Understand how a feature is being used
Ask Fullstory MCP how your users are engaging with any part of your product and it will pull the data, surface behavioral patterns, and tell you what's working and what isn't — with session evidence where it matters.
Example questions
What you get back
A structured report with the key numbers, behavioral patterns, friction points, and what's working. For example:
Wishlist feature usage (last 30 days)
By the numbers
- 3,200 unique users added at least one item
- 41% returned within 7 days and added more
- 18% of wishlist creators later completed a purchase (vs. 9% baseline conversion for non-wishlist users)
How people use it
- 62% add a single item, never return
- 24% build a list of 3 or more items across multiple sessions
- 14% use it as a comparison tool: add, remove, then buy outside it
Friction
- Removing an item requires 3 clicks. 11% of users attempt removal twice before succeeding.
- Wishlist link is hidden in the account menu on mobile. Mobile return-to-wishlist rate is 19% lower than desktop.
What's working
- The "added to wishlist" toast is reliably effective. Click-through to the wishlist via the toast is 34%.
- Users who add 3 or more items convert at 2.4x the baseline rate.
Follow-up questions
Example follow-up
Reviewed 8 sessions where users attempted to remove a wishlist item.
Pattern A (5 of 8 sessions): heart icon confusion
- Users click the heart icon expecting it to remove the item
- The heart icon is decorative; the actual remove control is a small "x" in the upper corner
- Users click the heart 2 to 4 times, scan the page, then find the "x"
Pattern B (3 of 8 sessions): confirm-dialog hesitation
- Users find the "x" on the first try
- A confirm dialog appears asking "Remove this item?"
- Users hesitate 6 to 14 seconds before confirming, suggesting the dialog is doing work the button placement could do better
Suggested investigation: A/B test removing the confirm dialog and making the heart icon itself the toggle.
Behind the scenes
The MCP routes the chain above through: build_segment, build_metric, compute_metric, update_metric, and get_sessions.