A/B testing forms: the guide for beginners

Your contact form converts at 3%. Not bad. But could it do better? 4%? 6%? The only way to know: test. A/B testing of forms allows you to compare two versions of the same element to identify which one performs better. A simple, scientific, and remarkably effective method.

In this guide, you’ll discover how to set up your first A/B tests on your forms. No technical jargon. No complex setup. Just solid foundations to start optimizing today.

What is A/B testing for forms?

A/B testing (or split testing) consists of presenting two different versions of a form to distinct groups of visitors. Version A stays the same (control). Version B includes a modification (variant).

The principle is simple:

50% of visitors see version A
50% see version B
You measure which version generates the most submissions

After collecting enough data, you keep the winning version. Then you test something else. And so on.

Why test rather than guess?

Your intuition deceives you. Often.

That green button you find more attractive? Maybe your visitors prefer blue. That “phone” field you consider essential? It might be driving away 20% of your prospects.

Data doesn’t lie. A well-conducted A/B test eliminates opinion debates and assumptions. You know what works because you’ve measured it.

Some telling figures:

A button color change can increase conversions by 21%
Removing a single field can improve completion rate by 26%
Modifying a CTA text can double clicks

Elements to test first on your forms

Not all elements have the same impact. Here’s where to start for quick results.

Submit button text

This is often the simplest and most profitable test. “Submit” is generic. It communicates no value.

Test these alternatives:

“Get my quote” vs “Request a quote”
“Start for free” vs “Create my account”
“Get my consultation” vs “Book my slot”

First-person buttons (“My”) generally outperform second-person ones. But don’t take my word for it. Test.

Number of fields

Basic rule: each additional field reduces your conversion rate by approximately 4%.

Tests to run:

Complete form vs minimal form (email only)
With or without phone field
First name + Last name vs Full name in a single field

Ask yourself: do you really need this information now? Or can you collect it later?

Form title

The title guides the action. It can reassure or create friction.

Test examples:

“Contact us” vs “Ask your question”
“Quote request” vs “Get your quote within 24 hours”
With response time vs without mentioned delay

A benefit-oriented title generally converts better than a descriptive title.

Position on the page

Your form’s placement directly influences its visibility and completion rate.

Variations to test:

Above the fold vs at the bottom of the page
Fixed sidebar vs integrated into content
Scroll pop-up vs static form

On mobile, consider a sticky button that stays visible during scrolling.

Trust elements

Visitors hesitate. They wonder if their data will be protected, if they’ll receive spam, if someone will actually respond.

Elements to test:

With or without visible GDPR mention
With or without customer testimonials
With or without guaranteed response time

With Skedox, you can easily create multiple versions of your forms and track their respective performance. Ideal for launching your first tests without technical complexity.

Methodology: how to conduct an effective A/B test

A poorly conducted test produces false results. Here’s the 5-step method for reliable tests.

Step 1: Define a clear hypothesis

Don’t test randomly. Formulate a precise hypothesis.

Bad approach: “I’m going to test the button.”

Good approach: “By replacing ‘Submit’ with ‘Get my free guide’, I will increase the submission rate by 15% because the benefit will be more explicit.”

A well-formulated hypothesis includes:

What you’re changing
The expected result
The supposed reason

Step 2: Test a single variable

This is the golden rule. If you change the button text AND its color AND its size, you won’t know what made the difference.

One test = one variable.

Want to test multiple elements? Run multiple successive tests. Or use multivariate tests (but that’s another story).

Step 3: Calculate the required sample size

To get statistically significant results, you need a minimum volume of data.

Practical rule:

Minimum 100 conversions per variant
Ideally 250+ for more reliability
Minimum duration: 7 days (to cover weekly variations)

With 1,000 visitors per month and a 5% conversion rate, you get 50 monthly conversions. You’ll therefore need about 4 months for a reliable test. Or more traffic.

Step 4: Let the test run long enough

Don’t check results every day. Don’t conclude after 48 hours because version B “seems” better.

Wait for:

Statistical significance (95% confidence minimum)
At least one complete cycle (week or month depending on your activity)
A sufficient volume of conversions

Early results are often misleading. Patience pays off.

Step 5: Analyze and document

Once the test is complete, document everything:

Initial hypothesis
Test duration
Traffic volume per variant
Conversion rate per variant
Result: winner, loser, or inconclusive
Decisions made

This documentation will prevent you from redoing the same tests and will build your knowledge base.

Classic mistakes to avoid

Concluding too quickly

You have 30 conversions and version B leads by 20%? That’s not significant. Random fluctuations can create significant gaps on small samples.

The solution: use a statistical significance calculator. Many free tools exist online.

Testing during atypical periods

A test launched on December 24th will produce skewed results. Holidays, public holidays, major events affect visitor behavior.

The solution: test during periods representative of your normal activity.

Ignoring segments

Your version B wins overall. But did it win on mobile AND desktop? In France AND in Belgium? In the morning AND in the evening?

The solution: analyze results by segment. A version can perform differently depending on context.

Giving up after an inconclusive test

No clear winner? That’s not a failure. It’s valuable information: this element probably doesn’t have a major impact on your conversions. Move on.

Tools to get started with A/B testing forms

Several options are available depending on your technical level and budget.

Solutions integrated into form platforms

The simplest solution. Skedox allows you to create and compare multiple versions of your forms with integrated analytics. You directly visualize conversion rates per version without complex configuration.

Dedicated A/B testing tools

Google Optimize (free), VWO, Optimizely, or AB Tasty allow testing any element on your pages. More powerful, but more complex to set up.

Homemade solution

With some code, you can create your own testing system. But beware of statistical biases and maintenance.

To get started, prefer an integrated solution. You’ll save time and avoid configuration errors.

A/B testing: concrete examples and results

Example 1: Reducing the number of fields

Context: A B2B quote request form with 8 fields.

Test: Version A (8 fields) vs Version B (4 fields: email, name, company, message).

Result: +34% submissions for version B. Lead quality did not decrease significantly.

Example 2: Button text

Context: A newsletter signup form with the “Subscribe” button.

Test: “Subscribe” vs “Get my free tips”.

Result: +52% signups for the second version. The explicit benefit made the difference.

Example 3: Adding a trust element

Context: A contact form without response time indication.

Test: Without mention vs “Response guaranteed within 24 hours”.

Result: +18% submissions. The promise of a quick response reduced hesitation.

Create your first test right now

You have the basics. Time for action.

Your plan for this week:

Choose a form to optimize
Identify the element to test first (start with the button)
Formulate your hypothesis
Create your variant
Launch the test and wait for results

Don’t seek perfection. The first test is rarely perfect. The important thing is to start, learn, and iterate.

Conclusion: A/B testing forms, a worthwhile investment

A/B testing forms isn’t reserved for large companies with data teams. With the right methodology and tools, any SMB can optimize its forms scientifically.

Start small. Test one element. Analyze the results. Start again. In a few months, you’ll have considerably improved your conversion rates.

Ready to launch your first tests? Discover Skedox and create optimized forms with integrated analytics. Track your performance, identify improvement opportunities, and make data-driven decisions.

Your visitors give you valuable information with every interaction. It’s up to you to listen to them.

A/B testing forms: complete guide for beginners