Rebuy's A/B Testing is a tool for optimizing conversion rates (CRO) that enables you to experiment with on-site widgets and the underlying data sources. It empowers them to make decisions based on data, ensuring a customized customer experience that drives scalable revenue growth and meets you customers' expectations for personalization.

We offer two distinct types of experimental tools: one that drives widgets and another that serves as a general tool, allowing you to manage elements within your storefront through CSS and JS. 🛠️✨

Rebuy encourages you to utilize this tool to enhance you customer experience, boost average order value (AOV), increase revenue, extend customer lifetime value (LTV), and potentially reduce customer acquisition costs (CAC).

Those who are seeking to achieve the following goals can benefit from A/B Testing:

Want to learn more about Rebuy features? Book a demo with our team today!

The setup process for experiments is effortless, taking less than 5 minutes. Once a test is completed, the winning widget, determined based on the selected Experiment Goal during setup, will automatically be used for all incoming traffic. Accessing and reviewing reports is always convenient, directly from the A/B Testing Dashboard.

The A/B Testing dashboard provides a centralized view of ongoing, draft, and completed experiments, along with their respective metrics. From the dashboard, creating new experiments and getting started is simple. Completed tests can be accessed and reviewed at any time without the worry of them being automatically archived.

Test up to 9 variations of widgets against a control widget. Variations can consist of different widgets, the same widgets with different data sources, or even no widget at all. Note: If "Do Not Show Anything" is selected, certain metrics may be unavailable (e.g., comparing conversion rates across variations requires a widget for conversions).

Schedule experiments to start and run for a specific duration or initiate an experiment immediately, allowing it to run indefinitely until manually ended. If there's a need to conclude a scheduled test earlier, it can be easily ended with a simple click.

Currently, you have the option to conduct A/B testing on all widget types except for the following:

We are actively working on enhancing our capabilities for future iterations of A/B testing to accommodate a broader range of widget types. Keep an eye out for new releases with additional A/B testing capabilities!

When conducting an A/B test, it is necessary to track attributions in order to receive the maximum benefit from your test. Attribution tracking supports conversion optimization in a number of ways, such as helping you determine the effectiveness of different widgets, providing an understanding of user behavior on each page, and optimizing ad spend with the most profitable channels.

Because of the importance of attribution tracking for A/B tests, it is not recommended to disable the attribution tracking for widgets in your A/B test. Doing so will cause the test to fail, and will not provide the information you need from your test.

For more information on A/B testing with widgets, please see our user guide on A/B testing with widgets.

Although A/B testing offers extensive customization options to meet your specific requirements, we have compiled a best practices guide to provide helpful tips and tricks for conducting a successful A/B test.

A/B Testing feature allows for up to 10 total variations (1 control and 9 variations).

Yes you can test different widget types against one another.

Currently Post Purchase A/B testing is housed within the Post Purchase flows. You can follow the Post Purchase Flow guide to set up your A/B testing experiment. A/B testing for post purchase is a bit different due to the nature of the widget type so it remains separate at this time.

You have the option to test Smart Cart using the general A/B experiments feature within Rebuy. Alternatively, if you prefer a third-party solution, you can utilize Experiments.js, a JavaScript package developed by our team. With Experiments.js, you can choose your preferred A/B testing suite, whether it's Google Optimize, VWO, Optimizely, or even your own custom solution, providing you with flexibility in your testing approach.

Yes, you can technically A/B test data sources. If you wish to do so, you will need to build out a duplicate widget and have each widget pointed to the respective data source. You can duplicate the widget by following these steps.

Once you have the duplicate widget, you can adjust your data source rules to power each widget. From there you can use those two widgets to A/B test against each other, which will ultimately A/B test each data source.

Yes, if you would like to A/B test the placement of a specific widget. If you wish to do so, you will need to build out a duplicate widget and have each widget pointed to the respective dynamic placement. You can duplicate the widget by following these steps.

First you will need to make sure you set up each widget as a dynamically placed widget which you can do by following the steps in the Dynamic Placement Selectors guide.

In this example, we will outline building two PDP widgets. One that will live above the add to cart button, and the other to be under the add to cart button.

Once you have the two widgets built out with the dynamically placed selector configured, you can proceed to building out your A/B test.

The "winning" A/B test is chosen by the revenue or conversion (whichever you have chosen during the setup of the A/B test) of the page load results and not the specific performance of a widget. To learn more about the reporting structure for A/B testing, check out the Dashboards & Reporting article.

Yes, the widget(s) should be in live mode for the A/B test. However, as a reminder the widgets are not to be installed themselves. You must use the placeholder ID for experiments. Installing the placeholder ID is simple and you can follow the steps in this guide.

Regrettably, it's not possible to disable the A/B testing attributes, unlike the Smart Cart and Widgets, which offer that option. These attributes are essential for accurately tracking user sessions and obtaining reliable A/B testing outcomes.

If you want to verify your widget A/B testing experiment is firing you can follow these steps:

If the _r_experiment cookie is not present, then the A/B test did not load on the page you are on. This could be due to various reasons such as incorrect implementation or targeting settings. However, if the _r_experiment cookie is present, that means that an experiment did load in. You can verify which experiment by viewing the entire cookie to find the experiment ID. If you see multiple experiment cookies, this can happen if there are multiple A/B tests running on the same page.