Change management
What is user adoption?
User adoption is the extent to which the people a system was bought for do their work in it, in the way it was designed. A login proves very little. What counts is whether the work has moved.
Updated
Two kinds of user adoption
The phrase is used for two different problems. This page is about the first.
- An organisation rolls out a system to its own staff
- People are told to use it, and can still work around it
- The question is whether the work has moved into the system
- Owned by change, training and IT teams
- A software company wants customers to take up its product or a new feature
- Customers who do not see the value cancel
- Measured through activation, feature use and retention
- Owned by product and customer success teams
The formulas look alike, which is why the two get mixed up. The causes of failure differ. An employee cannot cancel the new purchasing system, but she can keep emailing her orders to a friendly buyer for years.
What adoption means beyond logins
Prosci, the firm behind the ADKAR model, breaks the people side of a project's return into three factors. They make a good definition of adoption because each can be measured separately.
How quickly do people start working the new way after go-live?
How many of the people affected end up using it?
How well do they perform, against what the design expected?
Prosci's Tim Creasey adds a caution: "The three factors are universal, but their measures aren't." A login count within a week of go-live is a fair signal of speed. It tells you nothing about how many purchase orders went through the system or how many had to be corrected. You have to choose a measure for each factor, for your own system and your own tasks.
How to calculate user adoption rate
Adoption rate = people doing the task in the system ÷ people who should be doing it × 100
Both halves of that sum need care. The top number should count people who completed the task, not people who logged in. The bottom number should count the people whose job includes the task, which is usually far fewer than everyone with a licence. The IRIS glossary makes the same point about the denominator: use the realistic target audience.
Say your company replaced emailed purchase requests with a purchasing system three months ago. IT issued 600 licences. Of those 600 people, 150 raise purchase requests as a regular part of their job. Last month:
- 138 of the 150 logged in. That is 92%.
- 96 of the 150 raised at least one requisition in the system. That is 64%.
- Purchasing placed 1,240 orders. 810 began as a requisition in the system, which is 65%. The other 430 arrived by email.
- 22% of requisitions were sent back for correction.
Report the 92% and the project looks finished. Divide the 138 logins by all 600 licences and you get 23%, which looks like a disaster. Neither figure describes what happened. The honest summary is that 64% of the people and 65% of the orders have moved, and about one requisition in five needs rework.
What is a good user adoption rate?
There is no trustworthy benchmark. The IRIS glossary says that 80% or more in the first month is generally considered positive for software that staff are required to use, but it gives no source and says itself that no universal benchmark exists. We could not find an independent study worth citing, so this page offers no number.
A better target comes from the process. If the system is meant to be the only way to raise a purchase order, the target is every order, and the useful figures are how far short you are and which teams account for the gap. Set the target per task, with a date, and agree it before go-live.
User adoption metrics: which to track and which mislead
| Metric | Verdict | Why |
|---|---|---|
| Logins or active users | Misleads alone | People log in to look around, then send the request by email anyway. Useful only as an early sign in the first week. |
| Licences assigned | Misleads | It counts what IT did. |
| Training completions | Misleads | It shows who attended. The Kirkpatrick model shows how far attendance is from a change in what people do. |
| Share of the target group doing the task in the system | Track | This is the adoption rate. Report it per task and per team. |
| Share of transactions going through the system | Track | It catches the side routes, such as the 430 emailed orders above. |
| Rework or error rate | Track | The proficiency measure. High adoption with high rework moves the cost to whoever fixes the errors. |
| Help desk tickets per task, per week | Track | They should rise at go-live and then fall. A team that has raised none may not be using the system. |
| Time to complete the task | With care | Compare it with what the design expected, and expect the first weeks to be slow. |
| Satisfaction surveys | With care | Good for finding causes. A happy user who still emails orders has not adopted anything. |
The two vendor lists we read, BambooHR's six KPIs for HR systems and WalkMe's eight adoption metrics, cover most of the rows above. Neither gives a formula or says who belongs in the denominator.
The share of each role that has completed its key tasks unaided in a practical check is the only adoption figure you can have before the system is live. It shows who cannot yet do the task, while there is still time to do something about it.
Why adoption stalls
Each cause leaves a different pattern in the numbers, so check them before you book more training.
- The old route is still open. Purchasing still accepts emails, so logins are high and transactions low. Close the mailbox. More training will not help.
- People were shown the task and never did it. Adoption is reasonable and rework is high. This is the gap between Knowledge and Ability in ADKAR, and the cure is practice.
- The new way is slower for the person doing it. A two-line email became a form with 12 fields, and the benefit lands in finance. Shorten the form if you can. If you cannot, the requester's manager has to insist.
- Managers are not using it. Approvals sit for a week, and requesters learn that the system is the slow way to get something bought.
- Support ended too soon. Tickets drop to zero in week three because the floor-walkers went back to their jobs and nobody knows who to ask.
- You are counting the wrong people. The 450 occasional users drag the rate down and hide the fact that the core group is fine.
The ADKAR page has a table for diagnosing which of its five elements a team is stuck on. Change management covers sponsorship and communication, which sit behind the manager problem.
A user adoption strategy in four parts
A strategy for adoption has to cover the time before go-live as well as after it, and each part should move a metric you have chosen to track.
- Training on tasksTeach each role the 10 to 20 tasks it will do, close to go-live. The method is in how to train employees on new software, and the planning document is the software training plan template.
- Practice before it countsPeople should have completed each key task themselves before go-live, in a software simulation or a training environment. This is the part that moves the rework rate.
- Support in the first weeksFloor-walkers for the first fortnight, one place to ask, and guides within reach. A digital adoption platform adds prompts inside the live system, which helps most with rare tasks and with changes after go-live.
- Measurement that leads to actionReview the adoption rate per task every week for the first two months, by team. Agree in advance what happens at each threshold, including the date the old route closes.
A dashboard nobody acts on changes nothing. Decide who sees the numbers and what they are expected to do when a team sits at 40%.
A user adoption plan on one page
This is the purchasing example written as a plan. The targets are illustrations. Set your own from the process.
| Task | Who should do it | Measure | Target | Source of the number | Owner |
|---|---|---|---|---|---|
| Raise a requisition | 150 regular requesters | Share who raised at least one in the month | 90% by month 3 | System report | Purchasing manager |
| Place orders from requisitions | Purchasing team | Share of orders that began in the system | 95% by month 4, when the email route closes | System report, plus a count of the purchasing mailbox | Head of procurement |
| Get the requisition right first time | 150 regular requesters | Share sent back for correction | Under 10% | System report | Training lead |
| Approve a requisition | 40 approving managers | Share approved within two working days | 90% | System report | Finance director |
Add a review date to each row and the plan is complete. For a fuller treatment of measuring whether the training itself did its job, see how to measure training effectiveness. End user training compares the methods you can put behind the first two parts of the strategy.
Practice, and a readiness number, before go-live
Recourse covers the second part of the strategy. It turns a screen recording of a task into a Demo to watch, a Practice mode on a replica of the screen with hints that escalate, and an Evaluate mode that scores an unaided run against a pass mark. The scores give you the pre-go-live figure described above. Recourse does not measure usage in your live system, so the adoption rate itself has to come from the system's own reports. It is not a digital adoption platform and puts no guidance on live screens. It is not an LMS, and it exports SCORM 1.2 only.
Start freeQuestions people ask
What is user adoption?
How do you calculate user adoption rate?
What is a good user adoption rate?
What is the difference between user adoption and engagement?
How do you measure user adoption?
How do you increase user adoption?
Sources
- Prosci, Tim Creasey: 3 People-Side Factors That Define or Constrain Project ROI, for speed of adoption, ultimate utilisation and proficiency, and the quotation.
- IRIS glossary: Adoption rate, for the formula, the advice on the denominator and the 80% figure.
- BambooHR, on measuring HRIS adoption, as an example of a vendor KPI list.
- WalkMe, 8 user adoption metrics, as a second example.
Keep reading
Last reviewed 21 November 2026. Recourse publishes these guides for learning, customer education and change teams. Details drawn from public sources can change.