The PICK model: compare improvement ideas

Several improvement ideas to choose between? Use this page to compare their expected impact and effort, then choose a manageable next step.

PICK sorts ideas by expected impact and effort: Possible, Implement, Challenge and Kill. It makes assumptions visible; it does not prove an idea will work or replace judgement about risk and obligations.

Manage competing commitments with Focus and prioritisation →

The four categories

“High” and “low” are relative to your goal and available capacity. Agree what they mean before sorting ideas.

I — Implement

Higher expected impact · Lower effort

Consider early testing or action after checking evidence, risk and dependencies—not automatic implementation.

C — Challenge

Higher expected impact · Higher effort

Explore feasibility, resources and a smaller test. Greater effort is not, by itself, a reason to reject a necessary change.

If the difficulty is conflicting needs, try TRIZ separation: could different times, places, conditions or parts meet both needs? Check feasibility rather than assuming separation makes the idea easy.

P — Possible

Lower expected impact · Lower effort

Keep available, but do not let easy tasks crowd out more useful work. Check whether they enable another change.

K — Kill

Lower expected impact · Higher effort

Normally decline, stop or redesign. Check obligations and enabling work before dropping it; arrange a safe handover if needed.

Impact and effort are estimates—not measured results or a precise ranking within each category.

Run a useful PICK discussion

    Check what happened and what helped bring it about before choosing a solution. Do not build the plan around an assumed cause.
  1. Agree the problem and desired result. Check what happened and what helped bring it about See - What are system conditions? → before choosing a solution. Do not build the plan around an assumed cause.
  2. Define impact and effort. Include likely customer benefit, possible harm, implementation, ongoing support, scarce skills, cost and time. Identify essential safety or legal work separately.
  3. Sort ideas with the people affected. Record the reasoning, evidence and uncertainty. Disagreement may mean you need information or a test—not a louder vote.
  4. Check connections. Some ideas depend on other changes. Ask whether the proposed combination is sufficient; do not assume every helpful idea is necessary or cherry-pick only the easiest.
  5. Choose a manageable next step. Name an owner, capacity, expected effect and review date. Agree what stops or waits. Revisit the sorting as evidence and circumstances change.
Common traps to avoid
  • Everything looks high impact and low effort. Check estimates against experience, dependencies and support demands. Keep uncertainty visible.
  • Easy wins become the whole plan. A small local gain may not improve the overall result. Check what is limiting progress.
  • “Kill” becomes personal. Discuss the future value and cost of the work, not the worth of its proposer. Past spending alone is not a reason to continue.
  • The matrix becomes a backlog of promises. An attractive idea is not a commitment. Limit active work and communicate what is deferred.

Test the idea—not just its rating

Before starting, agree what should change, by how much and when. Use an appropriate test and measures of benefit, implementation and possible harm.

A suitable chart or Bootstrap CUSUM analysis can help identify a performance change. It cannot, by itself, prove the idea caused that change. Consider timing, other changes and data quality; no detected change is not automatically proof of no benefit.

Choose useful measures →

Keep the tools in perspective

PICK compares ideas. Focus and prioritisation makes room to deliver them. Investigation and testing help check the explanation and response. These are complementary tools—not mandatory stages or a guaranteed route to success.

Make room to finish →
Return to Where might the difficulty lie? →