Decision Matrix Presentation: Explain the Scores and the Trade-offs

A decision matrix presentation should explain why you recommend an option, not just announce which row has the largest number. The audience needs to see the trade-off, the assumptions behind the scores, and what would change your recommendation.

Prepare the calculations first. You can then use Presenti’s text-to-presentation workflow to turn your checked notes into a slide draft. Keep the scoring in your working file: a generated presentation is not a substitute for evaluating the options.

When a weighted matrix belongs in a presentation

Use a weighted decision matrix when you are comparing feasible answers to the same question and several criteria matter. For example: which of three approaches should a team use to handle incoming service requests? The method is less useful when the options solve different problems or the evidence is too incomplete to support ratings.

ASQ’s decision matrix guidance describes evaluating options against weighted criteria, using consistent rating directions, and discussing the resulting totals. The highest total supports a conversation; it does not settle the decision automatically.

This is different from a two-axis prioritization chart. A quadrant chart positions items along two dimensions. A weighted matrix combines several explicitly defined criteria. Neither turns judgment into objective truth.

Separate non-negotiable requirements from preferences

Before assigning weights, check the requirements that every acceptable option must meet. A high score for convenience cannot compensate for failing a mandatory security requirement or missing an immovable launch date.

For the fictional example below, the team requires an approved data-handling arrangement and delivery within twelve weeks. All three options have passed those checks. An unresolved requirement would be marked “not yet eligible,” not given a low score and averaged away.

The remaining decision is how much to value workflow fit, affordability, delivery speed, and ongoing maintenance. Agree on those priorities before looking at the totals. Otherwise, it is easy to adjust the weights until a preferred option wins.

A worked example: three options, four criteria

The team is choosing between extending its current system (A), adopting a hosted system (B), and building a custom tool (C). These are invented teaching inputs, not Presenti customer results or vendor ratings. Each approach covers the same service-request process.

Use a scale from 1 to 5 where a higher rating is always more desirable. In this example, “affordability” means lower cost, and “maintenance simplicity” means fewer recurring staff hours. Do not label a column “cost” and accidentally reward the highest spend.

Calculate each total by multiplying each rating by its weight, then adding the four contributions. For B: (5 × 0.40) + (3 × 0.25) + (3 × 0.20) + (4 × 0.15) = 3.95. In a spreadsheet, store the weights as percentages and use SUMPRODUCT on matching rating and weight ranges. Do not multiply percentage-formatted cells by 100 again.

B leads A by only 0.05 on this five-point scale. That is a small modeled difference, not a 95% confidence level, a probability of success, or evidence that B is universally better. The useful finding is that B offers stronger workflow coverage while A is cheaper and faster.

Make the ratings reproducible

Keep a short rubric beside the working matrix. Here is the complete rubric used for the example:

• Workflow fit: one point per target task supported, out of five defined tasks. A supports three, B five, and C four.
• Affordability: estimated cost as a share of the preferred budget: up to 70% = 5; above 70–85% = 4; above 85–100% = 3; above 100–115% = 2; above 115% = 1. The separate absolute spending limit still applies. A is estimated at 60%, B at 90%, and C at 110%.
• Delivery speed: up to four weeks = 5; above four–six = 4; above six–eight = 3; above eight–ten = 2; above ten–twelve = 1. A takes four weeks, B eight, and C ten.
• Maintenance simplicity: recurring work per week: up to 0.5 hours = 5; above 0.5–1 = 4; above 1–2 = 3; above 2–4 = 2; above 4 = 1. A requires two hours, B one, and C four.

These bands are choices made for this example, not an industry standard. For a real decision, record the estimate owner, evidence, date, and uncertainty. An unsupported value is “unknown,” not zero. Resolve it or show a plausible range before claiming a ranking.

Also check for double counting. If “delivery speed” and “time to launch” measure the same thing, using both gives that preference extra influence. A weighted total also assumes that trade-offs between the chosen ratings are acceptable. Keep requirements that cannot be traded outside the sum.

Show when the recommendation changes

Now suppose the decision-makers place more importance on staying below the preferred budget. Move fifteen percentage points from workflow fit to affordability; keep the other weights unchanged. Do not change the ratings.

The preferred option changes from B to A. That does not invalidate the matrix. It identifies the decision the meeting actually needs to make: is complete workflow coverage worth the extra cost and delivery time?

A useful recommendation is conditional: “Choose B if covering all five tasks remains the priority. Choose A if affordability takes precedence and the two unsupported tasks have an acceptable workaround.” Explain the cost and feasibility of that workaround before asking for approval.

Turn the analysis into two main slides

Slide 1: the recommendation and its trade-off. Use a message title such as “B covers all five target tasks; A is cheaper and faster.” Show the three totals with the scale clearly labeled, then name the two differences that matter. Do not hide A’s strengths or use a green check mark to imply that B is already approved.

Slide 2: the condition that could change the choice. Show the two weight scenarios and the resulting leader. End with the precise question: “Should we prioritize full coverage or lower cost for this release?” If evidence is missing, ask for the investigation needed instead of approval to proceed.

Keep the full ratings, rubric, requirement checks, source references, and calculation file in the appendix or accompanying material. The audience should be able to inspect them without reading twelve scoring explanations on the opening slide. Use labels as well as color, and check the matrix at the actual presentation size.

Prepare the slide draft from checked notes

For a Presenti draft, paste the decision context and checked findings rather than asking the AI to choose a winner. The following brief gives it a bounded writing task:

Create a presentation comparing three service-request approaches. The figures are fictional teaching data. A scores 3.90, B 3.95, and C 2.80 out of 5 under the original weights. B covers more target tasks; A is cheaper and faster. Under the affordability-first weights, A scores 4.20, B 3.65, and C 2.50. Draft two main slides: the trade-off, then the decision about priorities. Put the full rubric and matrix in an appendix. Preserve every number. Do not invent evidence, prices, probabilities, or customer outcomes.

Review the draft against your calculation file. Keep “out of 5” with the totals, retain the conditional recommendation, and remove any wording that turns estimates into facts. Build or paste the final matrix in the format your chosen editor supports; this workflow does not assume that Presenti calculates weighted scores or provides a dedicated decision-matrix template.

The finished deck should make disagreement specific. If the audience challenges a score, discuss its evidence. If they challenge a weight, discuss the priority. If an option fails a requirement, take it out of contention. That is more useful than defending a total as though the number made the decision for you.