---
title: "A/B testing in RightMessage"
description: "You'll see what you can A/B test, where each test lives, how tests reach a verdict, and which plan each kind needs."
canonical: "https://rightmessage.com/help/ab-testing-in-rightmessage"
last-updated: 2026-11-06
---

# A/B testing in RightMessage

You'll see what you can A/B test, where each test lives, how tests reach a verdict, and which plan each kind needs.

Find out which version of a page, an audience's message, a question, an offer, or a whole flow gets more people to convert, then keep the winner. RightMessage tests on the pages you already have, checks results every week, and tells you in plain words when a version has won, lost, or can't be told apart.

For example, [WorkForge](https://workforge.rightmessage.com) could test a trust-first homepage headline against its usual one, try a shorter pricing button for its lead audience, and compare two wordings of its team-size question, all from one project.

## What you can test

| If you'd like to test…                                   | Run                                                                                      | Where you set it up                 | Plan                   |
| -------------------------------------------------------- | ---------------------------------------------------------------------------------------- | ----------------------------------- | ---------------------- |
| The page everyone sees when no audience matches          | An Original page test, with your page as it is as Control and up to three variants       | The campaign editor                 | Starting needs **Pro** |
| What one audience sees on a page                         | An audience test, with the audience's own changes as Control and up to three challengers | The campaign editor                 | Starting needs **Pro** |
| A RightMessage-hosted landing page, or one element on it | A page test                                                                              | The page editor's **Test** panel    | Starting needs **Pro** |
| Whether a whole campaign beats your page as it is        | The campaign holdback                                                                    | The campaign's **Testing** tab      | **Scale** and **Pro**  |
| A question's wording, or a form's or offer's content     | A question, form, or offer test                                                          | The question, form, or offer editor | Every plan             |
| A different order or length of flow                      | A flow test                                                                              | The flow canvas                     | Every plan             |

You can set up a page or audience test on any plan and start it once you're on **Pro**.

## Find every test in one place

**A/B Testing**, under **Optimize** in the sidebar, lists every test in your project: page and audience tests from your campaigns, and question, form, offer, and flow tests. Tests are grouped into **Active tests**, **Draft tests**, and **Completed tests**. A test you are still setting up, or one that starts on your next publish, is under **Draft tests**. Each page and audience test shows its campaign, goal, start date, weeks running, status, and a one-line verdict such as "Direct pricing button beat Control." Choose a card to open the campaign's **Testing** tab.

![A/B Testing page with the Display and Filter controls, and Active tests listing a Lead audience test ready to end, an Original page test and a Customer audience test collecting evidence, each with its goal, start date and weeks running](https://rightmessage.com/help/screens/ab-testing-index.light.webp)

Type in **Filter** to search test names and types. **Display** narrows the list by kind (**Page tests**, **Audience tests**, **Offer tests**, **Question tests**, or **Flow tests**) and by status. Choose **Needs decision** to see only the tests waiting on you, such as a test that's ready to end.

![A/B Testing Display menu open, with a Display type of All tests and a Status of All statuses](https://rightmessage.com/help/screens/ab-testing-filters.light.webp)

A campaign's own tests also appear in the **A/B tests** section of its **Overview**, with **View in Testing** to manage them.

![A/B tests section of a campaign’s Overview, with the Customer audience test collecting evidence, the Lead audience test ready to end, and View in Testing](https://rightmessage.com/help/screens/campaigns-overview-ab-tests.light.webp)

The **Tests** card on your [Dashboard](https://rightmessage.com/help/dashboard-reference) lists running tests too, and [Needs attention](https://rightmessage.com/help/needs-attention-reference) tells you when a test pauses, stops counting visitors, can't reach a verdict, or is ready to end.

## How a page or audience test runs

1. **Set up.** Build the versions in the editor, pick the conversion goal that decides, and check the estimate of how long a verdict might take. See [run an A/B test on a page or an audience](https://rightmessage.com/help/run-an-ab-test-on-a-page-or-audience).
2. **Start.** **Start test** is a draft change. The test begins when you publish, and the card reads **Starts on Publish** until then.
3. **Collect evidence.** RightMessage checks the results every Friday. The earliest verdict comes at the week 2 check, about three weeks in, and many tests take longer.
4. **Get a verdict.** A version wins when its whole likely range beats Control and loses when it's all worse. With no verdict by week 26, the test is inconclusive. See [how verdicts work](https://rightmessage.com/help/campaign-experiment-decisions).
5. **End and keep a version.** Tests never end on their own. When the card reads **Ready to end**, or you've waited long enough, end the test and keep the version you want. An Original page winner can stay deployed until your website has the change. See [end a test and deploy the winner](https://rightmessage.com/help/end-a-test-and-deploy-the-winner).

![Review test window for an Original page test in setup, with the goal, the split between Control and Variant A, the time-to-verdict estimate and Start test](https://rightmessage.com/help/screens/editor-review-test-dialog.light.webp)

Low traffic never stops you from testing, but it can make a verdict slow or unlikely. The estimate says so before you start. See [read A/B test results and test with low traffic](https://rightmessage.com/help/read-ab-test-results-and-test-with-low-traffic).

Question, form, offer, and flow tests work differently: they show a confidence figure rather than weekly verdicts, and you choose the version to keep. See [test questions, forms, and offers](https://rightmessage.com/help/split-testing-questions-and-offers) and [test whole flows](https://rightmessage.com/help/a-b-testing-and-split-testing-flows).

## Analyze tests in your own tools

RightMessage can send each visitor's test version to Google Analytics 4, Google Tag Manager, and PostHog, so you can break down any report by version. See [analyze tests in GA4 and Google Tag Manager](https://rightmessage.com/help/analyze-tests-in-ga4-and-google-tag-manager) and [analyze tests in PostHog](https://rightmessage.com/help/analyze-tests-in-posthog).

## Let Righty set it up

[Righty](https://rightmessage.com/help/righty-your-ai-assistant) can set up a page or audience test, write the challenger copy, and read the verdict back to you. Connected AI assistants can do the same through MCP, with [campaign tools](https://rightmessage.com/help/mcp-tools-campaigns) for page and audience tests and [split test tools](https://rightmessage.com/help/mcp-tools-tests) for question, form, offer, and flow tests.

RightMessage keeps testing simple on purpose: no multivariate tests, redirect tests, or bandits. See [what RightMessage testing doesn't do](https://rightmessage.com/help/what-rightmessage-testing-doesnt-do).
