> ## Documentation Index
> Fetch the complete documentation index at: https://meta.niceshare.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Decision Matrix

> A decision matrix evaluates options against weighted criteria to choose the best path. Learn the Pugh method, scoring, and when it fails.

<Info>
  **Category**: Models<br />
  **Type**: Decision Support Tool<br />
  **Origin**: Stuart Pugh, 1980s; formalized in *Total Design* (1991)<br />
  **Also known as**: Pugh Matrix, Criteria Matrix, Grid Analysis, Weighted Decision Matrix
</Info>

<Note>
  **Quick Answer** — A decision matrix is a structured table that compares several options against the same set of weighted criteria. It turns competing priorities into visible scores so a team can see why one option wins instead of relying on gut feeling or the loudest voice.
</Note>

## What is a Decision Matrix?

A **decision matrix** is a decision-making tool that lists your options along one axis and your decision criteria along the other. Each option is scored against every criterion, the scores are multiplied by the weight of each criterion, and the totals reveal which option best matches what you actually value. The core insight is that "the best choice" is not a personality contest; it is the option that scores highest against criteria you have agreed on in advance.

> A decision matrix evaluates and prioritizes a list of options.

This definition comes from the American Society for Quality (ASQ), which teaches the matrix as part of its quality toolbox. The method is especially useful when no single option dominates: one choice is cheaper, another is faster, and a third is more reliable. The matrix forces the trade-offs out into the open.

### Decision Matrix in 3 Depths

* **Beginner**: Draw a table. List your options as rows and the things that matter—cost, quality, speed—as columns. Score each cell from 1 to 5, add the rows, and see which option tops the list.
* **Practitioner**: Assign weights to the criteria before scoring. If cost matters twice as much as speed, give cost a 2 and speed a 1. Multiply each score by its weight, then sum. This prevents a trivial criterion from overturning an important one.
* **Advanced**: Treat the final score as the start of a conversation, not a verdict. Sensitivity analysis—slightly changing the weights—shows whether the winner is robust or merely an artifact of how you weighted the first draft.

## Origin

The modern decision matrix traces back to **Stuart Pugh**, a Scottish design engineer at the University of Strathclyde. In the early 1980s, Pugh developed a concept-selection method for engineering teams who had to choose between competing designs before detailed data was available. His version, often called the **Pugh Matrix**, used a simple datum or baseline design and scored alternatives as better (+), same (0), or worse (−) on each criterion. The method was published in *Total Design: Integrated Methods for Successful Product Engineering* (1991).

A parallel line of work produced the **weighted decision matrix**. In 1980, **Thomas L. Saaty** published *The Analytic Hierarchy Process* (AHP), a mathematical method for deriving weights from pairwise comparisons. AHP and related multi-criteria methods turned the matrix from a simple scoring grid into a family of rigorous decision-support techniques used in procurement, public policy, and engineering.

Today, the terms overlap. A "Pugh matrix" often implies a baseline comparison with simple symbols, while a "weighted decision matrix" uses numerical scores and explicit weights. Both share the same purpose: make trade-offs visible and debatable.

## Key Points

A decision matrix is only as good as the thinking that goes into the criteria and weights. Garbage criteria produce garbage rankings.

<Steps>
  <Step title="List criteria before you list options">
    Decide what "good" means before you know which option you prefer. Common criteria include cost, quality, risk, speed, ease of implementation, and strategic fit. Involve stakeholders so one person’s hidden priority does not dominate later.
  </Step>

  <Step title="Assign weights that add up to a fixed total">
    ASQ recommends distributing 10 points across the criteria. A criterion that matters twice as much as another gets roughly twice the points. This step is where most of the real negotiation happens.
  </Step>

  <Step title="Score each option independently">
    Rate every option against every criterion using a consistent scale, such as 1 to 5. Watch for scale direction: a high score should always mean "more desirable," so use "low cost" rather than "cost."
  </Step>

  <Step title="Multiply, sum, and then discuss">
    Multiply each score by its criterion weight, add the weighted scores for each option, and rank the results. The highest score is the leading candidate, but the team still decides whether the margin is meaningful and whether any non-scored factor should change the choice.
  </Step>
</Steps>

## Applications

Use a decision matrix whenever a group must choose one path from several plausible ones and the choice depends on more than one factor.

<CardGroup cols={2}>
  <Card title="Choose software or vendors" icon="laptop-code">
    Score vendors on cost, security, integration, support, and roadmap. A cheaper vendor may lose if security and support matter more to your team.
  </Card>

  <Card title="Prioritize product features" icon="star">
    Compare features by user value, development effort, strategic alignment, and technical risk. The matrix keeps "cool idea" from beating "valuable idea."
  </Card>

  <Card title="Pick a career or relocation option" icon="briefcase">
    Weight salary, commute, growth, stability, and family fit. The matrix surfaces which factors you are actually treating as most important.
  </Card>

  <Card title="Select a project portfolio" icon="folder-tree">
    Rank proposed projects by expected return, resource load, risk, and alignment with company goals. The result is a defensible funding order.
  </Card>
</CardGroup>

## Case Study

### NASA’s Artemis III Landing Site Selection

In 2024, a team of researchers published a multi-criteria decision-making analysis to help NASA choose landing sites for the Artemis III mission near the lunar south pole. The challenge was not a shortage of options; it was an excess of them. The team evaluated **1,247 potential locations** inside 13 candidate regions using criteria such as surface visibility, line-of-sight to the lander, sunlight exposure, Earth communication, and geological value.

Using a **TOPSIS** multi-criteria method—a weighted decision-matrix approach that ranks options by their distance from an ideal solution—the analysis identified a leading site known as **DM2 (Nobile Rim 2)** at approximately **84.20°S, 60.70°E**. The result was not a hunch about a pretty crater; it was the location whose weighted profile best balanced safety, science, and communication constraints. The researchers also ran sensitivity analyses to confirm that the winner stayed robust when the weights or method changed.

The lesson for everyday decisions is the same. Artemis did not pick the moon’s most famous spot; it picked the spot that scored highest against the mission’s actual criteria. When your own options all look reasonable, a decision matrix turns the conversation from "which one feels right?" into "which one best matches what we said mattered?"

## Boundaries and Failure Modes

A decision matrix can look more objective than it really is. Three limits deserve attention:

1. **The weights encode opinion, not truth.** A matrix does not discover what matters; it records what the group agreed matters. If the weights are manipulated to make a preferred option win, the tool becomes theater.

2. **It struggles with missing or soft criteria.** Factors like "team morale," "brand risk," or "regulatory goodwill" are hard to score numerically. Ignoring them because they do not fit the grid is a common failure mode.

3. **The highest score is not automatically the answer.** ASQ explicitly warns that the matrix is a decision-support aid, not a replacement for judgment. A winning option with a hidden deal-breaker should still be rejected.

A frequent misuse is filling out the matrix after the decision has already been made. If the team starts with "we all know Option B is best," the scoring step becomes rationalization, not reasoning.

## Common Misconceptions

<AccordionGroup>
  <Accordion title="Misconception: A decision matrix removes human judgment">
    The matrix organizes judgment; it does not eliminate it. People still choose the criteria, assign the weights, and interpret the result. The value is transparency, not automation.
  </Accordion>

  <Accordion title="Misconception: The option with the highest score must be chosen">
    The highest score is the leading candidate. If it has a non-negotiable flaw—legal risk, ethical concern, or a dependency that cannot be met—the team should still override the number.
  </Accordion>

  <Accordion title="Misconception: More criteria always make a better matrix">
    A bloated matrix with twenty criteria becomes noise. The goal is to capture the few dimensions that genuinely separate the options, not to make the table impressive.
  </Accordion>
</AccordionGroup>

## Related Concepts

These tools sit next to the decision matrix in the same problem space: choosing well when multiple factors compete.

<CardGroup cols={3}>
  <Card title="Eisenhower Matrix" icon="table-cells" href="/models/eisenhower-matrix">
    A 2×2 matrix that prioritizes tasks by urgency and importance. Useful when the criteria are fixed and the goal is quick categorization.
  </Card>

  <Card title="Decision Tree" icon="sitemap" href="/models/decision-tree">
    Maps decisions and uncertain outcomes as branching paths. Better than a matrix when timing, probability, and sequential choices matter.
  </Card>

  <Card title="Expected Value" icon="calculator" href="/models/expected-value">
    Calculates the long-run average outcome of risky options. Pair it with a matrix when probabilities and payoffs are available.
  </Card>

  <Card title="Pareto Principle" icon="chart-pie" href="/laws/pareto-principle">
    Suggests that a few criteria often drive most of the decision. Use it to avoid overloading the matrix with minor factors.
  </Card>

  <Card title="Opportunity Cost" icon="scale-balanced" href="/models/opportunity-cost">
    The value of what you give up when you choose one option. A matrix scores options; opportunity cost reminds you what the runner-up offered.
  </Card>
</CardGroup>

## One-Line Takeaway

<Tip>Before you argue about which option is best, agree on what "best" means—then let the decision matrix carry the argument.</Tip>


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.