KlientBoost
Apexure
MarketingExperiments
Venture HarbourA multi-step form splits one long form into a sequence of short screens, usually with one or two questions per screen and a progress indicator.
In the lead-gen and SaaS tests in our database, splitting a long form into steps usually won, often by a lot, but the one clean loss shows the rule: steps help when they make the form feel easier, and hurt when they just add pages before the value.
This page pulls out what won, what lost and what the patterns are.
The full grid of every tagged test sits below the analysis, and you can browse everything else in all tests or the element index.
Multi-step form best practices, from the tests that won
The strongest evidence comes from teams that tested the same form more than once.
HomEquity Bank, which sells reverse mortgages to Canadians aged 55 and over, ran a sequence of tests on one inquiry form and published each step.
- Splitting a one-page form into seven steps lifted the inquiry conversion rate by 154% (2015).
- Cutting those seven steps to four then added 53% more conversions (2016).
- Adding a progress tracker to the multi-step form added 35% (2016).
- A later version that replaced the tracker with one explaining what happens after you submit added another 20% (2017).
The lesson is not "more steps is better".
Seven steps beat one page, and four steps then beat seven.
The first win came from making the form look approachable; the second came from removing steps that added clicks without adding ease.
Reassured, a UK life insurance broker, found the same direction with a cleaner test.
Its multi-step quote form, one question per step with a progress bar on top, lifted form submissions by 31.23% over a long single-page form across about five weeks.
The lift was larger on desktop than on mobile.
Apexure's work for Reclaim My PPI Tax shows the value of easy questions first.
A four-stage claim form that opened with the simplest questions moved end-to-end conversion from 1.77% to 4.33%.
Only 25% of visitors finished step one, so Apexure then tested a gateway question before the form, which moved conversion again from 4.33% to 6.09%.
Both ran for about a month on paid traffic split in Unbounce, though the case study gives no traffic counts.
Put together, these winners share five habits:
- Start with the easiest, least personal question. Ask about the need (coverage type, project size, loan amount) before name, email and phone.
- Keep each step to one idea. One or two questions per screen, with large tap targets for multiple-choice answers.
- Use as few steps as the questions need. HomEquity's second win came from cutting steps, not adding them.
- Tell people where they are and what comes next. A progress tracker helped, and explaining the next steps after submitting helped more.
- Ask for contact details last. By then the visitor has put in effort and has a reason to finish.
For field-level choices inside each step (labels, optional fields, phone number fields), see the form fields hub.
Progress bars: helpful, but not automatically
Two of our tests isolate the progress indicator.
Besides HomEquity's +35%, Tinkoff Bank tested two progress-bar designs on a four-step credit card application with 190,000+ visitors over 39 days.
A green bar that filled up as the applicant progressed lifted completed applications by 12.8%; a yellow bar with a progress line lifted them by 6.9%.
Both bars also showed an approval likelihood that rose as people filled in the form, so this tested a motivating message as well as a progress display.
Research on surveys is less kind to progress bars.
A 2013 meta-analysis by Villar, Callegaro and Yang of 32 randomized web-survey experiments found that a constant-speed progress indicator did not significantly reduce drop-offs.
Indicators that moved fast at the start and slower later reduced drop-offs, while slow-to-fast indicators increased them.
The practical reading: a progress bar on a long form that looks barely started after step one can discourage people.
Short forms, early quick steps, and a bar that shows real movement are the safer bet. Test it rather than assuming it helps.
Multi-step vs single-step form: when each wins
Most published multi-step tests compare a new multi-step form with an old, long single-page form.
That comparison favors the multi-step version, because the new form usually also has better questions, better order and better design.
Keep that in mind when you read a big lift.
The clearest loss in our database makes the point.
MarketingExperiments tested a three-step free trial signup on RegOnline's homepage against the existing two-step version.
The three-step variant asked for contact information first, then showed two more forms, and completed leads fell 24.5%.
MarketingExperiments later reached +141% on that page by removing navigation and walking visitors through the product's value first, but the extra step on its own lost.
Britannica's test points the same way. Removing one whole page from the subscription registration path lifted free trial conversion by 117%, paid subscriptions by 61% and revenue by 130%, in a 2008 MarketingExperiments test.
So steps are not the active ingredient. What wins is lower perceived effort and a better question order.
What loses is extra pages, especially ones that ask for personal data before the visitor sees value.
| Situation | Likely better | Why, based on the tests |
|---|---|---|
| Long qualifying form (quotes, insurance, mortgages, B2B inquiries) | Multi-step | HomEquity, Reassured and Reclaim My PPI Tax all won by breaking up long forms |
| Short form, 3-4 fields (newsletter, simple demo request) | Single step | Steps add clicks with no ease to gain; Britannica won by removing a page |
| Signup where value comes after the form (free trials) | Fewest pages possible | RegOnline lost 24.5% when a step was added before the trial |
| Paid traffic arriving cold | Multi-step with an easy first question | Reclaim's gateway question lifted conversion from 4.33% to 6.09% |
| Form that already has steps | Fewer steps, plus a progress or next-steps cue | HomEquity gained from cutting seven steps to four, then from trackers |
Why an easy first step might work (and what research really shows)
The usual explanation is the foot-in-the-door effect. In Freedman and Fraser's 1966 study, 52.8% of people who first answered a short phone survey later agreed to a two-hour home visit, versus 22.2% of people asked cold (36 people per group).
Later reviews were more cautious. A 1984 meta-analysis by Dillard, Hunter and Burgoon found a modest average effect (a correlation of about 0.17) even under the best conditions, and found it needed a prosocial topic.
A sales inquiry is not a prosocial request, so the psychology is at best a plausible story, not a guarantee.
The A/B tests are the better guide here.
They show easy-first, one-question-per-step forms winning on long lead forms.
They do not show that any form gets a lift from being split up.
Multi-step form conversion rate: what the numbers look like
There is no reliable industry benchmark for multi-step form conversion rates.
Published figures mix landing page visits with form starts, and vendors selling form tools tend to publish their best cases.
The before-and-after rates in our database give a sense of range instead.
| Test | Before | After | Evidence |
|---|---|---|---|
| Reclaim My PPI Tax, four-stage form (Apexure) | 1.77% | 4.33% | A/B test, about a month |
| Reclaim My PPI Tax, gateway question (Apexure) | 4.33% | 6.09% | A/B test, about a month |
| Venture Harbour consulting enquiry form | 0.96% | 8.1% | Before/after over several years |
| Venture Harbour B2C finance lead form | 11% | 46% | Before/after, method not given |
Treat the Venture Harbour figures with care. They are before/after numbers from years of iteration, not split tests, and the company later sold a multi-step form tool.
The same caution applies to the many agency case studies tagged with this element.
KlientBoost reports lifts such as +143% for Pike13 and +329% for Juniper Networks after moving clients to multi-step landing page forms, but those numbers cover whole paid-media engagements in which ads, keywords and pages changed together.
Performics India's +40% for a paint cost calculator cut from five steps to three came from a one-week test with several changes at once.
These cases show that teams keep choosing multi-step forms; they do not tell you how much the steps alone added.
Measure the rate that matters for your form: completed submissions divided by visitors who saw the form, plus the drop-off at each step.
A form that converts better but sends weaker leads is not a win, so track lead quality or sales accepted leads too.
How to test a multi-step form
- Instrument every step first. Record how many people start, reach each step and submit. The step with the biggest drop is where your test should focus, as Apexure found with its 25% step-one completion.
- Test one structural change at a time. Single page vs steps, then step count, then question order, then progress cues. HomEquity's sequence shows how much each change can be worth on its own.
- Keep the questions identical when testing format. If the multi-step version also drops fields, you will not know which change caused the lift.
- Size the test before you start. Lead forms often get modest traffic. Use the sample size calculator to see how long you need to detect a realistic lift; Tinkoff needed 190,000+ visitors and 39 days for its result.
- Check mobile and desktop separately. Reassured's lift was higher on desktop.
- Follow leads downstream. Compare lead quality, contact rate and sales, not just submissions.
Most landing page builders and form tools can build step-based forms with conditional logic.
We compare them in our multi-step form builders guide, and the form optimization guide covers the rest of the lead form.