# What RightMessage testing doesn't do

Get straight answers about the kinds of tests and statistics RightMessage doesn't offer, and what to use instead.

Available on the Scale and Pro plans.

RightMessage is a personalization platform with built-in A/B testing. You can test the page everyone sees and each audience's changes, with plain-language weekly verdicts. It isn't an enterprise experimentation suite, so here's what it leaves out.

## Can I run a multivariate test?

No. Each version is a whole alternative, and a test has at most four versions: Control plus up to three. RightMessage doesn't test every combination of separate element changes. Put the changes you want to compare together into whole versions instead.

## Can I split traffic between two different URLs?

No. Campaign tests change content on the same page; there are no split-URL or redirect tests for campaigns. Build the alternative as a version of the page in the [campaign editor](https://rightmessage.com/help/run-an-ab-test-on-a-page-or-audience).

## Does RightMessage shift traffic toward the winner automatically?

No. There are no bandits or auto-allocation. Each version keeps the split you set, even or custom from 1% to 99%, until you end or restart the test.

## Does it show a "chance to beat" or "% confident"?

No. There's no Bayesian "chance to beat", and campaign tests never show p-values, "% confident", or countdowns. Each version has a likely range and a plain verdict: **Winner**, **Did worse**, **No clear winner**, or **Collecting evidence**. Question, form, offer, and flow tests still show their older [confidence display](https://rightmessage.com/help/split-test-results-reference).

## Can I see significance in real time?

No. Results are checked every Friday, and each week's visitors get 7 days to convert before they count. The earliest verdict comes at the week 2 check, about three weeks after the test starts. Because the checks are weekly and planned in advance, looking every week doesn't raise the chance of a false winner. See [how verdicts work](https://rightmessage.com/help/campaign-experiment-decisions).

## Will a test end itself and roll out the winner?

No. Tests never end on their own, and RightMessage never rolls out a winner for you. When a test is ready, the card says **Ready to end** and you choose the version to keep. See [end a test and deploy the winner](https://rightmessage.com/help/end-a-test-and-deploy-the-winner).

## Is there server-side testing, an SDK, or feature flags?

No. Tests run through the RightMessage snippet in the browser. To stop visitors seeing the original before their version, add the `rmcloak` class to tested elements. See [prevent a flash of original content](https://rightmessage.com/help/how-to-prevent-flash-of-original-content-fooc-with-the-rmcloak-css-class).

## Can I keep visitors out of two tests at once?

Not across campaigns: there are no mutually exclusive experiment groups. Within one campaign, an audience a visitor matches wins on the elements it changes, and the Original page test covers the rest.

## Are there heatmaps or session recordings?

No. Pair RightMessage with a heatmap tool if you need them. RightMessage sends which version each visitor saw to [Google Analytics, Google Tag Manager, and PostHog](https://rightmessage.com/help/analyze-tests-in-ga4-and-google-tag-manager), so you can break down other reports by version.

## Does RightMessage build audiences for me automatically?

No. You choose the audiences, from your segments, survey answers, and email platform or CRM data. RightMessage doesn't create machine-learned cohorts.