Feature

Form A/B testing: discover scientifically what converts

Fullmetrix lets you A/B test multiple variants of your forms and popups to identify the version that maximizes your conversion rate. Design, copy, field order, button wording variants: test everything, measure in real time and automatically deploy the winning version.

Try for free

The problem

Why most merchants don't A/B test their forms

A/B testing tools are complex and expensive

Optimizely, VWO or Google Optimize (discontinued) require installation, configuration and often a hefty subscription. To test a simple button change on a newsletter form, that's a disproportionate effort. Result: 90 percent of merchants never test their forms and leave tens of percentage points of conversion on the table.

Results are hard to interpret

Without statistical significance, you risk concluding a variant is winning when the difference is actually due to chance. Classic tools require understanding p-values, confidence intervals and minimum sample sizes. Without these statistical foundations, decisions are made on noisy and therefore false results.

Impossible to test multiple forms simultaneously

An average merchant has 5 to 10 active forms on their site (newsletter, contact, exit popup, quiz, lead magnet). Testing each form individually by hand would take months. Without a centralized tool, most forms are never optimized and stay frozen for years.

The solution

How Fullmetrix democratizes form A/B testing

Variant creation in one click

From any Fullmetrix form, duplicate it in one click to create a variant. Modify what you want to test (title, button, field order, design) on the variant. Fullmetrix automatically creates an A/B test with 50/50 traffic split between the original version and the variant.

Statistical significance calculated automatically

Fullmetrix calculates in real time the statistical significance of each test (p-value, confidence interval, sample size). The dashboard clearly shows when a variant can be declared winner with 95 percent confidence, or if results aren't yet significant and you need more traffic.

Automatic deployment of the winning version

When a variant reaches statistical significance and beats the original version, Fullmetrix can automatically deploy the winning version to 100 percent of traffic. The next test can start immediately on another element to optimize. Optimization becomes a continuous and automated process.

How it works

Launch your first A/B test in 4 steps

Choose a form to test

Identify a form with enough traffic volume (at least 500 impressions per week recommended). High-traffic forms reach statistical significance faster and allow quick iterations.

Create a variant

Duplicate the form and change a single element at a time for a clean test: title, button, field order, image, main color, wording. Avoid changing multiple elements simultaneously so you can attribute the gain to a specific cause.

Launch the test

Start the test. Fullmetrix automatically splits traffic 50/50 between the original version and the variant. Each visitor always sees the same version (consistency) thanks to an identifier stored in local storage. The test runs until statistical significance is reached.

Analyze and deploy

Monitor results in real time in the dashboard. When a variant reaches 95 percent statistical confidence, Fullmetrix flags the winner. Deploy the winning version in one click and launch the next test. Optimization becomes iterative and measurable.

Key features

Everything you need to A/B test effectively

One-click variant creation

Duplicate a form and modify the elements to test. Fullmetrix automatically manages traffic split and tracking of each version.

Statistical significance

Automatic calculation of p-value, confidence interval and minimum sample size. No statistics knowledge required, results are interpreted for you.

Multivariate tests

Test up to 4 variants in parallel to speed up iterations. Fullmetrix automatically adjusts statistical significance to account for multiple comparisons.

Automatic winner deployment

Configure Fullmetrix to automatically deploy the winning version to 100 percent of traffic as soon as it reaches significance. Continuous optimization, zero manual intervention.

Test history

Keep full history of all past tests with results, dates, variants. Learn from each test to build your own optimization playbook.

Results segmentation

Analyze results by segment: desktop vs mobile, new vs recurring, organic vs paid. A variant can win on mobile and lose on desktop, Fullmetrix reveals it.

Integrations

Compatible with your Fullmetrix forms

Fullmetrix forms

Native A/B tests for all forms created in the Fullmetrix editor, without extra configuration.

Popups and embeds

A/B tests on exit popups, scroll popups, embed forms in content, load popups.

PrestaShop, WooCommerce, Shopify

The same Fullmetrix snippet handles tracking and A/B testing on all supported platforms.

Meta, Google Ads

Segment results by ad campaign to identify variants that convert better depending on the source.

FAQ

Frequently asked questions about form A/B testing

How much traffic is needed for significant results?+

It depends on your current conversion rate and the size of the effect being tested. As a rule of thumb, you need at least 1000 impressions per variant to detect a 20 percent gain with 95 percent confidence. Fullmetrix calculates the required sample size automatically.

How many variants can I test at once?+

Fullmetrix lets you test up to 4 variants in parallel on the same form. Beyond that, traffic is too dispersed to get significant results in a reasonable timeframe. To test more variants, run successive tests keeping the winning version.

Does the same visitor always see the same variant?+

Yes. Fullmetrix uses an identifier stored in local storage to guarantee that a visitor always sees the same variant throughout their journey, even if they return multiple times. This guarantees test validity and a consistent experience.

When is a test considered finished?+

A test is finished when it reaches 95 percent statistical significance, meaning the probability that the observed difference is due to chance is less than 5 percent. Fullmetrix automatically signals the winner and suggests deploying the winning version.

Can I test multiple elements at once in a variant?+

Technically yes, but it's discouraged. If you change both the title and button and the variant wins, you don't know if it's thanks to the title, the button or the combination. The most informative tests change one element at a time.

Do A/B tests slow down my site?+

No. The Fullmetrix snippet is lightweight (under 15 KB) and asynchronous. Variants are selected client-side with no additional server request. The impact on load time is imperceptible and doesn't affect your Core Web Vitals.

Explore