Insights is the page that answers "how are we actually doing?". More usefully, it lets you click any answer to see the conversations it came from.

What Insights is

A single screen of numbers about your support, built from your threads. There is nothing to set up: it starts working the moment you have conversations.

You reach it from Insights in the left-hand sidebar. It has three tabs: Overview, which is everything below, Exports, and Advanced, which is marked coming soon.

The three filters

Along the top are three controls, and every number on the page obeys all three.

The date range: Which period you are looking at. Headline numbers compare against the previous period of the same length.
For: Everyone, or one person. This is how you look at a single teammate's work.
Channel: All, or one channel. Worth using, because live chat and email behave so differently that averaging them together hides both.

Change the channel filter before drawing conclusions about response times. A team that looks slow overall is often fast on chat and slow on email, which is a different problem with a different fix.

What is on the page

Six sections, in a deliberate order: from what customers experienced, to what your team did, to what they were talking about.

Headlines: The three numbers that summarise the period: first response, resolution time and satisfaction.
Threads by status: Where every conversation from the period stands right now.
AI performance: How much of the work Chloe carried.
Common topics: What customers wrote about.
Common frustrations: Where conversations turned negative.
Kudos: Real customer words from five-star ratings, with the teammate who earned them.

Every number opens its threads

This is the part worth knowing, because it changes Insights from a report into a tool.

Nearly every row is clickable, and clicking opens the conversations behind it. A star rating row opens the threads rated that way. A status row opens the threads in that status. A topic opens the threads about it.

So the useful move is never to look at a number and think about it. It is to click it and read three of the conversations.

Numbers tell you something changed. Only the conversations tell you why, and the why is what you can act on.

Exports

The Exports tab produces raw files of your workspace data, for when you want to work somewhere else.

1Open Insights, then Exports.
2Click Start an export.
3Pick a dataset. Today that is threads and their resolution outcomes, with more coming.
4Choose the scope: a time window and the channels you want.
5Download the file once, or have it delivered on a schedule by email or webhook.

Export files expire after seven days, and the download links are private. If you need a file kept, save it somewhere of your own rather than relying on the link.

How often to look

Weekly is about right for most teams, and the honest reason is sample size. Support numbers are noisy day to day, and one difficult customer can move a satisfaction score several points. Reacting to daily movement mostly means reacting to noise.

Look daily only at the two things that are about individual conversations rather than averages: overdue threads and common frustrations. ---

Reading your Insights overview