Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/bastibuck/contao-ab-testing

A/B testing with Contao Open Source CMS
https://github.com/bastibuck/contao-ab-testing

ab-testing analytics contao conversion split-testing visitor

Last synced: 3 months ago
JSON representation

A/B testing with Contao Open Source CMS

Awesome Lists containing this project

README

        

# Contao: A/B Testing

[![Latest Stable Version](https://poser.pugx.org/bastibuck/contao-ab-testing/v/stable)](https://packagist.org/packages/bastibuck/contao-ab-testing)
[![Total Downloads](https://poser.pugx.org/bastibuck/contao-ab-testing/downloads)](https://packagist.org/packages/bastibuck/contao-ab-testing)
[![License](https://poser.pugx.org/bastibuck/contao-ab-testing/license)](https://packagist.org/packages/bastibuck/contao-ab-testing)

This extension provides a new page type to evenly split up visitor traffic between the selected test-pages. At the same time it makes sure that returning visitors will be shown the same page they have visited before.

This way A/B tests can be performed in Contao Open Source CMS.

## Features

- Select one or more pages to be tested
- test pages are rendered within the the new page type so all versions have the same url
- Set a cookie and its' expiring time to identify returning visitors
- adds a new Content Element to set a Matomo Custom Variable easily

## Notice

This extension **doesn't** provide analytics! To get insights on statistics of your visitors you will need to use an analytic software like Google Analytics, Matomo or similar.

## How does the extension work?

![A/B Testing visitor flow](../screens/AB-Test-Extension.jpg)

The visitor flow is visualized in the graphic above and shows how the extension works.

1. Visitor **a** is the first visitor to our A/B test. The A/B test page checks which test-page was last visited by any visitor. Since visitor **a** is the first visitor there will be no result and **a** will be rendered the first of the test-pages **V0**.
2. Visitor **b** is the second visitor. Checking which page was rendered before will now return the ID of test-page **V0**. Therefore visitor **b** will be given the second test-page **V1**.
3. When visitor **c** visits our A/B test the check of the last visited page will return the ID of the second page **V1** since the last visitor was shown this one. Visitor **c** will therefore be rendered test-page **V0** again as **V1** was the last page in the test.
4. Visitor **d** is a returning visitor. He visited the A/B test before and was given **V0**. By this time a cookie was saved on his device containing the value of the test-page he got to see. On the returning visit the A/B test checks if the cookie exists and if so gets its' value and renders the same page he has seen before. In this case this will also be **V0** although the next page to be visited would be **V1** again (if it wasn't a returning visitor). Additionally the last visited pge won't be saved in the case of returning visitors so the regular visitor order won't be messed up.
5. Visitor **e** is a first time visitor, has no cookie with a page ID value and therefore triggers the check of the last visited page which returns the ID of page **V0** (visitor **c** was the last one to visit it). Therefore visitor **e** will be shown test-page **V1**.