Days 1–15: validate and select priorities
Review the findings with the people who do the work. Confirm which observations are supported by records and which are hypotheses. Deal with urgent safety or compliance concerns through the appropriate responsible professionals rather than waiting for a planning cycle.
For the remaining work, compare operational impact, effort, dependencies and the cost of delay. Choose a small number of priorities that the team has capacity to own. Keep other findings in a visible backlog instead of starting them all.
Days 16–30: define the first change
For each priority, write the problem, current baseline, action owner, available resources and a practical test for success. Agree what the owner can decide and which matters need escalation.
Choose one team or workflow for the pilot. For a delayed handoff, the first change may be a shared checklist and a named receiving person. Identify the information that must be complete before work moves, then record how often the handoff needs correction today.
- Problem and evidence
- Owner and decision authority
- Action, dependency and due date
- Baseline and measurement definition
- Review date and stop conditions
Days 31–60: run the pilot and remove obstacles
Review progress at a short weekly meeting. Ask what changed in the work, where the team struggled and which decision is needed next. Record the answer rather than collecting a long list of status updates.
If the pilot does not work, distinguish poor execution from a poor assumption. A checklist may fail because the required information is unavailable, not because staff need another reminder. Change the design when evidence shows it is unsuitable.
Days 61–75: compare evidence before expanding
Compare the pilot with the baseline using the same definitions. Note differences in workload, staffing and customer mix. Ask the team whether the improvement can be maintained during a busy period.
Check for unintended effects. Faster approval is not progress if errors increase or responsibility becomes unclear. Decide to adopt, revise or stop. Expand only when the people receiving the change understand it and have the resources to use it.
Days 76–90: hand over and choose the next cycle
Update the operating instructions, name the person who maintains them and agree a review date. Make the current version easy to find and archive outdated copies. Include a way to record exceptions and propose changes.
Close the cycle with a short record of what was tried, what the evidence showed and what remains unresolved. Choose the next priority from the backlog using what you now know. Do not call the entire audit implemented just because the first set of actions is complete.
Treat the 90 days as a learning and implementation cycle. A few changes that the team can sustain are more useful than a long action list with no owners.
