Back to Features

User journeys

Follow each user through your app: every page and action in order, with an AI summary of what they did and where they got stuck.

Everyone who used your product

One list of every person who has been in the app, with their email or, for people who have not signed up, the id they are known by. Each row carries the country, the stage they are at, how many sessions they have had and when they were last in.

Search runs over emails, ids and the summaries themselves, so you can look for a person or for a behaviour — everyone whose summary mentions the cancel flow, for example.

The summary

Every user gets a few sentences about what they actually did, written after the sessions, not from a template. It names the behaviour and what it suggests: started the cancel flow and picked a budget reason, then went off to explore analytics, then rage-clicked on Videos.

The date it was written sits next to it, so an old summary reads as old. A brand-new user says the summary is not ready yet rather than inventing one from a single visit.

The numbers on a person

Total sessions, total time, average session length, days active, the stage they are in now, and when they were last seen. Enough to tell a heavy user having trouble from a casual one who never got started — two situations that look identical in an aggregate chart.

Sessions, day by day

Sessions are grouped by day, newest first. Each one shows the time it started, how long it lasted, the device and browser, and how many events and pages it covered.

Next to that is a sentence describing what the session was: processed three catches, opened each, clicked Copy, Reply now and Complete, three times over. You can read a month of someone’s use without opening anything.

Inside a session

Open a session and you get the timeline: every click and every page in order, with the time and the exact address. Repeated sequences are described as sequences rather than listed twenty times, and the detail under each line says what was clicked and where.

What it notices

  • Rage clicks — the same thing clicked repeatedly, which usually means it looked clickable and was not.
  • Cancel flows, including the reason the person selected.
  • Sessions that bounce after a few interactions on one page.
  • Switching between sections, which tells you what they were looking for and did not find.

When to open it

Before answering a support message, so you already know what they tried. Before a call with someone who is about to cancel. And after a release, to watch the first few people meet the thing you shipped.

When the same stall shows up across several people, it stops being a story about one user and becomes a fix.

You shipped a product.

Let's get it earning.

Get started