Understanding A/B Testing

A/B testing (sometimes called split testing or bucket testing) is a technique for measuring how users respond to different variations of your content. The idea is straightforward: present two or more versions of a page element, such as a headline or call-to-action button, and measure which one drives better results.

Using Tags to Segment Your Traffic

Tag-based segmentation makes it possible to set up A/B experiments quickly within a privacy-focused analytics platform. Tags attach metadata directly to individual page views, which means you can label each visit with the variant that was shown and then filter your dashboard accordingly.

To implement this, you add data attributes with a data-tag- prefix to your analytics tracking snippet. Once tags are flowing in, they appear in your analytics dashboard and become available as filters.

A Practical Example

A common use case is testing two different headlines on a landing page. Here is the general approach:

Create two (or more) headline variants
Randomly assign each visitor to a variant, storing the assignment in localStorage so the experience stays consistent across sessions
Attach the variant identifier as a tag to your page view tracking snippet
Track conversion actions (like button clicks) as custom events

This setup lets you correlate specific variants with conversion outcomes.

Evaluating Your Experiment

Once data starts accumulating, navigate to the tag section of your analytics dashboard. Select the tag key and value corresponding to each variant to filter all metrics down to that segment. If your analytics tool supports a comparison mode, you can place two variants side by side to quickly spot differences in engagement and conversions.