Skip to main content

Understand why users are behaving a certain way

When you see something unexpected in your data: a spike in drop-offs, a confusing usage pattern, a frustration signal without a clear explanation. Ask Fullstory MCP to investigate. It will read real sessions and tell you what's actually happening.

Example questions​

Why are so many users going back and forth between the product page 
and the cart without checking out?
We're seeing a spike in checkout abandonment on the shipping step.
What's causing it?
Users are searching for "return policy" right before they abandon checkout.
What are they looking for that we're not showing them?
Why did our conversion rate drop on mobile after last Tuesday's release?

What you get back​

A behavioral explanation grounded in what users actually did: not just where they dropped off, but what they were trying to do, what they found (or didn't find), and what a fix might look like.

Shipping step abandonment: Investigation

Reviewed 6 sessions from users who abandoned at the shipping step.

What happened:
All 6 users got to the shipping step and paused at the delivery
options section. 4 of them scrolled back up to the order summary
before abandoning, suggesting they were looking for information
that wasn't surfaced at the decision point. 2 scrolled down to
the bottom of the page, then left without selecting a shipping method.

What they were doing:
The scroll-back-to-summary pattern is consistent enough to suggest
users needed more context before they could choose a shipping tier.
3 users spent over 30 seconds on this step before abandoning,
well above the median time of 8 seconds for users who complete it.
Two users navigated away from the site entirely within seconds of
leaving the shipping step.

Suggested direction:
The hesitation and scroll behavior points to a missing piece of
information at the shipping selection step. The session evidence
alone can't confirm exactly what's missing; the visual replay
would let you see what the shipping options actually look like
at that moment and what context users may have been looking for.

Following a thread​

The most useful investigations often come from noticing something odd and just asking about it directly:

Users who visit our size guide page have a lower conversion rate 
than users who don't. That seems backwards. Why?
We have a 40% drop-off rate on step 2 of our checkout, 
but step 2 is just entering an email address. What's happening?
Some users are spending 10+ minutes on the product page before 
adding to cart. What are they doing?

The AI will pull the relevant sessions, read through them, and explain the behavior. You don't have to form a hypothesis first, or know what to look for.

When you want to see the session yourself​

Every session the AI references includes a direct link to the replay in Fullstory. Use it when you want to see the visual state of the page: its layout, rendering, and what elements were visible at the moment something happened, rather than just the event sequence.