Skip to main content

A/B Testing

If you want to test two versions of an agreement — different wording, layout, or consent options — you can run them side by side on the same page. Click tracks each version independently so you can compare acceptance rates and decide which performs better.

Setting up a test​

  1. Create and publish the templates you want to compare.
  2. Go to Click → Deployments → + New Deployment.
  3. Set the domain, path, and environment.
  4. Enter a Variant label — e.g. control.
  5. Create and activate the deployment.
  6. Repeat for your second version using a different label — e.g. variant-a — on the same domain and path.

Both deployments can be active at the same time because they have different variant labels. Click routes users between them and tracks results separately. For background on how domain and path matching works, see Domain & Path Configuration.

Use clear, consistent labels

Labels like control, variant-a, and variant-b appear directly in your analytics reports. Consistent naming makes comparison much easier.

Coming Soon

Screenshot: click-ab-variants-list, screenshot of the Deployments list showing two active deployments on the same domain/path with variant labels "control" and "variant-a" and separate view/acceptance counts

Measuring results​

Go to Click → Analytics and use the Variant breakdown to compare views, acceptances, and acceptance rates side by side. See Metrics for a full explanation of each number, and Reporting to export variant data.

Coming Soon

Screenshot: click-ab-analytics, screenshot of the Analytics view showing a variant comparison table with columns for Views, Acceptances, and Acceptance Rate for "control" vs "variant-a"

Ending a test​

  1. Deactivate the losing variants using the Active toggle on each deployment.
  2. Keep the winning variant active. You can remove its label to simplify it back to a standard deployment.
Coming Soon

Screenshot: click-ab-end-test, screenshot of the Deployments list after a test concludes, showing the losing variant with the Active toggle off and a greyed-out row, while the winning variant remains active

Evidence is preserved

Every evidence bundle records which variant was shown to each user. Historical records are kept even after you deactivate the other variants.

Next: Scheduled Deployments →