The Kepner-Tregoe Matrix: Clear Thinking for Tough Decisions (That You Can Stand Behind)

Download: KT Decision Analysis Worksheet (step-by-step guide with scoring tables) — make a copy of this google doc and fill it out yourself.

You are an expert decision coach using the Kepner-Tregoe (K-T) Decision Analysis framework.
Guide me through making a high-stakes decision in 7 steps.
Ask me one question at a time, wait for my answer, then move to the next.
Here’s the structure to follow:

1. Clarify Purpose – Ask me what decision I’m trying to make and what success looks like (short + long term).
2. Musts vs. Wants – Help me list objectives. Classify each as a “must” (mandatory, measurable, realistic) or “want.”
3. Weight Wants – For each want, have me assign a weight from 1–10 based on importance.
4. Screen Alternatives – Ask me to list possible options. Eliminate any that fail a “must.”
5. Score Alternatives – For remaining options, help me score each want (0–10). Multiply by the want’s weight, then total scores.
6. Assess Risks – For the top-scoring option, ask what could go wrong short/long term, what happens if we’re close to a must, and where assumptions/data might be weak.
7. Make Choice + Next Step – Ask if I accept the risks. If yes, confirm the choice. If not, revisit the next option. Then ask me to name the smallest next step to move forward.

Throughout, remind me to connect each must, want, and trade-off to my core values.

1. What is going on? – Situation Appraisal
2. Why did it happen? – Problem Analysis
3. What should we do? – Decision Analysis
4. What lies ahead? – Potential Problem (and Opportunity) Analysis

• DTS (days to solve) down ~1 day per case
• TMPI (time minutes per incident) down 27 minutes
• Customer satisfaction up 3.3%
  More than 7,000 team members now apply a shared playbook. Translation: fewer heroics, more systems. My kind of party.

• 74% reduction in average time-to-restore
• Fewer global incidents and tighter variation
  Baseline first, then streamline steps and comms, then train the cadence. That’s how you make “calm under pressure” an SOP, not a personality trait.

• A must = non-negotiable, measurable, realistic (safety threshold, legal compliance, margin floor).
• A want = desirable outcome ranked by importance (brand lift, speed, innovation, experience).

1. Clarify purpose
   What decision are we making? Success looks like… (short and long term).
2. List objectives, split into musts and wants
   • Musts: mandatory, measurable, realistic. One “no” and it becomes a want.
   • Wants: rank 1–10 for importance (that ranking = your value judgement).
3. Screen alternatives
   If an option fails any must, it’s No-Go. No exceptions, no “favourite child” exemptions.
4. Score remaining options against wants
   For each want: best option gets 10, others 0–9 relative performance. Multiply by the want’s weight. Sum totals.
5. Assess risk starting with the top score
   What could go wrong (near + long term)? What happens if we’re right on the edge of a must? Where might our data be shaky?
6. Make the best balanced choice
   If you accept the risks for the top option, choose it. If not, move to the next score. This isn’t perfection; it’s clarity with accountability.
7. Plan action
   Smallest, obvious next step. Owner. Deadline. Then communicate the why—including the values that drove the trade-offs.

Grab the worksheet to do all of this in minutes: weights, scores, and a tidy risk page you can show your board. Download the Google Doc here.

• Stakes are high; outcomes affect people, customers, safety, or significant dollars.
• Options are many, time is tight, bias is likely.
• You need transparent reasoning to get buy-in.

• You need divergent, creative leaps (start with ideation, then evaluate with K-T).
• Information is unknowable, and speed trumps rigour (make a reversible call, gather data, revisit).
• It’s lunch. Please don’t run a must/want matrix on tacos.

“Managers usually do an unsatisfactory job of setting priorities because they approach each issue as it comes along, on the merits presented by the narrow context of the issue itself.”
— Charles H. Kepner

• Speed matters more than precision. In a crisis where seconds count, use a playbook, not a 9-step matrix.
• You don’t have the data. K-T depends on solid facts. If you’re guessing anyway, the spreadsheet won’t save you.
• Creative leaps are required. This is not a brainstorming tool. If you’re designing the next iPhone, start with divergent thinking first.
• Low-stakes choices. Please don’t use a must/want scale for tacos. That’s not leadership—it’s a lunch delay. (And if you’re scoring tacos on a spreadsheet, you might need a vacation.)
• When values aren’t clear. If your team hasn’t agreed on what truly matters, the weighting system becomes a matter of politics rather than clarity.

• Time-intensive. You need patience and discipline—it’s not built for “move fast and break things.”
• Rigid. It requires sticking to the process to get the payoff, which can frustrate leaders who thrive on intuition.
• Data-heavy. If you don’t have good information, the method bogs down. Garbage in, garbage out.

• How to set musts (non-negotiables) and wants (preferences).
• Why does weighing your wants upfront keep bias out?
• How to eliminate options that don’t pass the sniff test (sorry, Ferrari, two doors won’t cut it).
• How to score and calculate the best option (Fiesta beats Corsa—math wins, bias loses).

1. Name the decision.
2. Write three musts (measurable).
3. Write five wants and weigh them.
4. List 3–4 options, score them, then do a quick risk pass on the winner.
5. Decide. Communicate the values behind the trade-offs. Ship.

Download the KT Decision Analysis Worksheet — linked here — and run the play with your team this week. If Olympians need to see a cue ~13 times before it sticks, your organization probably needs at least that many reps too. Inculcate the cadence until clear thinking becomes culture.