Dashboard and scans
The two scan scopes, what the numbers mean, and how to unstick a stalled scan.
What this tab is for. Running a scan and reading the summary. Everything else in the app reads from the snapshot a scan produces, so nothing works until you have run one.
License Waste Manager
One button, two scopes — and a third way in
There is a single Scan Users button. Clicking it asks what the scan should cover; both answers run the same phases, and the difference is which population they read. The same dialog also offers the third way to fill a snapshot: importing the admin-console CSV exports, which needs no API key for the data and takes seconds on tens of thousands of accounts.
| Scope | Who it reads | When to use it |
|---|---|---|
| Licensed users only | The people who hold a product licence — the seats you are billed for. | Routine use. Run it before every review. |
| Everyone in the directory | The same, plus accounts in your directory that hold no licence: guests, external accounts and leftovers from an integration. Noticeably slower on a large organisation. | Before a licence renewal, a formal audit, or a directory clean-up. |
Scans run in the background in phases and report progress. Only one scan or bulk action runs at a time. If a run was interrupted, Force Unlock releases the lock so you can start a new one.
The phases a scan goes through
The progress panel names the phase it is on. Two of them depend on the organisation API key, and one only runs for the wider scope — which is why the same scan can look different on two sites.
What the four numbers mean
- Billable Seats
- Seats you are billed for, counted once per product. Somebody with Jira and Confluence is two seats. This is the number that lines up with your invoice.
- Licensed People
- How many people hold at least one product, out of everyone the scan read. The same population as above, counted as colleagues rather than as seats — both are correct, and they answer different questions.
- Dormant Seats
- Seats with no activity for longer than your inactivity threshold. These are your candidates.
- Seats Dormant
- The dormant share as a percentage — not a monetary figure. The app has no price data. Multiply it, or the dormant seat count, by your own per-seat price before quoting a saving to anybody.
Utilisation by product
For each product: seats you are licensed for, seats in use, and the percentage. This tells you which of two very different actions to take.
- Utilisation is low and users are dormant — reclaim the seats. That is what this app does.
- Utilisation is low because you bought a bigger tier than you need — reduce the subscription with Atlassian. Reclaiming seats will not save you anything until you do.
Atlassian Guard — a different bill, answered on the same screen
Guard is billed per managed account, not per seat: every person whose email domain your organisation verified costs you Guard money, whether or not they ever open a product. The Guard block answers the questions that follow from that, and every figure is a link — click it and the Users tab opens on exactly those people.
- Managed accounts
- What Guard is charged on. Not the same population as “managed” in the account filter — verifying a domain makes an account manageable; being billed is a separate fact the scan reads separately. On one real customer, 8,551 of 37,686 managed accounts were billed.
- Paying Guard, holding no seat
- Secured, but not using any of your products. A licence scan cannot see these — they hold no seat to be found by — so this figure needs the wider scan scope or a CSV import, and shows an em dash until it has one.
- Paying Guard, dormant
- Billed for Guard and inactive past your threshold. The double waste.
- Paying Guard, working elsewhere
- Accounts on your verified domain whose product access lives on instances that are not yours — you pay Guard for them wherever they work. The sites are listed by name with a count each, and each one clicks through to the people who reach it.
Two more views under the tiles
- Inactivity Distribution
- Seats bucketed by how long they have sat idle: active (under 30 days), 30–60, 60–90, 90–180, 180+ and never active. This is the picture that tells you whether you have a slow drift or a cliff, and it is what makes the case for a threshold.
- Users by email domain
- People, licence holders, dormant seats and the share of seats, grouped by email domain — the fastest way to see a contractor population or an acquired company. Atlassian withholds the email address on many accounts; those are grouped as Email not visible rather than dropped.