The monthly support review habit that improved our product more than any user interview

We do user interviews. We run surveys. We track analytics.

But the single habit that has most consistently improved our product in the past year is a 60-minute monthly support review.

Here's exactly how we run it:

Who's in the room:
Support person, one product person, me (founder). Three people max. More than that and it becomes a meeting instead of a conversation.

What we review:

We pull the last month's tickets and look at three things only:

1. Volume by category — what types of questions increased or decreased vs last month?

2. The five tickets that took longest to resolve — why did they take long? What was hard about them?

3. The five tickets where the customer seemed most frustrated — what pattern do they share?

What we don't do:

We don't review individual performance. We don't celebrate good CSAT scores. We don't spend time on tickets that went smoothly.

We focus entirely on friction, confusion, and patterns.

What comes out of it:

Every single monthly review has produced at least one product or documentation change.

Not because we're doing anything clever. Because when you look carefully at where customers struggle, the answers are usually obvious once you're in the room together.

Some examples of changes that came directly from this review:

— We rewrote our plan comparison page after noticing that "what's the difference between plans?" kept appearing even though we had a comparison table. The table existed. Customers weren't finding it. We moved it.

— We added an in-app prompt after noticing that several customers had emailed asking about a feature that was two clicks away from where they were looking. They didn't know to look there.

— We changed how we communicate upcoming renewals after noticing that several "how do I cancel?" emails came in the week before annual renewal dates. The emails weren't about wanting to cancel. They were about confusion about what was about to happen to their billing.

All of these took less than one sprint to fix.
All of them reduced tickets in those categories noticeably the following month.

The review itself takes 60 minutes.
The fixes usually take less than a sprint.
The impact compounds because fewer tickets means more time for the review to find the next thing.

If you're not doing a monthly support review, it's the highest-leverage 60 minutes I'd recommend adding to your calendar.

What does your current process look like for extracting insight from support data?