# Test questions, forms, and offers

You'll compare two versions of a question, form, or offer, see how visitors respond, and choose which version to keep.

Available on the Pro plan.

Find out which wording gets more people to answer, click, or submit a form without building a separate flow. For example, [WorkForge](https://workforge.rightmessage.com) could compare `How big is your team?` with `How many people will use WorkForge?` while keeping the same answers and branches.

A test runs everywhere that question, form, or offer appears, not just in the flow you opened it from. Make sure a published flow or widget can show it to visitors. These tests are separate from [testing personalized pages against their original content](https://rightmessage.com/help/a-b-testing-your-website-personalization).

## 1. Create your alternative

Start with one change you'd like to compare. Keep the original as **Control**, then edit the alternative.

### For a question

1. Open **Questions** and select your question.
2. Open **A/B test**, then choose **Create A/B test**.
3. Edit the alternative's question wording or description. For a choice question, you can also change answer wording and images.
4. Compare the alternative with **Control**.

The answers, their order, segments, and branches stay the same in both versions. To compare question order or the number of steps, use a [whole-flow test](https://rightmessage.com/help/a-b-testing-and-split-testing-flows).

### For an offer or form

1. Open **Offers** or **Forms**, then select the item.
2. In its full editor, open **A/B test** and choose **Create A/B test**.
3. Select **Variant B** in the version selector.
4. In **Edit**, change the **Headline**, **Description**, **Button text**, or **Image** you want to compare.
5. Check the selected version in the preview beside the content.

Changes save automatically. Keep the alternative selected while editing so you don't change the shared original. A link offer can also test a different destination URL or new-tab setting. Form fields and post-submission actions stay the same in every version.

## 2. Check it before you start

Review both versions in the editor, then [test the visitor path on your site](https://rightmessage.com/help/best-practices-for-testing). Make sure visitors can reach the question, form, or offer and continue afterward. Creating a test won't make a hidden widget appear.

Finish your edits before starting. A running question test locks its wording, answers, display, behavior, and **Auto-Segment** rules. A running offer or form test locks its content, fields, and actions everywhere it's used.

## 3. Start the test and publish

1. Choose **Start test**.
2. Review the lock warning and confirm **Start test**.
3. Choose **Publish** to include the active test on your site.

Question tests split traffic evenly between versions and remember the assignment in the visitor's browser. Offer and form tests assign visitors randomly across versions.

> **Starting and publishing are separate**
>
> Publishing a draft test doesn't start it. Once you start a test, anyone publishing the project can include it. Don't use **Start test** as a private preview step.

## 4. Compare how visitors respond

Open **A/B Testing** under **Optimize**. You can also choose **Open in A/B Testing** from the question, form, or offer to go straight to its test.

For questions, compare visitors, answer rate, and lift. For offers and forms, compare how many people saw each version and how many converted. **Best Performer** names the current leader. It doesn't mean you need to end the test.

![Running team-size question split test showing visitors, answers, best performer and statistical confidence](https://rightmessage.com/help/screens/reporting-split-test-results.light.webp)

Use [Read and manage your split tests](https://rightmessage.com/help/split-test-results-reference) for the statistical details, version limits, preview options, editing locks, and deleting a draft you don't want to run.

## 5. Keep the version you choose

For a question:

1. Open its **A/B test** tab and choose **End test**.
2. Choose **Keep current wording** for control, or select the alternative's wording.
3. Confirm **End test**.
4. Review the draft and choose **Publish**.

For an offer or form:

1. Choose **End test** in its editor, or **End Test** on its central test card.
2. Select **Control** or an existing alternative.
3. Review your choice and confirm the end action.
4. Choose **Publish**.

Keeping control leaves the original in place. Choosing an alternative applies its content to the shared question, form, or offer after publication. Ending the test alone doesn't publish that choice.

When visitors see the version you kept wherever the item appears, the change is live. The completed test keeps a saved comparison and can't restart. Create a new test when you have another change to compare.