Most change programmes report on the plan. Milestones met, training sessions delivered, system switched on, budget on track. The status report goes green, and leaders reasonably assume the change is working. Then, six months later, the old spreadsheets are still in use, the new process has three unofficial workarounds, and the promised benefits haven't arrived.
This guide covers what to measure during a change, how to measure readiness before you start, how to track adoption while it happens, and how to tell whether the change will last. It is part of our wider guide to change leadership.
What should you measure during a change?
Measure change at three levels. Prosci describes them as organisational performance, which asks whether the change delivered the results expected; individual performance, which asks whether people adopted the new way and use it well; and change management performance, which asks how well the change itself was led (Horlick 2024).
| Level | The question it answers | Example measures |
|---|---|---|
| Organisation | Did the change deliver what it promised? | Cost, cycle time, error rates, customer complaints, revenue, safety incidents |
| People | Have people adopted the new way, and do they use it well? | Usage, proficiency, speed of adoption, workarounds still in use |
| Leadership of the change | Did people understand it and feel supported? | Understanding of the reasons, confidence, trust in leaders, readiness scores |
In our experience, most organisations measure the first level, because the business case demands it, and the third as a set of completed activities. The middle level is where changes succeed or fail, and it is the one most often left unmeasured.
Does measuring change make a difference?
The evidence suggests it does. In Prosci's 12th edition benchmarking research, 63% of participants said they measured compliance and overall performance during their change. Of those, 76% met or exceeded their project objectives, compared with 24% of those who didn't measure (Horlick 2024).
That is an association, and part of it will reflect better-run organisations doing many things well at once. Measurement still earns its place, because it is how leaders find out early that something isn't working. The same research found the main reason organisations didn't define success was a lack of agreement on goals and objectives, cited by 40% of participants (Horlick 2024). Agreeing what success looks like is often the first useful conversation a leadership team has about a change.
What is the difference between leading and lagging indicators?
Leading indicators show whether a change is on track to work, early enough for you to act. Lagging indicators show whether it worked, after the fact. You need both.
| Leading indicators (early warning) | Lagging indicators (results) |
|---|---|
| Do people understand why the change is happening? | Are the business results the change promised showing up? |
| Do they believe it will work, and that leaders back it? | Has the old way stopped, or are workarounds still in use? |
| Are people trying the new process or system? | Have error rates, cycle times or costs moved? |
| What are people raising in team meetings and one-on-ones? | Has turnover or absence changed in the affected teams? |
| Are managers confident explaining the change? | Are customers noticing a difference? |
Leading indicators are softer and harder to measure, so they are often skipped. They are also the only ones that give you time to change course.
How do you measure readiness before a change?
Readiness can be measured before launch, and it is worth doing. Armenakis and Harris (2009) summarise readiness as resting on five beliefs: that change is needed, that this change is the right one, that the organisation can do it, that leaders support it, and that it will benefit the people being asked to change. Holt et al. (2007) developed and tested a survey scale built on four of these: confidence that people can make the change, belief that the change is appropriate, belief that management supports it, and belief that it will benefit them personally.
A simpler test comes from Beckhard and Harris (1987). Change happens when dissatisfaction with the current state, a clear vision of the future and credible first steps together outweigh the resistance to change. The three drivers multiply, so if any one is close to zero, the change stalls however strong the others are. Leaders can turn that into three questions for any group affected:
- Do people feel the cost of staying where we are?
- Can they describe where we're going?
- Can they see practical first steps they believe in?
A weak answer to any of the three tells you where to put your effort before launch. Our own Change Score Card builds on the same idea, using short anonymous surveys to track these drivers, and resistance, as a change progresses.
How do you measure adoption during a change?
Adoption is about behaviour: whether people are doing things the new way, and how well. Prosci breaks it into three measures: how quickly people adopt the change, how many end up using it, and how proficient they become (Horlick 2024). In practice, that means a mix of:
- Usage data. System logins, transactions processed the new way, forms completed in the new format.
- Workarounds. Whether the old spreadsheet, form or shortcut is still in use. Ask the people doing the work, because they will know.
- Quality. Error rates, rework and complaints in the new process compared with the old.
- Pulse surveys. A few short questions on understanding, confidence and support, repeated every four to six weeks.
- Conversations. Managers and frontline leaders hear things long before they show up in data. Ask them regularly.
It also helps to know why people are going along with a change. Herscovitch and Meyer (2002) distinguished between people who support a change because they believe in it, because they feel they ought to, and because they feel they have no choice. People who believed in the change or felt obliged to support it showed more support than those who simply felt they had no alternative. Compliance can look like adoption in the usage data while hiding people who will drop the new way the moment pressure eases.
Watch the load as well. Researchers have built a validated measure of change fatigue, and found that fatigue wears people down and weakens their commitment to the organisation (Bernerth, Walker and Harris 2011). If several changes are hitting the same teams, measure that too. Our guide to change fatigue covers how.
How do you know a change is holding?
A change has held when the old way has stopped, the new way has become normal, and it survives pressure: a busy month, a new manager, a staff shortage. That takes longer than most project timelines allow.
Research on individual habits gives a sense of scale. In one study, it took an average of 66 days for a new daily behaviour to become automatic, with individual times ranging from 18 days to a predicted 254 days (Lally et al. 2010). Organisational change involves many people, systems and incentives shifting together, so it usually takes longer still. Kotter (1995) lists declaring victory too soon among the classic errors in leading change.
A few practical checks:
- Measure at three, six and twelve months after launch, well beyond the go-live date.
- Look for the old way. If the workarounds are creeping back, the change hasn't held.
- Test it under pressure. Check what happens in the busiest period, or after a key person leaves.
- Check the incentives. If targets, rewards or promotions still favour the old way, the new way won't last.
- Keep reinforcing. Recognise people using the new way well, and keep fixing the things that make it hard.
What are the common mistakes in measuring change?
- Measuring activity instead of behaviour. Training attendance and emails sent show effort. They don't show change.
- Trusting the green status report. A plan can be on schedule while people work around it. Ask the people doing the work.
- Measuring only at the end. By the time lagging indicators move, it is often too late to change course.
- Surveying without acting. People stop answering surveys that lead nowhere. Share what you heard and what you changed.
- Surveying too often. A short pulse every four to six weeks is usually enough during the busiest period.
- Claiming every result for the change. Other things move the numbers too. Be honest about what the change did and didn't cause.
How do you build a simple measurement plan?
A measurement plan doesn't need to be elaborate. For most changes, one page is enough:
- Agree what success looks like. Write down the two or three results the change is meant to deliver, and how you will know.
- Pick a few adoption measures. Choose what you will track to see whether people are using the new way, and how well.
- Measure readiness before launch. Use a short survey or the three Beckhard and Harris questions with each affected group.
- Set a rhythm. Decide when you will check each measure, and who will look at the results.
- Name an owner for each measure. Someone who will act when it moves the wrong way.
- Keep going after go-live. Plan checks at three, six and twelve months.
- Close the loop. Tell people what the measures show and what you are doing about it.
If you want an independent read on how a change is tracking, talk to us about our change management work. Our guide to the leader's role in change covers what leaders can do when the measures show a problem, and our guide to resistance to change covers what is usually going on underneath the numbers.