Measuring Change Success

A change goes live, and the project dashboard turns green: “90% of staff have logged into the new system.” Champagne. Except logging in once tells you almost nothing about whether the change worked. Measuring change is where a lot of well-run projects quietly fool themselves, because they count the easy thing instead of the meaningful thing. This module is about measuring what actually matters, and being patient enough to let the real results show up.

Adoption, usage, proficiency

There are three different things people lazily call “success,” and keeping them apart is half the battle.

Adoption is whether people have started using the new thing at all. Did they log in, create an account, run it once? It’s the lowest bar, and it’s the one most often reported, precisely because it’s easy to measure and looks good.

Usage is whether they keep using it, regularly, as part of real work. One login is adoption; logging in every day and actually entering data is usage. Usage is where a lot of changes fall apart — adoption spikes at launch and then quietly collapses as people drift back to spreadsheets.

Proficiency is whether people use it well — using the features that create the value, not just poking at the surface. A team can adopt and use a new tool while ignoring exactly the capability that justified buying it. Proficiency is the hardest to measure and the closest to actual benefit. The trap is celebrating adoption and assuming proficiency followed. It rarely does on its own.

Leading vs lagging indicators

To track change in time, split your measures into two kinds. A leading indicator is an early signal that predicts where you’re heading — login rates, training completion, support tickets, pulse-survey sentiment. It moves quickly and lets you steer while there’s still time. A lagging indicator is the actual outcome you cared about — cost saved, cycle time reduced, errors down, revenue up. It’s the real prize, but it shows up late and slowly.

You need both. Lagging indicators alone are like driving by looking only in the rear-view mirror: by the time the number tells you the change failed, it’s too late to fix. Leading indicators let you react early — if usage is sliding in week three, you intervene before the lagging benefit ever fails to appear.

Surveys, used honestly

Pulse surveys are a useful leading indicator if you ask real questions and act on the answers. “How confident do you feel using the new process?” tells you about proficiency and desire long before the cost savings show. But surveys curdle fast if people answer honestly and nothing changes. A survey nobody acts on trains people to stop telling you the truth.

When the benefits actually show

Here’s the expectation to set early: things often get worse before they get better. This dip — productivity falling right after launch as people fumble with the new way — is normal and predictable. Benefits arrive on the far side of it, weeks or months later, once usage becomes proficiency. Judging a change by its first fortnight guarantees you’ll panic at exactly the wrong moment.

Rule of thumb: measure usage and proficiency, not just adoption — and expect a dip before the upside. Logging in once is the start of the story, not the end of it.

Spot it: adoption, usage, or proficiency?

Read each situation and decide for yourself, then tap a card to flip it and check your answer.

Sort the measures

Drag each item into the bucket it belongs to — or tap an item, then tap a bucket. Hit Check placement when you’re done.

How to use it

When someone reports a change as a success, ask which thing they measured. Useful phrases: “Is that adoption, or are people actually using it day to day?” “What’s our leading indicator telling us, not just the lagging one?” “Are we past the expected dip yet, or still in it?” When a survey comes round, push for action: “If people flag this, what will we actually do about it?” Knowing the difference between someone logging in and someone using a tool well makes you the person who can tell whether a change really worked, instead of the person clapping at a green dashboard.