Configuring A/B Testing

A/B, or multivariant, testing provides out of the box capabilities to test how the existing template(s) used by a domain perform compared to up to three variants. This is scoped to the specific geolocations so that you may respect different jurisdictional guidance related user journey testing where required. For additional information, see Cookie Guidance.

For more information on optimizing your A/B testing, see A/B Testing Best Practices.

To configure A/B testing

  1. On the Cookie Consent menu, select Testing Center. The Testing Center screen appears.

  2. Click the Add New button. The General Information screen appears.

  3. Enter the Test Name, Domain, and a Description (optional).
    Important:

    You can only run one test per domain. If you have an existing test running on a domain, that domain will not display for selection.



  4. Click the Next button. The Variants Setup screen appears.

  5. For each geolocation rule group applicable to the domain, select the template you want to test for each variant. You can add up to three different templates/variants to each rule.

    A/B testing will use only the templates added as variants. Audience logic overrides any standard template (i.e., audience logic will display a configured variant instead of the template set within Geolocation Rules).

    Tip:

    You can enable or disable testing per geolocation rule group using the Enable Test setting.



    Important:

    On each template, you must enable the Override Template Styling setting before A/B testing can function as expected. For more information, see Customizing the Banner Template and Customizing the Preference Center Template.

  6. Click the Next button. The Objectives Setup screen appears.

  7. Select the time duration (in days) for your test.
    Tip:

    Enable the Email Notifications after Completion of Test setting if you would like to be emailed when the test is completed.



  8. Click the Finish button. The Test Details screen appears.


    Note:

    You can edit the test at any time using the Edit button on the Test Details screen.

  9. To start the test, click the Publish Production button. For more information, see Publishing and Implementing Cookie Consent Scripts.
    Important:

    One your test has begun, you cannot edit the test configuration (duration, success criteria, etc.).

  10. Once the test has started you can monitor results under the Report tab. When the test is over, you will be able to update your template and optimize your CMP based on the results.

To manually end a test

You can end your tests manually before they complete, if necessary.

  1. On the Cookie Consent menu, select Testing Center. The Testing Center screen appears.

  2. Click the name of a test with the Running status. The Test Details screen appears.

  3. Click the End Test button. The End Test modal appears.


  4. Select an option to either

    • End the test early and collect results; or

    • Cancel the test and discard results.

  5. Click the Confirm button. You will be prompted to publish your production scripts to end the test. For more information, see Publishing and Implementing Cookie Consent Scripts.

To view and export testing reports

  1. On the Cookie Consent menu, select Testing center. The Testing center screen appears.

  2. Click the name of a test with the Completed status. The Test Details screen appears.

  3. Select the Report tab. The Report tab appears.

  4. Select a default audience.
  5. View the reporting metrics. The Report tab includes information on the following:

    • Total visitors

    • Opt-in percentage by variant. category, and over time

    • Opt-out percentage by variant, category, and over time

    • Not given percentage by variant, category, and over time

  6. To export the results, click the Export results button. The Export results modal appears.

  7. Select the file types (.csv or .pdf) you want to exports the report as.
  8. Click the Export button. A notification appears when the export is ready.
  9. Click the Notifications icon . The Notifications popover appears.

  10. Click the export to download it to your device.
Important:

OneTrust now offers enhanced A/B testing exports that leverage Insights reporting. This enhancement includes

  • More reliable percentage calculation, avoiding misleading results.

  • Full visibility into user behavior without hidden gaps.

  • Greater transparency into the entire dataset.

  • More realistic representation of segment performance.

  • On-screen dashboards and exported reports that match exactly.

The Insights Reporting for A/B Testing Exports feature is under Public Preview. Please contact OneTrust Support to request early access.

Frequently Asked Questions

For more information on optimizing your A/B testing, see A/B Testing Best Practices.

What does the A/B testing functionality do?

With A/B testing, you can run different variant tests per different location. This means you will be able to add a variant per geolocation rule.

Is A/B testing limited to the Cookie Banner or can you adjust the Preference Center?

The variants can be different for both the Cookie Banner and Preference Center because variants are configured at the template level.  It is recommend that you only change one aspect at a time so you can understand what is affecting your opt-in rate.

How does the randomization for the A/B test occur?

This happens in the code you have via JavaScript function. We do not perform algorithmic selection.

How is the information captured in the A/B test stored?

The information is stored the same way as information captured through standard consent logging.

Is this data anonymized?

Yes, personally identifiable information is not used. It is based on a geolocation.

Is the A/B testing cookie-based?

Yes, we can load cookies pre-consent.

What parameters are available to test?

Currently, just country (geolocation) is available to test. Newer versions will be able to test age and a custom variable for a user.

What is the maximum number of variants we can test today?

Three variants can be tested under one control group.

Can you edit the details of a test (duration, success criteria, etc.) while a test is running?

No. Tests cannot be edited after launching.

Why does audience logic display a different template than what's configured in Geolocation Rules?

Only the templates added as A/B testing variants will be displayed. Audience logic overrides the domain's standard template configured in Geolocation Rules.