/Docs

FullStory Integration

The FullStory integration attaches experiment enrollment data to FullStory sessions as custom events. Once connected, you can filter session recordings in FullStory to watch only visitors who were in a specific experiment variation — a powerful way to understand the qualitative experience behind your quantitative A vs B results.

Setup

1

Go to Project Settings > Integrations

Open your project, click Settings, then click the Integrations tab.
2

Enable FullStory

Find the FullStory card and toggle it on.
3

Enter your Org ID (optional)

Enter your FullStory Org ID. You can find this in FullStory under Settings > General. If your site already has the FullStory snippet installed (which sets window.FS), you can leave this blank and A vs B will auto-detect the initialized FullStory instance.
4

Save

Click Save. Experiment data will appear in FullStory sessions immediately on the next enrollment.

What is sent to FullStory

When a visitor is enrolled in a variation, A vs B calls:

javascript
1FS.event('avsb_experiment', {
2 experiment_id: 'exp_abc123',
3 experiment_name: 'Homepage Hero Test',
4 variation_id: 'var_xyz789',
5 variation_name: 'Variation B',
6 event_name: 'experiment_viewed'
7});

Viewing experiment data in FullStory

After the integration is running, experiment data appears in FullStory as custom events on each session timeline. You can use this data in several ways:

  • Segment recordings — in the Sessions search, filter by the custom event avsb_experiment and add a property filter for variation_nameequals "Variation B". This shows you only the sessions from visitors who were in that variation.
  • Funnels — create a FullStory Funnel that starts with the avsb_experiment event and includes your conversion steps. Break the funnel down by variation_name to compare drop-off rates between variations.
  • Session timeline — within any individual session recording, the experiment enrollment event appears on the timeline, showing you exactly when during the session the visitor was bucketed into the experiment.
Watch sessions from each variation
One of the most valuable uses of this integration is watching 5–10 session recordings from each variation side by side. Quantitative results tell you what happened; session recordings tell you why. Look for behavioral differences — does a visitor in Variation B interact with the new element as intended, or do they ignore it?