← Back to Knowledge Hub

How to measure change adoption when the metrics are hard to pin down, and what leading indicators actually predict success

Most organisations measure change adoption by counting system logins and training completions. These metrics tell you what happened, not what will happen. The real skill in adoption measurement is knowing which indicators predict success before it arrives, and which only confirm it after the fact. This guide covers both: how to measure adoption when the metrics feel elusive, and which leading indicators genuinely predict whether a change will stick.

Why Adoption Metrics Are Hard to Pin Down

Measuring change adoption is harder than measuring project delivery because adoption is a human phenomenon, not a technical one. A system can be live without being used. A process can be documented without being followed. Training can be completed without being applied. The gap between technical completion and genuine adoption is where most measurement frameworks fail.

The core challenge is that adoption lives in the space between what people do and why they do it. A login count tells you someone accessed the system. It does not tell you whether they completed their work there, whether they reverted to a spreadsheet afterwards, or whether they will log in again tomorrow. Adoption measurement requires you to look beyond activity data and into behaviour, capability, and intent.

The following are the six most common situations where adoption metrics feel impossible to define. Each one has a practical approach. Click any challenge to see how to measure what feels unmeasurable.

When the change is cultural, not procedural
When you cannot isolate the change from other variables
When the data does not exist yet
When people game the metrics
When leadership wants a single number
When the change affects different groups differently

Leading or Lagging? Classify the Metrics

The most important distinction in adoption measurement is between leading indicators, which predict whether adoption will happen, and lagging indicators, which confirm whether it did. Most organisations over-index on lagging indicators and discover problems too late to fix them. Test your understanding: classify each of the 15 metrics below as leading or lagging, then reveal the correct answers.

Number of people who completed training
System login rates in the first 30 days
Manager confidence scores from pulse surveys
Reduction in errors or rework after go-live
Employee sentiment towards the change
Percentage of processes completed using the new method
Number of support tickets or help requests
Business outcome improvement (revenue, cost, speed)
Attendance at change champion briefings
Proficiency assessment scores at 90 days
Number of workarounds or shadow processes detected
Leadership communication frequency about the change
Stakeholder readiness assessment scores
Customer satisfaction scores post-change
Percentage of managers who can articulate the change rationale
Classify all 15 metrics to reveal answers (0 of 15 done)

The Adoption Measurement Curve

Adoption is not a single event. It is a progression through five stages, and each stage requires different metrics. Most organisations only measure the middle: they track whether people are using the system. But adoption starts before usage and extends beyond it. Click any stage to see what to measure, how to measure it, and what signal tells you it is time to move on.

Building Your Adoption Measurement Framework

A strong measurement framework is not a list of metrics. It is a system that connects leading indicators to lagging outcomes, covers all stages of the adoption curve, and provides actionable signals at every point. The following principles should guide its design.

Start with the outcome and work backwards

Define what success looks like in business terms. Then identify the behaviours that would produce that outcome. Then identify the conditions that would produce those behaviours. Your leading indicators are the conditions. Your lagging indicators are the outcomes. The behaviours are what you are trying to change.

Measure at the segment level, not the aggregate

Organisation-wide averages hide the variation that matters. A team at 95% adoption and a team at 15% adoption produce a meaningless 55% average. Segment your metrics by role, function, geography, and tenure. The segments with the lowest scores are where your attention should be focused.

Balance quantitative data with qualitative insight

Dashboards tell you what is happening. Conversations tell you why. A declining usage trend is data. Understanding that usage is declining because the new system does not handle a critical edge case is insight. Build structured qualitative collection into your framework, not as an afterthought but as a primary data source.

Define thresholds and response plans in advance

Every metric should have a threshold that triggers action. If manager readiness drops below 60%, what do you do? If first-use drop-off exceeds 40%, what is the response? Defining these thresholds before you need them prevents the paralysis that occurs when data reveals problems that nobody has a plan for.

Measure frequently enough to act

A monthly adoption dashboard is a historical report, not a management tool. If your rollout is happening in weekly waves, you need weekly or even daily measurement of leading indicators. The frequency of measurement should match the speed at which you can intervene. Data that arrives too late to act on is not measurement. It is post-mortem.

Retire metrics that have served their purpose

Not every metric is relevant at every stage. Awareness metrics matter in the first weeks. By month three, they should be replaced by adoption and proficiency metrics. A measurement framework that never changes accumulates noise and dilutes attention. Review and retire metrics quarterly.

Is Your Measurement Framework Fit for Purpose?

Use this checklist to assess whether your current adoption measurement approach will give you the insight you need to intervene early and sustain results long-term.

0 of 10 complete

This topic is part of Execution, the fourth pillar of the TCA Change Model.

Explore the Full Model