How can you measure the success of change management?

Start with a sentence about the difference you expect in day-to-day work. In a fictional business, complaints should reach the responsible department without unclear forwarding. The number of additional queries needed is then one possible process measure. Whether employees know the right route is a complementary question about their experience.

Decide who can make a decision based on the result. A metric without someone responsible produces reports, but no reliable next step. Also specify when a decision is needed: for example, after trialling a new handover rule or before commissioning more training.

In its approach to measurement, Prosci distinguishes between delivering change activities, adopting the change and organisational results. This does not mean that all levels can be measured through a survey. You can find the original article at Metrics for Measuring Change Management. The example below is our own practical application.

First separate three questions. Was the action carried out? Does the changed work step function? How do those affected experience the conditions needed for it? This distinction prevents high workshop attendance from accidentally being used as evidence of a better process.

The information does not have to be held in the same system. An existing process analysis, documentation of your action and a team survey can be considered side by side. But check that they cover the same period and scope. A company-wide figure does not automatically fit a change within a single team.

Look at the situation before an action, where suitable information is available. Document the period, scope and known special circumstances. For complaints, an unusually large order or a change in product group may affect comparability, for example. These events belong in the interpretation, rather than in a footnote added afterwards.

If there is no baseline, start with an initial observation and acknowledge the gap. Do not invent a previous situation from memory. Initial feedback can still help you choose an action. It simply does not yet allow a reliable before-and-after comparison.

Define a process measure so that it can be collected in the same way later. What counts as a follow-up query? Which cases are included? Who records them? Without this rule, an apparent decrease may result from a different counting method. This precision is often more important than another chart.

Ask about conditions you can respond to. Our own example statements are: “I am clear about the step needed when a handover is incomplete”, “I receive support when I have difficulties with the new process” or “Enough time is available to try out the new way of working”.

Each statement should address one subject. “The new collaboration is clear, simple and efficient” mixes several assessments. A low value leaves it unclear whether the criticism concerns the explanation, execution or result. Choose a narrower statement that fits the next decision.

A scale from one to five can organise agreement if its endpoints are clearly described. The examples here are not validated measurement instruments. An average of, for example, 3.2 is also neither a percentage nor a statement about the share of the team that can carry out the process. It describes the responses to this specific statement.

Record the assumption before you start: a clarified exception rule is intended to improve people's sense of direction and reduce avoidable follow-up queries. Note when it was introduced and whether those affected could actually use it. Then check the same information sources, keeping collection as consistent as possible.

For repeated questions, compare the same wording and the same question version. Rewording can change responses even when the work has not changed. Look beyond the average to the distribution and participation, where available. A small difference may not be practically meaningful.

A more favourable trend after an action is an indication, but not yet proof of causation. Other events may have contributed. If follow-up queries fall and clarity is rated more positively, test the explanation against typical cases. If the two sources diverge, investigate that difference rather than picking the figure that fits.

For a financial decision, you need the cost and effort of the action and a clear connection to the operational goal. The time spent on a case exercise may be known internally. Whether it actually saved processing time must be checked through suitable data collection. Better mood cannot be converted directly into euros.

A lack of effect can also be useful. Perhaps the exception rule was not the decisive obstacle. You can then reconsider another similar action and first examine technical access or capacity. The benefit here comes from a better-informed decision, rather than automatically calculated savings.

For a transformation programme, first assign measures to the relevant work step. A single overall value can hide opposing trends. Putting transformation into practice in day-to-day work explores how to consider multiple initiatives and dependencies across departments.

A short entry may be enough: goal, work step being considered, workflow metric, statement rated by the team, action, responsible person and next review point. Add below what you infer from the information and what remains open. This connects measurement back to the decision.

Do not let the fields become a documentation requirement without benefit. If a measure does not influence a decision, consider whether you need it at all. If a specific work case can resolve an open question, you may not need a new series of surveys.

The article on the change process provides the organisational framework. For longer-term shared patterns, Measuring company culture offers the relevant further reading. What you want to examine determines the measurement, rather than the availability of a metric.

End each review with a sentence stating the decision. For example: “We will keep the exception rule and examine the remaining bottleneck in data access.” Name who is responsible and the next point for observation. This makes even a mixed result manageable.

Also record an alternative explanation. Perhaps the cases became easier, or another team provided temporary support. You do not have to rule out every conceivable cause. Initially, it is enough to avoid overlooking a plausible alternative and investigate more closely when making an important investment decision. A short, honest report is more useful than an apparently conclusive success figure.

EngageBeat Fokus shows team feedback and responses to the same question over time. Questions rotate, so a selected question will not automatically appear the following week. Results are available after the week closes, with at least five valid responses per question. You add operational process information outside the app. The Change Management and Transformation pages show the use cases; EngageBeat does not provide automatic proof of ROI.