A lead form is the set of fields a visitor fills in to become a lead: a quote request, demo booking, brochure download or contact form.
The forms that convert best ask only for what you need to follow up, explain why each answer matters, and reassure people at the moment they hesitate.
In our database, removing a field, splitting a long form into steps or adding a trust line beside the submit button has lifted completions anywhere from 5% to several hundred percent, and adding the wrong field has cut leads.
This guide goes through the form tests in the LiftArchive lead-gen collection and pulls out what held up, what failed and how to test your own form.
What a lead capture form has to do
A lead capture form has two jobs that pull against each other.
Marketing wants more submissions. Sales wants enough detail to qualify and route each lead.
Every field you add helps the second job and usually hurts the first.
So the first decision isn't design. It's which fields you actually use.
If no one reads the "company size" answer before calling the lead, that field is pure friction.
The test results back this up. When Qualicorp removed fields that asked for the same information twice, completed quote requests rose 19.31% (run with VWO).
When BliVakker.no cut 3 fields from a 17-field registration form, sign-ups rose 11%.
Going the other way, MarketingSherpa found that adding one extra field to a guide download form caused a significant drop in leads.
Fewer fields is not a law, though. In the same Qualicorp program, making the profession field required lifted completed requests 12.52%.
When a field helps people get a relevant answer (here, the right health plan), it can read as useful rather than nosy.
Form design best practices, backed by tests
Here are the patterns that show up again and again in the form field tests we track.
Each one comes from a real test, but most are single results from one site, so read them as strong hypotheses, not rules.
| Practice | Test | Result | Ran by |
|---|---|---|---|
| Cut duplicate or unused fields | Removed repeated fields | +19.31% completed quote requests | VWO |
| Let people skip hard questions | 'Don't remember passport details?' option | +35.8% applications | Tinkoff Bank |
| Simplify awkward inputs | Simpler date-of-birth field | +16.3% sign-ups | Vandebron |
| Replace long dropdowns with search | Search field instead of a 450-gym dropdown | +3.4% conversion | PureGym |
| Mark required vs optional fields | Asterisks on required fields | +5.89% form completions | Micro Focus |
| Keep labels visible | Floating labels on a quote form | +16.93% completed requests | VWO |
| Reassure about data use | Privacy message on the quote form | +10% quotes | Groupama |
| Add proof next to the form | Student testimonial above the brochure form | +50% submissions | The Reference |
Two themes run through this table. The first is lowering effort: fewer fields, easier inputs, an escape route for questions people can't answer right away.
The second is lowering doubt: telling people what happens to their data and showing that others like them went ahead.
Trust content works best right next to the submit button, where the hesitation happens.
POSist made its demo form narrower and placed logos and testimonials beside it, and demo sign-ups rose 20%.
For more on what counts as proof, see our page on trust signals.
The headline above the form matters too. Kaya Skin Clinic changed its form heading and CTA to offer an "expert opinion", and conversion went from 4% to 9.5% (run by Ultraviolet Digital).
People fill in forms for what they get, so name the outcome, not the action.
Form optimization: single-step, multi-step and where the form sits
Form optimization goes beyond field count. Two structural choices often move results more than any single field: whether the form is split into steps, and where it sits on the page.
Multi-step forms
Splitting a long form into short steps, starting with easy questions, is one of the strongest patterns in our multi-step form tests.
Venture Harbour moved a consulting enquiry form to a multi-step layout and conversion went from 0.96% to 8.1%.
Reassured's multi-step life insurance quote form with a progress bar lifted submissions 31.23%.
Tinkoff Bank added a progress bar to a credit card application and submissions rose 12.8%.
Fewer steps can also win. Performics India cut a paint cost calculator from 5 steps to 3 and leads rose 40%.
And multi-step forms can lose: MarketingExperiments tested a three-step free trial sign-up on RegOnline's homepage and completed leads fell 24.5%.
The lesson is that steps help when each one is short and the first one is easy.
They hurt when they add clicks without making any question easier.
Splitting the form also takes some load off working memory, which our page on cognitive load covers in more depth.
If you need a tool to build one, compare the options on our multi-step form builders page.
Form placement
Moving the form up the page often helps.
HDFC ERGO put the full form above the fold on a health insurance page and leads rose 25%.
Warner Music Group moved the sign-up form on artist sites and sign-ups went from 6.2% to 18.4%.
But hiding the form can win as well.
Northmill tucked a loan application behind an "Apply" button and applications rose 7%.
Compare Courses moved its enquiry form above the fold and other leads rose 159.52% while form leads fell.
Watch every conversion type, not just the one you changed.
Web form best practices for usability
Some web form best practices come from usability research, not from A/B tests.
They are worth following even if you never test them, because they prevent errors and support accessibility.
- Put labels outside the field and keep them visible. Nielsen Norman Group's Katie Sherwin lists the problems with using placeholder text as labels: people forget what a field asked for, can't check their answers before submitting, and may mistake a placeholder for a pre-filled value. The article treats floating labels as a compromise and always-visible labels outside the field as the clearest option. Qualicorp's floating-label win above came against a weaker starting design, so it doesn't contradict that advice.
- Count your fields before you count your steps. Baymard Institute's checkout research found the average checkout in 2024 had 5.1 steps and 11.3 form fields, while most sites need only 8. Baymard concluded that the number of fields affects usability far more than the number of steps. That research is about ecommerce checkouts, but the same logic applies to lead forms.
- Use the right input type. Dropdowns are slow when the list is long, and free-text fields are slow when the answer has a fixed format. Supersonic replaced a free-text property search with two dropdowns for Casa Mineira and leads rose 57.25%, while PureGym went the other way, from a dropdown to search. Pick whatever needs the fewest keystrokes for that answer.
- Label optional fields, or remove them. Springest added "optional" labels and social proof to a brochure form and saw a 5.21% lift.
Clever formats don't always help. Patrick McKenzie tested a madlibs-style sign-up form, where fields sit inside a sentence, and it cut trial sign-ups 22%. Plain, familiar layouts are a safe default.
Contact form best practices
A contact form is usually the lowest-intent lead form on a site, so the bar for effort should be the lowest too.
Name, email and a message box is enough for most businesses. Anything else can wait for the reply.
Visibility often matters more than design. Cosentino added a "Contact" link to its main menu and leads from the contact page rose 140%.
On mobile, Digital NRG moved Trent's Drains' contact form higher and added a call-us line, which brought in more form leads (no exact figure was shared).
Give people another option when the form isn't right for them.
EZ Texting put a live chat widget next to its sign-up form and sign-ups rose 31%.
For people who start a long form and leave, IMB Bank's exit pop-up offering to save progress increased saved loan applications 52%.
Finally, don't judge a contact form on submissions alone.
A form that brings in twice as many leads but half as many qualified ones hasn't helped.
Track lead quality or sales outcomes for each variant where you can.
How to test your own form
- Audit every field. For each one, ask who uses the answer and when. Remove or defer anything nobody uses before the first call.
- Find where people drop off. Field-level analytics show which question stalls people. That field is your first test.
- Test one structural change first. Field count, steps or placement usually moves more than button color or wording.
- Measure downstream. Count qualified leads, not just submissions, as EDP did when it spread form fields through the funnel and got more than 25% more leads.
- Run it long enough. Lead forms often have low traffic, so plan for weeks, not days, and decide your sample size before you start.
Browse more examples in our lead form test collection, the lead-gen A/B test results roundup and the full test database.
If the form sits on a landing page, our landing page best practices guide covers the rest of the page.
Venture Harbour
MarketingExperiments