Effort vs Impact Matrix: Turn a Diagnosed Cause Into a Change Worth Piloting

The Effort vs Impact Matrix screens solutions by implementation effort versus expected impact to identify quick wins and strategic bets. That is the whole job. It does not diagnose a problem and it does not design a fix. It takes a list of candidate changes you already have and sorts them into an order you can defend.

You reach for it in the Improve phase of RDMAICS, after Analyse has given you a cause you believe in. By that point you usually have more ideas than capacity. Some came from the team, some from a supplier, some from a manager who read something. Without a screen, the loudest idea wins. With a screen, the idea with the best return for the work wins.

The tool is also a conversation device. Scoring forces people to say out loud how hard something will be and how much it will move the measure. Disagreement surfaces early, at the scoring table, instead of halfway through a build.

Everything below uses the sample data the tool loads when you open it. The case is fictional. It is an accounts payable process where invoice handling is slow and error prone, and the Analyse phase has pointed at manual capture, manual routing and unreliable vendor data.

Six candidate solutions sit on the worksheet. They range from a large system change to a small data clean. They are not equivalent in cost, risk or time, and that is the point. You need a single comparable view before you commit anyone's week to anything.

The six candidates are OCR-based invoice capture, auto-routing with delegation, automated 3-way match rules, a vendor portal with validation, a mobile approval app, and a vendor master cleanup. Each one plausibly addresses part of the diagnosed cause. Only scoring separates them.

Each row needs a solution name, an impact score from 1 to 10, and an effort score from 1 to 10. Impact is how much the change moves the problem measure. Effort is the work to get it live, including build, test, training and change management. Keep the scales consistent across the whole list or the matrix means nothing.

In the sample, OCR-based invoice capture scores impact 9 and effort 6. Auto-routing with delegation scores impact 8 and effort 3. Automated 3-way match rules score impact 7 and effort 4. The vendor portal with validation scores impact 6 and effort 7. The mobile approval app scores impact 4 and effort 8. Vendor master cleanup scores impact 5 and effort 2.

Score as a group, not alone. Ask the person who will build the thing to score effort. Ask the person who owns the measure to score impact. Where the two differ by more than a couple of points, talk it through before you write a number down. The talk is often worth more than the score.

The tool plots impact against effort and assigns each row to a quadrant. Every point on the chart comes from one complete worksheet row, so an incomplete row simply will not appear. Three of the six land as Quick Win, two as Major Project and one as Time Sink.

Auto-routing with delegation sits at effort 3 and impact 8. Automated 3-way match rules sit at effort 4 and impact 7. Vendor master cleanup sits at effort 2 and impact 5. All three are quick wins and all three carry the decision to implement now. OCR-based invoice capture sits at effort 6 and impact 9 and is a major project marked to prioritise. The vendor portal with validation sits at effort 7 and impact 6, also a major project, marked phase 2. The mobile approval app sits at effort 8 and impact 4 and is a time sink marked defer.

The shape of the picture is the message. The highest impact item on the board, at 9, is not the cheapest, at effort 6. The cheapest item on the board, at effort 2, returns a middling impact of 5. The tool offers an AI review that explains the pattern in the plotted result if you want a second reading.

The matrix lets you size relative gain without a business case. Auto-routing with delegation returns an impact of 8 for an effort of 3. The mobile approval app returns an impact of 4 for an effort of 8. You do not need a model to see which of those two deserves a pilot slot this quarter.

This is how a diagnosed cause becomes a change worth piloting. You take the cause, list the candidate responses, score them on one scale, and pick the smallest change that moves the measure most. Then you pilot that change and watch the measure. If the measure moves, your impact score was honest. If it does not, you learn something about your scoring, which improves every matrix after it.

Sequencing matters as much as selection. Vendor master cleanup at effort 2 and impact 5 is modest on its own, but clean vendor data makes later work easier. Running the small item first is often what makes the large item at impact 9 achievable.

Four decisions appear in the sample: implement now, prioritise, phase 2 and defer. Each one is a commitment of a different size, and the matrix makes the size visible to everyone in the room.

Implement now applies to the three quick wins. Start them, finish them, and measure. Prioritise applies to OCR-based invoice capture, the impact 9 item. It needs planning and funding, so it enters the pipeline with a named owner. Phase 2 applies to the vendor portal with validation at impact 6 and effort 7. It stays on the list but waits. Defer applies to the mobile approval app, where impact 4 does not justify effort 8 today.

Defer is not reject. Effort scores change when platforms change and impact scores change when the process changes. Review deferred rows at the next cycle rather than deleting them.

The most common mistake is scoring impact on enthusiasm. Someone wants the mobile app, so impact creeps up. Anchor every impact score to the measure from the Measure phase and the creep stops.

The second mistake is underscoring effort. Teams count build time and forget training, data migration, policy updates and the weeks of hand holding after go live. If your effort scores are all low, you are probably measuring only the easy part.

The third mistake is treating the quadrant as the answer. The quadrant is a sorting aid. A quick win with impact 5 and a quick win with impact 8 sit in the same box and are not the same decision. Read the numbers, not just the label. The fourth mistake is scoring in isolation, where one person fills the worksheet and the team sees it only as a finished chart.

The matrix will not validate your scores. It plots what you type. If impact 9 is wrong, the chart still shows a confident point in the top region. Garbage in, confident chart out. The discipline has to come from the scoring conversation.

It will not tell you whether a solution addresses the real cause. That is Analyse work. A high impact score on a solution aimed at the wrong cause produces a well prioritised waste of time. Check the causal link before you score.

It will not account for dependencies, risk, cost in money, regulatory constraints or capacity. The sample shows six items with three marked implement now, but it says nothing about whether one team can run three changes at once. It will not predict the realised benefit either. Only the pilot and the measure can tell you that.

Use the matrix for what it is. A fast, visible, shared screen that turns a long list into a short, ordered one. Then move to Control and Sustain, where you prove the change held.