Good sign up page design asks for as little as possible, shows proof that the product is worth the effort, and removes every step that isn't needed to create the account.
In our database, cutting a whole registration page lifted Britannica's free trials by 117%, and Teamleader gained conversions from small additions like "No credit card required" and customer logos.
Some popular ideas have also lost, so test changes before you treat them as rules.
This page pulls together signup tests from our SaaS and lead-gen collections.
Each result is linked to its full write-up and credited to the team that ran it.
You can browse more in the full test database.
What the signup form itself should look like
The form is where most of the friction lives.
Every field is a question the visitor has to answer before seeing any value, so start by removing the fields you don't use on day one.
Kareo Marketing removed several unnecessary fields from the DoctorBase doctor sign-up form and got 30% more new doctor sign-ups.
Kareo didn't publish which fields went, but the lesson travels: if sales or onboarding can collect it later, leave it off the signup form.
Some fields can't be removed, but they can be made easier.
Vandebron's energy sign-up had a date-of-birth calendar that defaulted to 2000 and moved one month at a time.
Swapping it for a plain text field with a format hint and auto-inserted slashes gave +16.3% sign-ups and 80% fewer drop-offs on that field.
PureGym replaced a dropdown of 450+ gym locations with a search field and saw +3.4% conversion.
Small visual cues on the fields have also tested well:
- Teamleader added icons next to each field and moved free trial sign-ups from 32.84% to 37.11% (+13.1%).
- Micro Focus marked required fields with asterisks and a "required field" note, and form completions rose 5.89% over a 20-day, 50/50 test.
Not every clever form wins. Patrick McKenzie tested a "madlibs" form, where the fields sit inside a sentence, against a standard labeled form on Bingo Card Creator.
The standard form converted at 27.55% and the madlibs version at 21.73%, about 22% worse, across 736 participants.
The test only ran 34 hours, but it is a useful warning about copying design trends.
Labels matter too. Nielsen Norman Group recommends clear, visible labels outside the field rather than placeholder text, because placeholders vanish as people type and are hard to read for users with visual or cognitive impairments.
For more form-level evidence, see our form fields tests.
How a SaaS signup flow should be structured
A SaaS signup flow is every step between "Start free trial" and the first screen inside the product: the button, the form, email verification, any plan choice, and the first-run setup.
The number of steps often matters more than the design of any one of them.
The clearest example is Encyclopaedia Britannica Online. MarketingExperiments removed one whole page from the registration path, and the result was +117% free trial conversion, +61% paid subscriptions and +130% revenue.
Patrick McKenzie also reworked the Appointment Reminder signup process and reported about 50% more free trials, though the post doesn't say what changed.
Fewer steps isn't always the answer, though. In lead-gen, splitting long forms into steps often wins: HomEquity Bank turned a one-page inquiry form into seven steps and got +154% conversion rate.
The difference is what each step asks. Removing a page that only adds effort helps, while splitting the same questions into easy steps can help too.
Our multi-step form tests cover the second case in detail.
The offer behind the button shapes the whole flow.
Acuity Scheduling changed its main signup from a free-forever plan to a 14-day Professional trial with card details up front and saw +268.14% paid signups.
That changed the offer itself, not just the page, so treat it as a pricing test.
If you're weighing card-required trials, start with our free trial guide and freemium vs free trial comparison.
The flow also doesn't end at the form.
What happens on the first screen after signup decides whether a trial user turns into a customer, which our SaaS onboarding page covers.
Signup flow best practices backed by tests
These are the patterns with the strongest support in our database. Each links to the test behind it.
| Practice | Test | Result |
|---|---|---|
| Remove a registration step | Britannica (MarketingExperiments) | +117% free trials, +130% revenue |
| Say "No credit card required" | Teamleader | NL 48.7% to 54.8%; BE 33% to 38.6% |
| Add testimonials next to the form | Mida | 37% to almost 60% completion (+56.94%) |
| Add customer logos | Teamleader | 36% to 40.3% (about +12%) |
| Offer live chat beside the form | EZ Texting | +31% sign-ups |
| Cut competing calls to action | B2C online service (MarketingExperiments) | +265.6% clickthrough, about +27% conversion |
Answer the objection before it's asked
Many visitors hesitate because they expect a card form.
Teamleader's "No credit card required" line lifted sign-ups on both its Dutch and Belgian sites, a 12.5% headline lift.
A Hacker Safe security seal on Britannica's trial offer page added 13.2% to free trial take-up.
Put proof next to the form
Mida's testimonial test is striking, but it was based on about 200 new sign-ups, and Mida was testing its own product, so treat the size of the lift with care.
Teamleader's logo result was smaller and more typical.
Our social proof and trust logos pages collect more of these tests.
Give the page one job
A signup page with a login button, navigation and side offers gives visitors several ways out.
When Planoly removed the login button and added a testimonial and badges, CodaLift LLC measured +17% paid clients.
Micro Focus removed a "Contact Us" button and moved the trial button above the fold, and completions rose 12.37%.
Both were bundled changes, so we can't say which part did the work.
Don't assume polish wins
MarketingExperiments tested a bare white page with a plain name-and-email form against two polished versions with logos, testimonials and badges.
The plain control converted at about 14% against 7-9% for the treatments.
Proof helps when it answers a real doubt, and adding clutter around the form can still hurt.
Mida
MarketingExperimentsSignup form optimization: what to test and how to read it
Once the basics are in place, these are the technical details worth getting right.
They come from public standards, not our tests, so they're things to fix rather than to A/B test:
- Use the right autocomplete and input types. Google's web.dev guide recommends values like autocomplete="new-password", email and tel, plus the correct type so mobile users get the right keyboard (web.dev sign-up form best practices).
- Don't ask for email or password twice. The same guide suggests sending a confirmation code instead of making users retype contact details.
- Drop password composition rules. NIST's SP 800-63B (July 2025) says services "SHALL NOT impose other composition rules (e.g., requiring mixtures of different character types)" and must allow password managers and autofill (NIST SP 800-63B).
- Offer sign-in with an existing account. web.dev also recommends letting people sign up with a third-party identity provider, which saves typing a password at all.
When you test, measure the right thing. Kommunicate removed the email field from its homepage "Try for Free" button and got +25.53% clicks, but clicks are not completed signups.
A MarketingExperiments test on a subscription site went the other way: adding an email field to the trial call to action raised trial signups from 0.68% to 1.57% (+133%), with 32,782 and 30,479 visitors in the two arms.
Track signups, activation and paid conversion, not just the first click.
Button wording is another cheap test. On its Partner Hub join page, Booking.com compared "Join now", "Start registration", "List your property" and "Sign up today".
"Start registration" did best, while "Sign up today" cut clicks by about 10%.
Small copy changes on the final button can move results in either direction, so test them rather than picking the wording that sounds best.
Some tests will come back flat, and that's useful too.
Patrick McKenzie removed the mailing-list checkboxes from the Bingo Card Creator signup form and found no significant change in trials.
A flat result tells you that element wasn't what held people back, so you can move on to the next idea.
For a wider look at form tests outside SaaS, see our form optimization page.
For the page around the form, including headline and layout tests, see our SaaS signup page guide.