License utilization rate: how to measure it
License utilization rate is the share of the seats you pay for that someone actually uses: active users divided by paid seats, over a fixed window. A tool with 100 paid seats and 62 people who signed in over the last 30 days is running at 62 percent, and the other 38 seats are money leaving every month for nothing. The measure only means something once you fix three things: the window, what counts as active, and where the seat count comes from.
- Utilization = active users divided by paid seats, over a fixed window, per tool.
- Fix the window (30 days is the usual choice), and apply the same one to every tool.
- Active means a sign-in or a meaningful action, never a provisioned account.
- Below 70 percent on a renewing contract is worth acting on before the notice deadline.
- Measure per tool. A blended number across your stack hides the tool you should cut.
What is the license utilization rate formula?
The formula is deliberately plain.
Utilization rate = (active users in the window / paid seats) x 100
Everything difficult is in the three inputs, not in the arithmetic.
| Input | Where it comes from | The trap |
|---|---|---|
| Paid seats | The contract or order form, not the admin console | The console shows provisioned accounts, which can be higher or lower than what you bought |
| Active users | The vendor's usage or sign-in report | Vendors differ on what they call active, so read their definition before trusting the number |
| Window | Your choice, applied consistently | Changing the window between tools makes the comparison meaningless |
What counts as an active user?
This is where two teams measuring the same tool get different answers. Pick one definition and write it down.
- Provisioned: the account exists. This is not activity, and counting it gives you 100 percent every time.
- Signed in: the person authenticated at least once in the window. The usual working definition, and the easiest to get from any vendor.
- Did something: created, edited, sent or commented. The honest one for tools people open and abandon.
Sign-in is the practical default because every vendor reports it. Move to the stricter definition for expensive tools, where the gap between opening a tool and using it costs real money.
What is a normal license utilization rate?
There is no published benchmark that survives contact with a specific stack, so treat these as decision thresholds rather than industry figures. What matters is the action each band implies.
| Band | Reading | Action |
|---|---|---|
| 95 to 100 percent | Fully used, possibly constrained | Check whether people are sharing logins or waiting for a seat |
| 85 to 95 percent | Healthy | Nothing, keep measuring |
| 70 to 85 percent | Some slack | Reclaim before the renewal, do not yet cut the contract |
| Below 70 percent | Paying for people who are not there | Reclaim seats, then renegotiate the seat count at renewal |
A tool bought for a project that ended will sit far below 70 and stay there. That is not a utilization problem, it is a contract nobody cancelled.
How do you measure it without a spreadsheet per tool?
The work is not the division, it is keeping three moving numbers aligned per tool, every month. A worked example from the demo workspace: 24 licences across 6 vendors, 191 assets, and two tools whose seat counts had drifted from the contract because someone added seats mid-year.
- Record the paid seat count from the contract, with the renewal date and the notice deadline next to it.
- Pull the active user count from each vendor for the same window, on the same day each month.
- Divide, store the result, and keep the history. One month is a number; six months is a trend you can take to a renegotiation.
- Flag anything below your threshold far enough before the notice deadline to act on it.
The history is the part people skip and the part that wins the negotiation. A vendor will argue with one bad month. Six months of 58 percent is not an argument, it is a fact.
Utilization rate or adoption rate: which one are you measuring?
The two get used interchangeably and answer different questions, which is how a team ends up cutting seats from a tool that is working.
| Utilization rate | Adoption rate | |
|---|---|---|
| Question | Are we paying for seats nobody uses? | Are the people who should use this actually using it? |
| Denominator | Paid seats | Intended users |
| Owner | IT and finance | The team that bought the tool |
| Fix when low | Reclaim seats, resize the contract | Training, or the tool is wrong for the job |
Utilization is a money question. Adoption is a product question. Measure utilization first, because it pays for itself and does not need anyone's cooperation.
Which mistakes make the number wrong?
Five, and each one pushes the rate in a predictable direction.
- Counting provisioned accounts. Pushes the rate up, towards a useless 100 percent.
- Mixing windows. A 90 day window on one tool and 30 on another makes the first look healthier than it is.
- Taking seats from the admin console. The console shows what is configured, the contract shows what you pay for. Use the contract.
- Blending every tool into one figure. A stack averaging 88 percent can contain one tool at 40 percent, which is the one worth your afternoon.
- Measuring after the notice deadline. The number is then accurate and useless for another year.
What do you do with a low utilization rate?
In this order, because each step changes the next number.
- Reclaim. Remove seats from leavers and from people who never signed in. This costs nothing and usually moves the rate several points.
- Reassign. Give reclaimed seats to people waiting on a request instead of buying new ones.
- Downgrade. Some users need a cheaper tier, not a full seat.
- Renegotiate. Take the trend to the renewal and reduce the committed seat count. This only works before the notice deadline, which is why the deadline belongs next to the number.
Steps one and two are available today. Step four is available once a year, and only if you did not miss the window. See what to log for a renewal reminder and what your agreement allows on true-down.
How often should you measure?
Monthly for tools above a cost threshold you set, quarterly for the rest, and always once more in the month before a renewal notice deadline. Measuring more often than you can act is busywork; measuring once a year means finding out after the contract renewed itself.
Frequently asked questions
What is a good license utilization rate?
Most teams treat 85 to 95 percent as healthy. Above 95 percent, check whether people are sharing logins or waiting for seats. Below 70 percent on a renewing contract, reclaim seats and reduce the committed count at renewal.
Should I count provisioned accounts as active users?
No. A provisioned account only proves someone created it. Counting provisioned accounts returns close to 100 percent for every tool and hides exactly the waste you are looking for. Use sign-in as the minimum definition.
What window should I use?
Thirty days is the common choice because most vendors report it and most work cycles fit inside it. The window matters less than applying the same one to every tool, every month.
Does low utilization always mean I should cut seats?
No. Check for a rollout still in progress, a seasonal team, or a tool bought for a project that ended. The first two need time, the third needs cancelling rather than resizing.
Can I measure utilization without vendor admin access?
Partly. Without vendor usage data you can still compare paid seats against assigned people in your own records, which catches leavers and unassigned seats. It will not catch someone who has a seat and never opens the tool.
Read the rate instead of rebuilding it.
The demo workspace holds 24 licences with their paid seat counts, the people assigned and the renewal date next to each one.
Try the live demo →