Learning experience 03
Measure what matters
Meridian was measuring. That was the problem. Completion rates were at 94%. Satisfaction scores averaged 4.2 out of 5. Leadership considered the knowledge system a success. Meanwhile, new hires were still taking 14 weeks to become effective and managers were still answering the same questions every week.
The system wasn't broken. The measurement was. This is how I redesigned it.
What Meridian tracked before Source. Everything looks healthy. That's the problem.
These metrics measure activity, not behavior. None of them answer the only question that matters: did anyone make a better decision because of this system?
What I instrumented during the Source rollout. Behavior signals, not vanity metrics. Tracked weekly across all regions.
Repeat searches on the same topic is the most diagnostic signal — it tells you where the content isn't actually answering the question. It drove two rounds of content revision during rollout.
Lagging indicators measured at 6 months post-launch. Connecting Source adoption to actual performance change.
The measurement framework itself became a governance tool — the adoption indicators in phase 2 are now reviewed monthly by the L&D and operations teams to catch content gaps before they become performance gaps.
The framework
Activity metrics tell you what happened. Adoption metrics tell you if the system is working. Impact metrics tell you if it mattered. You need all three — in sequence, not simultaneously. Trying to measure impact before adoption is like checking exam scores before anyone studied.