How to build a 90-day operational improvement plan

A 90-day improvement plan should make execution more focused. It should not become a catalogue of every problem the business has noticed.

Choose a business outcome, identify the flow that produces it and select a small number of changes the team can realistically deliver. The three phases below are a practical planning pattern, not a guarantee of results within a fixed period.

Days 1–30: establish the diagnosis

Start with evidence from the work. Follow an order, request or project from start to completion. Record waiting, rework and unclear handoffs. Speak to the people responsible for each step.

Define the baseline using available records. Where information is missing, run a short observation period and label estimates honestly. A target without a baseline makes progress hard to interpret.

Use the operational bottleneck guide to test where the flow is constrained. Confirm the team agrees on what is complete and valuable, rather than simply counting activity.

Choose priorities with owners

Select a limited set of interventions. For each one, write the problem, expected effect, responsible owner, required resources and decision date.

An owner needs the authority to make the change or a clearly named escalation path. Responsibility without decision rights tends to leave the plan dependent on repeated informal approvals.

Identify what will be paused to create capacity. A plan added on top of a fully occupied team is unlikely to receive the attention assumed in its timeline.

Days 31–60: test the changes

Run bounded experiments before broad implementation. A clearer input checklist, a revised approval threshold or a dedicated review window can test a diagnosis without redesigning every process.

Measure the intended outcome and a balancing measure. For example, a faster response is not an improvement if incomplete answers create more work for another team.

Keep a short decision log. Record what changed, what happened and what the team learned. If a test fails, distinguish a weak diagnosis from poor execution before choosing the next step.

Keep the weekly review practical

Review a small number of measures and decisions. Ask what has moved, what is blocked and what needs leadership attention.

Avoid using the meeting to repeat every task in the plan. Focus on the assumptions that might invalidate the next action. Make decisions visible so that teams do not wait for the same clarification repeatedly.

If the intervention reveals a technology gap, define the workflow requirement before opening a vendor selection. Use a procurement scorecard to preserve consistent evaluation.

Days 61–90: standardize and reassess

For changes that work, document ownership, required inputs and normal review routines. Train the people who need to use the process and remove conflicting instructions.

Compare the result with the baseline and note remaining uncertainty. Do not claim a lasting improvement from an unusually quiet week or a temporarily elevated workload.

Reassess the constraint. A successful change can move the bottleneck, which means the next quarter should not automatically repeat the same priorities.

Finish with a decision, not a presentation

At day 90, decide which practices to retain, which experiments to stop and which larger commitments need evidence or investment. Assign the next operating owner.

The value of the plan is a clearer way of delivering work and making decisions. A small number of changes that become part of normal operations is more useful than a long list of initiatives marked as launched.