Get started pages

Get your first answer

Watch a real session, read the traffic behind it, and ask Home a question you could not have answered before.

Data arriving is not the same as data being useful. Once verify shows a session, these are the three things worth doing next — ending with a plain- language question on Home grounded in your own product.

Before you start

Verify should already show a session. Everything here reads the data you just recorded, so a project with nothing in it will show you empty states and teach you nothing.

Watch a real session

Open Replays and click the newest row.

Press play. Long sessions have a Skip inactive control that jumps the dead air, which is usually most of a real session.

The thing that makes a replay more than a video is the activity panel. Open it and you get the same timeline expressed as data: Events for what the person did, Console and Network for what the app was doing while they did it, Issues for anything that broke, and Web Vitals for how it performed. Scrubbing the player moves through all of them together.

This is the habit worth forming: when something is confusing, watch the session instead of reasoning about the funnel. It is faster, and it is evidence rather than inference.

Read the shape of your traffic

Open Dashboard and set the range to Today, then Last 7 days.

The KPI row across the top is your pulse: Visitors, Sessions, Pages per visit, Bounce Rate, and Active time. Revenue appears too, but only once Stripe is connected.

Below it, the cards break traffic down by Pages (with entry, exit, top, and inactive views), Sources, Events, Devices, and Geography. On a brand new project most of this is you, which is fine; the point right now is knowing where to look, not what the numbers say.

Ask Home a question

Open Home and type a real question:

What pages did people visit in the last 24 hours?

Home queries the project data and answers in plain language, and its answers link back into Dashboard, Replays, and Visitors. Follow one of those links — that round trip from question to evidence is the whole point of the product.

Try a second question that would have been annoying to answer with filters:

Which pages did people leave from most often this week?

Use New chat to start a clean thread. Past threads are kept under Chats in the sidebar.

Check Issues, and expect it to be empty

Open Issues. It defaults to the Runtime tab with status open.

Empty is a good result on day one: it means nothing has crashed. Do not read it as a broken install, because Replays already proved otherwise.

If something is there, open it. An issue is a group of similar errors, so what you get is the stack trace, the number of people affected, and links to replays of sessions where it happened. Watching the minute before a crash is almost always faster than reading the stack alone.

What good looks like

By now you have watched a session that actually happened in your product, seen your traffic broken down, and gotten an answer out of Home with a link back to the evidence behind it. That is the full loop the product is built around.

Next

Pick a Guide for the next real job (triage an error, watch a session, explain a dip), or identify users and track your own events so people and business moments show up in those workflows.