Solution Selection Matrix: Turn a Diagnosed Cause Into a Pilot Worth Running

The Solution Selection Matrix sits in the Improve phase of RDMAICS. In the product's own words, it lets you score candidate solutions against weighted selection criteria to choose the optimal solution set for piloting.

You reach for it at a specific moment. Analyse has finished. You have a cause you believe in. Your team has produced a list of things you could do about it. The list is too long to action and the arguments about what to do first are circular. The matrix converts that argument into numbers that everyone can see.

The point is not to replace judgement. The point is to make judgement visible. When you score Impact, Feasibility, Cost and Time separately, you force people to say which criterion they are actually arguing about. Most disagreements collapse once that is clear.

Everything below uses the sample data built into the tool. The case is fictional. It is there so you can see the mechanics without needing your own data loaded.

The scenario is an invoice processing cycle. The diagnosed cause points at manual handling and routing delays. Five candidate solutions are on the table.

Each one is plausible. Each one has a sponsor somewhere in the organisation. None of them can be judged by enthusiasm alone, which is why they get scored.

The worksheet takes one row per candidate solution. Each row carries four ratings on a one to ten scale: Impact on Y, Feasibility, Cost and Time. The tool then holds a Weighted Score for each row.

In the sample data, OCR-based invoice capture scores 9 on Impact, 8 on Feasibility, 6 on Cost and 6 on Time, giving a Weighted Score of 232. Auto-routing with delegation scores 8, 9, 9 and 9 for a Weighted Score of 268. Automated 3-way match rules scores 7, 8, 8 and 8 for 244. Vendor portal plus validation scores 6, 6, 5 and 5 for 176. Vendor master cleanup scores 5, 9, 9 and 9 for 226.

Two setup habits matter more than the arithmetic. First, agree the rating direction before anyone scores. A ten on Cost should mean the same thing to every person in the room, and you should write down whether a high number means cheap or expensive. Second, agree the criterion weights before you see any scores. Weights chosen after the fact tend to favour whichever option the loudest person already wanted.

The tool produces a ranked bar chart built from the worksheet. The axis runs from 0 to 280, with gridlines at 70, 140 and 210.

The ranking reads: Auto-routing with delegation at 268, Automated 3-way match rules at 244, OCR-based invoice capture at 232, Vendor master cleanup at 226, and Vendor portal plus validation at 176.

Read the gaps, not just the order. The top four sit between 226 and 268, a band of 42 points. The fifth sits at 176, well below the rest. That shape tells you something useful. There is no runaway winner at the top, but there is a clear loser at the bottom. The tool also offers an AI review that explains the result in plain language alongside the chart.

The tool states plainly that it ranks by the supplied weighted scores and does not independently recalculate them. It also prompts you to record criterion weights and rating direction in the selection rationale. Treat that as an instruction, not a footnote.

Ranking is only half the job. Before you pilot anything you should be able to say what change you expect and how you will see it.

Look at the Impact on Y column separately from the total. OCR-based invoice capture carries the highest Impact score in the set at 9, yet it ranks third overall at 232 because its Cost and Time scores are both 6. Auto-routing with delegation carries a lower Impact score of 8 but scores 9 on Feasibility, Cost and Time, which lifts it to 268.

That contrast is the real output of the matrix. The top ranked option is not the biggest lever. It is the best combination of lever and effort. If your sponsor wants maximum movement on Y and will fund it, the conversation changes and you may pilot the 9 on Impact instead. The matrix gives you the language to have that conversation with evidence on the table.

Vendor master cleanup is worth a second look for the same reason. Its Impact score is the lowest in the set at 5, but it scores 9 on Feasibility, Cost and Time, landing it at 226. That is the profile of groundwork. It may not move Y much alone, but it may make the other solutions work better.

Once the ranking is on screen, a small set of decisions becomes much easier to make and much easier to defend later.

Use the output to drive these conversations, and record the reasoning each time.

The most common mistake is scoring in a hurry at the end of a workshop. Tired people anchor on the first number said out loud. Score each criterion down the column rather than across the row, so you compare like with like.

The second mistake is inconsistent rating direction. If Cost sometimes means cheapness and sometimes means expense, your totals are meaningless. In the sample data Vendor master cleanup scores 9 on Cost and Vendor portal plus validation scores 5, and those numbers only make sense if everyone read the scale the same way.

The third mistake is treating a narrow gap as a verdict. The difference between 244 and 232 is twelve points out of a scale that reaches 280. That gap will not survive one person changing one rating by a point. Where scores cluster, use the matrix to shortlist and then use a different method, such as a pilot or a small test, to separate the finalists.

The fourth mistake is leaving out the rationale. Six months later nobody remembers why Feasibility was weighted the way it was. Write it down at the time.

The Solution Selection Matrix ranks what you give it. It does not verify your scores. The tool says so directly: it ranks by the supplied weighted scores and does not independently recalculate them. If your ratings are guesses, the ranking is a tidy presentation of guesses.

It does not tell you whether a solution will work. A Feasibility score of 9 is an opinion about feasibility. The pilot is what tests it. Nothing in the 268 attached to Auto-routing with delegation guarantees any result in the real process.

It does not size a financial benefit. There are no dollars in this tool and none in the sample data. If you need a business case, build one separately and use the matrix as an input to it rather than a substitute for it.

It does not generate options. The quality of the ranking is capped by the quality of the five rows you entered. If the best idea never made it onto the worksheet, the matrix will confidently rank the rest without ever noticing. Spend time on the option generation step before you spend time on the scoring.