The final question is not whether the change launched. The final question is whether it took.

Did behavior change? Did the old workflow lose power? Did managers reinforce the new standard? Did tools, incentives, metrics, and decision rights support the shift? Did shadow systems disappear, become official, or reveal design gaps? Did adoption survive pressure after the kickoff energy faded?

A change that cannot pass those questions is not complete. It is announced.

Why change decays

Change decays because the old system is patient.

The old system has habit, shortcuts, relationships, incentives, dashboards, templates, customer expectations, political bargains, and muscle memory. It does not need a launch plan. It already lives in the company.

The new system requires reinforcement until it becomes normal. If reinforcement stops too early, the company slides back toward the path with less friction.

Decay is especially likely when:

  • leadership attention moves on;
  • managers stop inspecting the new behavior;
  • the old tool remains available;
  • metrics still reward the old behavior;
  • exceptions are not reviewed;
  • early friction is not fixed;
  • new employees learn the workaround before the official process;
  • executives violate the change when urgent work appears.

The audit exists to catch this before the organization concludes that change is impossible.

When not to change

Good operators also know when not to change.

Do not launch a change when the behavior is unclear, the reason is weak, the company is already overloaded, the leadership team is not aligned on tradeoffs, managers cannot reinforce it, incentives contradict it, the transition risk is unmanaged, or the current system is ugly but functioning better than the proposed replacement.

Do not change because a new executive needs a signature move. Do not change because a tool was purchased. Do not change because the company is anxious and wants the feeling of motion.

Change consumes trust. Spend it on things that matter. The discipline to not change is part of the discipline to make important changes stick.

The adoption scorecard

Audit the change across ten dimensions.

1. Behavior. Can we name the old behavior and the new behavior? Is the new behavior observable in normal work?

2. Workflow. Does the new behavior fit the actual workflow, or does it require people to leave the flow of work to satisfy the change?

3. Incentives. Are goals, metrics, compensation, recognition, promotion, and executive attention aligned with the new behavior?

4. Tools. Do the tools make the new behavior easier, or do they create reporting theater while real work happens elsewhere?

5. Managers. Can managers explain, inspect, reinforce, correct, and escalate the change?

6. Decision rights. Is authority clear enough that people know who can decide, approve exceptions, or close the old path?

7. Transition design. Are dual-running, migration windows, fallback rules, and exception handling explicit?

8. Shadow systems. What unofficial workflows remain, and what do they reveal?

9. Metrics. Are we measuring adoption and outcomes, not just launch activity?

10. Ownership. Who owns the change after launch, including decay, fixes, exceptions, and reinforcement?

Score each dimension red, yellow, or green. Red means adoption is structurally at risk. Yellow means the change may work locally but is vulnerable under pressure. Green means the operating environment supports the desired behavior.

Then force a decision. Red dimensions require redesign, delay, or a narrower rollout. Yellow dimensions require named mitigations and a review date. Green dimensions can be scaled, but only if the supporting conditions remain true in the next team, region, customer segment, or operating context.

The executive review

The best adoption reviews are direct.

Ask: where is the old behavior still winning? Where is the new behavior working without reminders? Which managers are landing it? Which managers are not? Which exceptions are legitimate? Which exceptions are avoidance? Which metric changed? Which metric did not? What friction should we fix? What old artifact should we remove? What should we stop pretending has changed?

This is not blame. It is reality contact.

The operator's standard

A serious company does not judge change by the quality of the announcement. It judges change by repeated behavior under normal operating pressure.

If the change matters, design the adoption system. If you cannot design the adoption system, do not pretend the launch will do it for you. And if the audit shows the current system is better than the proposed one, have the maturity to stop, learn, and preserve trust for the next change.

Change takes when the new way becomes the easier, clearer, reinforced, and rewarded way to work.