> For the complete documentation index, see [llms.txt](https://docs.humblytics.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.humblytics.com/split-testing-overview.md).

# Split Testing Overview

Show different versions of a page to different visitors at the same time, and let the data tell you which one wins.

A split test, also called an A/B test, sends some visitors to your current page (**Control**, or version A) and others to a new version (**Variant B**). Humblytics counts how many people in each group do the thing you care about, like submitting a form or buying, and tells you how likely it is that one version is truly better.

In the app, go to **Optimize → Experiments**.

## Two ways to build a test

| Test type                  | How it works                                                                                                                                                                                                                                                                                                                 | Best for                                                |
| -------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------- |
| **Element test** (default) | Click **New test**, pick a page, and change it in the [visual editor](https://docs.humblytics.com/split-testing-overview/creating-a-b-page-variants-for-split-testing.md#visual-editor-tests): text, styles, images, show or hide, reorder. Humblytics applies the same edits for visitors in Variant B. Nothing to publish. | Headlines, CTA buttons, images, small layout changes    |
| **Split by URL**           | Send visitors to two different live URLs, such as `/pricing` and `/pricing-b`. You build and publish both pages first.                                                                                                                                                                                                       | Full redesigns, new landing pages, different page flows |

Both use the tracking script you already installed. There's nothing extra to add.

## How a test runs

1. **Pick a goal:** what counts as a win, like form submissions or payments. See [Conversion Goals](https://docs.humblytics.com/split-testing-overview/creating-a-b-page-variants-for-split-testing/conversion-goals.md) and [Engagement Goals](https://docs.humblytics.com/split-testing-overview/creating-a-b-page-variants-for-split-testing/engagement-goals.md).
2. **Build your variant:** in the visual editor, or as a second page.
3. **Optionally narrow the audience:** limit the test to certain devices, traffic sources, or countries. See [Audience & Location Restrictions](https://docs.humblytics.com/split-testing-overview/audience-and-location-restrictions.md).
4. **Launch:** Humblytics splits visitors between versions.
5. **Wait for enough data:** the [Test planner](https://docs.humblytics.com/split-testing-overview/using-the-humblytics-a-b-sample-size-calculator.md) tells you how many visitors you need. Tests stop on their own after 30 days unless you set a different limit.
6. **Call it:** when the test is ready, keep the winner and plan the next test.

## In this section

* [How to Set Up a Split Test](https://docs.humblytics.com/split-testing-overview/how-to-setup-a-split-test.md): the setup steps, click by click.
* [Creating A/B Page Variants](https://github.com/Humblytics/humblytics-docs/tree/main/split-testing-overview/creating-a-b-page-variants-for-split-testing/README.md): the visual editor, and building a second page for Split by URL tests.
* [Audience & Location Restrictions](https://docs.humblytics.com/split-testing-overview/audience-and-location-restrictions.md): run a test for one device, channel, or country.
* [How to Analyze Split Test Data](https://docs.humblytics.com/split-testing-overview/how-to-analyze-split-test-data.md): read results and decide what to ship.
* [Using the Test Planner](https://docs.humblytics.com/split-testing-overview/using-the-humblytics-a-b-sample-size-calculator.md) and [Deciding How Long to Run a Test](https://docs.humblytics.com/split-testing-overview/deciding-how-long-to-run-an-a-b-test.md): plan before you launch.
* [How Split Testing Works Under the Hood](https://docs.humblytics.com/split-testing-overview/how-humblytics-split-testing-works-under-the-hood.md): assignment, flicker, and the math.
