# A/B testing

> Run a controlled experiment on a form to find which headline, button text, or variant converts better.

A/B testing splits your form's traffic between a control (the current form) and a variant, then measures which version converts better. Use it to test copy changes, field order, or entirely different form designs.

> **Note** The A/B Test tab is available on saved, published forms. You must save and publish a form before you can run a test on it.

## Create a test

1. Open your form and select the **A/B Test** tab.
2. Click **New Test**.
3. Fill in the test details:
   - **Test Name** — a short label for your records (for example, "Headline Test Dec 2025").
   - **What to test** — choose the element to vary:

| Option | What changes |
|---|---|
| **Headline Test** | Tests two different headline texts |
| **CTA Test** | Tests two different submit button labels |
| **Field Order** | Tests two different arrangements of your fields |
| **Full Form** | Tests two completely different form configurations |

4. The **Control** row shows the current value for the element you chose. Enter the test value in the **Variant A (Test)** row.
5. Traffic splits 50/50 between control and variant automatically.
6. Click **Create Test**.

<!-- screenshot: the Create A/B Test dialog with Headline Test selected, showing Control and Variant A rows -->

## Manage a running test

While a test is running, the **A/B Test** tab shows:

- The test name and current status (**Running**, **Paused**, or **Completed**).
- A results table with conversion rate, views, and conversions for each variant.
- The winning variant (marked with a trophy icon) once a clear leader emerges.

To pause a running test, click **Pause**. To resume it, click **Resume**. To stop the test and lock in results, click **End Test**.

> **Important** Ending a test is permanent. The test moves to **Completed** status and can't be restarted. If you want to apply the winning variant, update the form manually in the editor.

## Read the results

| Column | What it shows |
|---|---|
| **Views** | How many visitors were shown this variant |
| **Conversions** | How many of those visitors submitted the form |
| **Conversion rate** | Conversions ÷ views for this variant |

The variant with the higher conversion rate is highlighted as the winner. Collect enough traffic before drawing conclusions — a low-traffic form may need days or weeks to reach a meaningful sample size.

## Past tests

The **A/B Test** tab lists your last 5 completed tests below the active test. Each row shows the test name, its final status, and the winning variant if one was declared. Click **Delete** to remove a past test record.

## Next steps

- [Form analytics](form-analytics) — Track overall conversion rate alongside test results.
- [Fields and conditional logic](fields-and-logic) — Redesign fields based on what your tests reveal.
- [Create a form](create-a-form) — Build a new form variant to use in a Full Form test.