basis of measuring form abandonment and of conditional logic.

Survey research adds some support. In a 40-question web survey shown with 1, 4, 10 or 40 items per screen, Toepoel, Das and van Soest (2009) (opens in a new tab) found, according to the paper’s abstract, that more items were left blank as more items shared a screen, and that people rated the multi-item layouts less favourably.

Not reliably. Survey experiments that compare paging with scrolling mostly find small or no differences in breakoff, meaning how many people quit, and we found no controlled study that compares one question at a time forms with single-page forms.

• SurveyMonkey’s summary of the research (opens in a new tab) says earlier studies on computers “didn’t find much difference between these two formats in terms of survey completion rate and break-off rate”. Some found scrolling led to more skipped questions.
• On phones, Mavletova and Couper (2014) (opens in a new tab) found lower breakoff with scrolling, “though not significantly lower”. Scrolling was significantly faster, had fewer technical problems and got higher ratings. A later meta-analysis by the same authors (opens in a new tab), pooling 14 studies, found no effect of scrolling design on breakoff.
• Toepoel’s study measured something different: as more items shared a screen, more items were left blank, but that is people skipping questions, not quitting. Multi-item screens were also quicker to finish.
• In checkouts, the Baymard Institute (opens in a new tab) found that “the number of form fields in a checkout impacts overall usability far more than the number of steps”.

So treat one question per screen as a pattern for clarity, phones and error recovery. Expect it to make a form easier to answer, not to rescue a form that’s too long.

A single page is often the better choice when the form is very short, when the same people fill it in again and again, or when answers depend on each other and people need to see them together.

The last row is a known limit of step-by-step designs. Nielsen Norman Group calls wizards “the classic example of staged disclosure”, and notes that staged designs struggle when steps depend on each other and people have to keep moving back and forth (NN/g on progressive disclosure (opens in a new tab)).

In checkouts, Baymard found the number of fields mattered more than the number of steps, though usability “does seem to suffer after 8+ checkout steps”. Treat that as a hint for other forms rather than a rule: cutting questions is a safer bet than reshuffling them.

Order questions from easy to hard, put related fields on one screen, show progress honestly, make every screen work from a keyboard and a thumb, and let people go back without losing answers.

Ribbit shows one question per screen on every form, and Typeform, Youform, Jotform and Paperform offer a one-question layout too. Tally, Fillout and Google Forms describe multi-page forms or sections instead, where you can put one question on each page yourself.

For a fuller side-by-side, see Ribbit vs Typeform, or the roundups of Typeform alternatives and Youform alternatives.

Every Ribbit form shows one question per screen, with multi-question pages for fields that belong together. You describe the form in one sentence and Ribbit’s AI form builder drafts the questions; then you edit everything by hand. Our guide to writing AI form prompts covers what to put in that sentence.

• Pages for related fields: a page puts several questions on one screen, in a stacked or compact layout.
• Progress: a thin progress bar and a “3 of 10” counter. On forms with logic the counter is hidden and the bar stays.
• Keyboard: Enter moves on, Ctrl or Cmd+Enter moves on from a long text answer, letter keys pick choices on single- and multiple-choice screens, and a single-choice answer moves on by itself.
• Back and errors: Back retraces the path the person took and keeps their answers. Each screen is checked when they move on, not at the end.
• No waiting between screens: the whole form loads at once, so moving between screens doesn’t wait on the network.
• Phones and accessibility: a mobile-first layout with tap targets of 48 px or more, built with keyboard and screen-reader support in mind; focus moves to each new screen’s input.
• Preview: try the form at desktop and phone size before it goes live.
• No account for respondents: people open your link and answer.

For a long run of same-scale ratings, put a few related statements on one multi-question page, so people aren’t tapping through a screen for each. Pro adds drop-off analytics for each question and partial answers, so you can see which screen people leave on (pricing).