The best form A/B tests in our archive made forms feel shorter and safer, not just prettier.
HomEquity Bank lifted conversions 154% by splitting a one-page inquiry form into seven steps, Britannica got +117% free trials and +130% revenue by cutting one registration page, and SafeSoft doubled leads by showing the price next to its form.
How we ranked them: we pulled every lead-gen and sign-up form test from the LiftArchive database, kept the winners that came from an A/B or split URL test with a published number, and ranked them by size of lift, moving cleaner single-change tests up.
We did not run these tests; every number comes from the company or agency that did, linked on each test page.
Many come from vendor case studies, which tend to publish wins, so read the sample notes.
For the how-to version of this page, see our form optimization guide.
For more results across lead gen, see lead gen A/B test results or the full research hub.
The ranked table
| Rank | Test | Change | Result | Who ran it |
|---|---|---|---|---|
| 1 | HomEquity Bank | One-page form split into 7 steps | +154% conversion rate | HomEquity Bank |
| 2 | Reclaim My PPI Tax | Long claim form split into 4 stages | 1.77% to 4.33% conversion | Apexure |
| 3 | Britannica | One registration page removed | +117% free trials, +130% revenue | MarketingExperiments |
| 4 | SafeSoft Solutions | Price shown next to lead form | +100% leads | SafeSoft Solutions |
| 5 | VenueSphere | Benefit-led subheadline above form | +69% form submissions | VenueSphere |
| 6 | Vendio | Form moved to its own page | +60% sign-ups | Vendio |
| 7 | Casa Mineira | Free-text search to 2 drop-downs | +57.25% form submissions | Supersonic |
| 8 | Hotel Institute Montreux | Testimonial above brochure form | +50% form submissions | The Reference |
| 9 | Tom's Planner | 4-field form above the fold | +44% sign-ups | Tom's Planner |
| 10 | Tinkoff Bank | Option to give passport details later | +35.8% applications | Tinkoff Bank |
| 11 | HomEquity Bank | Progress tracker on multi-step form | +35% conversions | HomEquity Bank |
| 12 | Reassured | One question per step plus progress bar | +31.23% form submissions | Reassured |
| 13 | Kareo (DoctorBase) | Unneeded sign-up fields removed | +30% new doctor sign-ups | Kareo Marketing |
| 14 | Travel company | 3 required fields, optional ones hidden | +26% conversions | InsightWhale |
| 15 | EDP | Contact fields asked later in funnel | More than 25% more leads | EDP |
| 16 | Trex | Form framed as calculator's next step | +23% lead form conversions | Seer Interactive |
| 17 | Qualicorp | Repeated fields removed | +19.31% completed quote requests | VWO |
| 18 | Vandebron | Typed date of birth, auto-formatted | +16.3% sign-ups | Vandebron |
| 19 | Tinkoff Bank | Progress bar with approval likelihood | +12.8% submissions | Tinkoff Bank |
| 20 | BliVakker.no | 3 of 17 fields removed | +11% form submissions | BliVakker.no |
The 20 best lead form A/B tests
1. HomEquity Bank: one-page form split into seven steps
The change: HomEquity Bank took its one-page reverse mortgage inquiry form and broke it into a seven-step form.
The result: +154% conversion rate on the inquiry form.
Run by HomEquity Bank and reported in a MarketingSherpa case study as an A/B test.
The audience was Canadians aged 55+, and the test design was not detailed.
Why it likely worked: One question at a time looks easier than a wall of fields, even when the total number of questions is the same.
People commit to the first easy step and then want to finish.
Try it if: Your form has more than five or six fields and your audience is not in a hurry, such as finance, insurance or home services.
2. Reclaim My PPI Tax: long claim form split into four stages
ApexureThe change: A UK tax rebate service replaced a single long claim form with a four-stage form.
Easier questions came first, and some fields sat side by side.
The result: Conversion rate went from 1.77% to 4.33%.
Run by Apexure as an A/B test on paid traffic in Unbounce over about a month.
Several form changes were bundled in one variant, and no traffic counts were given.
Why it likely worked: Starting with easy questions builds momentum before the harder ones about tax and personal details.
Splitting the form also hid its full length.
Try it if: Your form asks for sensitive or effortful details that people tend to abandon halfway.
3. Britannica: one page cut from the registration path
The change: Encyclopaedia Britannica Online removed one whole page from its subscription registration path. Nothing else about the offer changed.
The result: +117% free trial conversion, +61% paid subscriptions and +130% revenue. Run by MarketingExperiments as an A/B test.
Why it likely worked: Every extra page is another chance to leave.
This one carried through to paid subscriptions and revenue, not just sign-ups, which is rare.
Try it if: Your sign-up flow has a page that exists for the business (account preferences, extra profile data) rather than for the user.
4. SafeSoft Solutions: the price shown next to the lead form
The change: SafeSoft's PPC landing page had a contact form and no price.
The variant added a large green image showing the product price next to the form.
The result: +100% leads. Run by SafeSoft Solutions and published as a VWO case study of an A/B test. No sample size was given.
Why it likely worked: People hesitate to hand over contact details when they fear a sales call about a price they can't afford.
Showing the price answers that question before the form.
Try it if: You sell a product with a fixed price but hide it behind a "contact us" form.
5. VenueSphere: a blunt subheadline above the enquiry form
The change: VenueSphere replaced a polite subheadline (call us or fill in the form) with an attention-grabbing "Stop right now!" line promising they would do the venue hunting for free.
The result: +69% form submissions. Run by VenueSphere as an A/B test, published by VWO.
The test had close to 1,200 visitors, so treat it as a strong hint rather than proof.
Why it likely worked: The new line told visitors what they would get (free help) instead of what they had to do (fill in a form).
Try it if: The copy around your form describes the form instead of the benefit of sending it.
6. Vendio: sign-up form moved off the landing page
The change: Vendio's free-store landing page had registration fields embedded next to the marketing copy.
The variant removed the fields and used a "Signup Now" button that led to a separate form page.
The result: +60% sign-ups. Run by Vendio as an A/B test, published by VWO.
Styling also differed slightly and no sample size was given.
Why it likely worked: An embedded form can make the page feel like a demand before the visitor has read the pitch. A button is a smaller first commitment.
Try it if: Your landing page opens with a long form.
Note that Tom's Planner (entry 9) found the opposite with a short form, so test it rather than assume.
7. Casa Mineira: free-text search replaced by two drop-downs
The change: A Brazilian real estate site swapped a single free-text property search box for two drop-down menus: apartment type and location.
The result: +57.25% form submissions. Run by Supersonic as a split URL test with about 7,500 visitors.
Why it likely worked: A blank text box makes people think about what to type.
Drop-downs show the options and remove spelling errors and empty results.
Try it if: Your form starts with an open text field that has a small set of likely answers.
8. Hotel Institute Montreux: student testimonial above the brochure form
The change: A Swiss hospitality school placed a student testimonial directly above its free brochure request form.
The result: +50% form submissions, and AdWords cost per lead fell 54%.
Run by The Reference as a split URL test.
Why it likely worked: Proof from someone like the reader, placed at the moment of decision, makes the request feel safe and worth it.
Try it if: You have social proof somewhere on the site but not next to the form.
9. Tom's Planner: sign-up form above the fold
The change: Tom's Planner replaced two above-the-fold buttons (sign up and watch a demo) with a short four-field sign-up form and an arrow pointing to it.
The result: +44% sign-ups. Run by Tom's Planner as an A/B test with close to 3,000 visitors, published by VWO.
Why it likely worked: The form removed a click and a choice.
With only four fields, it looked quick enough to fill in on the spot.
Try it if: Your product is free to try and your form is short.
Compare with Vendio (entry 6), where a long embedded form lost.
10. Tinkoff Bank: a "Don't remember passport details?" option
The change: Tinkoff's credit card application asked for passport details at step two.
The variant added a checkbox that let applicants supply them later by phone or email.
The result: +35.8% conversion rate on application submissions.
Run by Tinkoff Bank as an A/B test with 265,000+ visitors over 23 days, published by VWO.
Why it likely worked: The form stopped people who didn't have a document at hand.
Letting them skip a field and finish later kept them in the funnel.
Try it if: One required field needs information people often don't have on hand (ID numbers, account numbers, policy numbers).
11. HomEquity Bank: progress tracker on the multi-step form
The change: After moving to a multi-step form, HomEquity Bank added a progress tracker to it.
The result: +35% conversions. Run by HomEquity Bank as an A/B test, reported by MarketingSherpa.
Why it likely worked: A multi-step form hides its length, which can worry people.
A tracker shows how close they are to the end.
Try it if: You already run a multi-step form and people drop off in the middle steps.
12. Reassured: one question per step on a life insurance quote form
The change: Reassured replaced a long single-page quote form with a multi-step form that asked one question per step and showed a progress bar at the top.
The result: +31.23% form submissions over about five weeks.
Run by Reassured as an A/B test, published by VWO. The lift was higher on desktop.
Why it likely worked: This combines the two ideas above (steps and a progress bar) on a high-intent insurance form.
Try it if: You want to try the multi-step pattern and need a second example from a regulated, lead-heavy industry.
13. Kareo (DoctorBase): fewer fields on the doctor sign-up form
The change: Kareo's marketing team removed several unnecessary fields from the sign-up form doctors used for its patient communication software.
The result: +30% new doctor sign-ups. Run by Kareo Marketing as "a few" A/B tests, published by Unbounce. The fields removed were not listed.
Why it likely worked: Busy professionals don't fill in fields that look like they are for the vendor's benefit.
Try it if: Your B2B sign-up form asks for things sales could collect on the first call.
14. Travel company: three required fields, the rest hidden
The change: A travel company's deal page form had eight mandatory fields in a multi-column layout.
The winner kept three required fields and hid the five optional ones under a dropdown.
The result: +26% conversions. Run by InsightWhale as an A/B test, published by Convert.
It also beat a version that showed the optional fields below the required ones.
Why it likely worked: The form looked short, but people who wanted to add detail still could.
Hiding optional fields beat showing them as optional.
Try it if: Sales wants the extra data but the form is losing people.
Make the extras optional and tuck them away.
15. EDP: form fields spread through the funnel
The change: Energy supplier EDP's lead funnel asked for name, phone and email early, before users had context.
The variant spread the required fields through the funnel, asking for the key ones near the offer stage.
The result: More than 25% higher lead conversion rate.
Run by EDP as an A/B test, published by VWO.
The result was given only as a lower bound.
Why it likely worked: Asking for a phone number before showing the offer feels like a trap.
Asking after the visitor has seen value feels like the next step.
Try it if: Your quote or calculator flow asks for contact details on the first screen.
16. Trex: lead form as the next step of the cost calculator
Seer InteractiveThe change: Trex's decking cost calculator had a lead form that felt like a separate, optional add-on.
The variant presented the form as the natural next step of the calculator.
The result: +23% lead form conversions. Run by Seer Interactive as the first of three A/B tests. No sample size was given.
Why it likely worked: When the form continues something the visitor is already doing, filling it in feels like finishing the task rather than starting a new one.
Try it if: You have a calculator, quiz or configurator and the lead form sits beside it rather than after it.
17. Qualicorp: repeated fields removed from step two
The change: Health plan seller Qualicorp asked for name and date of birth on the homepage and then again in step two of the quote form. The variant stopped asking twice.
The result: +19.31% completed quote requests (and +24.68% Request Contact clicks).
Run by VWO as a 14-day, 50/50 A/B test.
Why it likely worked: Being asked the same thing twice signals a broken system and wastes effort.
Try it if: Your form is split across pages or steps.
Check it for duplicate questions; it's a quick fix.
18. Vandebron: a simpler date-of-birth field
The change: Vandebron's energy sign-up used a calendar dropdown for date of birth that defaulted to 2000 and stepped month by month.
The variant removed the calendar, showed the expected format as a placeholder and inserted slashes automatically.
The result: +16.3% sign-ups, -8.5% exit rate on the step and -80% drop-offs on the date field.
Run by Vandebron as an A/B test, published by VWO.
Why it likely worked: One badly built field can sink a whole form.
Typing a birthday is faster than scrolling a calendar back decades.
Try it if: Your analytics show people stalling on one field. Fix that field before redesigning the form.
19. Tinkoff Bank: progress bar with approval likelihood
The change: Tinkoff tested two progress-bar designs above its four-step credit card application.
Both showed the current step and an approval likelihood that rose as users filled in the form.
The result: +12.8% submissions for the green filling bar and +6.9% for the yellow bar.
Run by Tinkoff Bank as an A/B test with 190,000+ visitors over 39 days, published by VWO.
Why it likely worked: The bar rewarded each answer with a visible gain.
The design mattered: one version did nearly twice as well as the other.
Try it if: Your form is multi-step and the reward for finishing is uncertain (approval, a quote, a match).
20. BliVakker.no: three fields cut from a 17-field form
The change: Norwegian cosmetics retailer BliVakker.no removed account number, phone and evening phone from the 17-field registration step in its purchase flow.
The result: +11% form submissions for this light version.
Run by BliVakker.no as an A/B test, published by VWO.
An ultra-light version with even fewer fields was also tested, and the source says stripping too much was not optimal.
Why it likely worked: Cutting fields that clearly weren't needed helped.
Cutting further didn't, which suggests some fields make the form feel legitimate.
Try it if: You run an ecommerce or account-creation form with fields nobody can explain the reason for.
Patterns across the winners
Steps beat one long page, when the questions are easy
Three of the top 12 split a long form into steps: HomEquity Bank, Reclaim My PPI Tax and Reassured.
All had long forms with a clear payoff.
Britannica shows the other side: steps help only when each one is short, and a page that adds nothing should go.
See multi-step forms for every test we have on the pattern, and multi-step form builders if your current tool can't do it.
Cut what the visitor can't see a reason for
Kareo, BliVakker.no, Qualicorp and the InsightWhale travel client all removed or hid fields that served the business, not the user.
BliVakker.no's source also says going further did not help, so the goal is fewer pointless fields, not the fewest possible.
Our form fields page has the full set.
Answer the fear next to the form
SafeSoft showed the price, Hotel Institute Montreux showed a student testimonial and VenueSphere promised free help.
Each answered "what happens if I send this?" in the spot where people decide.
More layouts like these are in our lead gen form examples.
Ask for contact details after you've shown value
EDP and Trex both moved the form later, so it felt like the next step of a quote or calculator.
Tinkoff let people skip a hard field and finish later.
Same idea: don't block people before they want what you offer.
For SaaS sign-ups specifically
Britannica, Tom's Planner, Vendio and Kareo are the SaaS-style tests here.
Their lesson is removing pages and fields from the path to the trial.
Browse more in sign-up page examples and the sign-up page guide.
Form tests that lost or went flat
- RegOnline's three-step free trial sign-up lost: -24.5% completed leads, run by MarketingExperiments. More steps are not automatically better.
- A madlibs-style sign-up form lost to a standard form: -22% trial sign-ups, run by Patrick McKenzie over 34 hours with 736 participants.
- Removing 2 sign-up form fields was flat: no lift, trended slightly worse, run by GrowthRock.
- A shorter sign-up form lost where it counted: +16% free sign-ups, but paid fell 0.97% to 0.72%, run by MarketingExperiments.
- A guide choice plus one extra form field lost: a significant drop in leads, run by MarketingSherpa.
More of these are in best practices that failed.
Real results
MarketingExperimentsMarketingExperimentsWhat to test next
- Audit for duplicates and dead fields. Ask who uses each field. Remove repeats (Qualicorp) and hide optional ones (InsightWhale travel client).
- Fix your worst field. Find the field with the most drop-offs in form analytics and rebuild it (Vandebron).
- Put proof or price beside the form. A testimonial, a price or a clear promise of what happens next (Hotel Institute Montreux, SafeSoft).
- Try steps on long forms. If you have more than five or six fields, test a multi-step version with easy questions first (HomEquity Bank, Reclaim My PPI Tax).
- Add progress feedback. Once you have steps, test a progress bar, and test more than one design (Tinkoff Bank).
- Measure past the form. Track qualified leads, trials that convert or revenue, not just submissions. A form that gets more junk leads is not a win.
Watch out for
Lead quality. Fewer fields can bring in more leads that sales can't use.
The paid content site above got more free sign-ups and fewer paid ones.
Agree on a quality metric before the test starts.
Small samples. VenueSphere's +69% came from close to 1,200 visitors, and several vendor case studies give no sample size at all.
Read statistical significance before you call a winner on low-traffic forms.
Bundled changes. Reclaim My PPI Tax changed several things in one variant.
You know the package worked, not which part did.
If your lead form matters to revenue and you want someone to plan and run these tests with you, Oliver Kenyon works with lead-gen and SaaS teams on exactly this.