Kepner-Tregoe Method

Thinking Tools5.2 Selecting the Right Decision ToolsAlso known as Kepner-Tregoe, KT Analysis

Developed by Charles H. Kepner and Benjamin B. Tregoe (The Rational Manager, 1965; The New Rational Manager, 1981).

Splits a tangled management situation into four separate processes, situation appraisal, problem analysis, decision analysis, and potential problem analysis, each with its own steps.

Three warehouse sites, screened on the musts and scored on the wants

A must is pass or fail: a site that fails one is out, whatever it scores. The sites still standing are scored 1 to 10 on four wants, each multiplied by its weight, and ranked by the weighted total. Airport Park scores highest and is still out, because it cannot open by July 1. Change a score or a pass and watch the ranking. Illustrative.

Operating by July 1

MustRiversideNorthgateAirport Park

Within a four-hour drive of 80 percent of customers

MustRiversideNorthgateAirport Park

Local labor pool for warehouse shifts

Weight9Riverside6Northgate8Airport Park9

Highway access for trucks

Weight7Riverside8Northgate7Airport Park9

Lease cost (10 is the cheapest)

Weight5Riverside7Northgate6Airport Park4

Room to expand

Weight4Riverside5Northgate7Airport Park8

Weighted total

Riverside165OutNorthgate179OutAirport Park196Out

Recommended siteNo site passes every must: widen the search or revisit a must.RiversideNorthgateAirport Park

In order

  1. 179RiversideNorthgateAirport ParkOut: fails a must (Operating by July 1).Out: fails a must (Within a four-hour drive of 80 percent of customers).
  2. 165RiversideNorthgateAirport ParkOut: fails a must (Operating by July 1).Out: fails a must (Within a four-hour drive of 80 percent of customers).
  3. 196RiversideNorthgateAirport ParkOut: fails a must (Operating by July 1).Out: fails a must (Within a four-hour drive of 80 percent of customers).

An example

A regional food distributor needed a third distribution center running before the holiday season, and the search had narrowed to three sites. Every meeting drifted: from the choice to the late deliveries that had started the search, and from there to what might go wrong after the move. The operations vice president split those into separate sessions and took the choice itself through decision analysis.

Two requirements were musts: the site had to be operating by July 1, and within a four-hour drive of 80 percent of customers. Four wants were weighted: the local labor pool at 9, highway access at 7, lease cost at 5, and room to expand at 4. Airport Park scored highest on the wants, 196 to Northgate's 179, but its landlord could not deliver the building before September, so it was out. Northgate went to the board with its scores. A separate session then listed what could go wrong there, a tight hiring market and a road project on the main access route, each with a preventive step and a fallback.

Why it works

Most management meetings mix four questions: what is going on, why something went wrong, what to choose, and what could go wrong with the choice. Each needs different information, and a team that keeps switching among them settles none. Separating them is the method's first contribution.

In the choice itself, the split between musts and wants does the work. A must is a line the decision cannot cross, so it screens options out before any scoring, and no score can buy back a failed requirement. Wants are weighted because they are not equally important, and fixing the weights before scoring makes the trade-offs explicit and keeps the favorite from setting them. The last process then puts the chosen option through a deliberate look at its risks while there is still time to act on them.

The thinking behind it

Charles Kepner Charles Kepner, a management consultant, founded the firm Kepner-Tregoe with Benjamin Tregoe in 1958. Their book The Rational Manager (1965) set out step-by-step methods for two jobs. Problem analysis starts by stating a problem as a deviation from what should be happening, which keeps a team from arguing about causes before it agrees on what went wrong. Decision analysis weighs the options against weighted criteria, the procedure the figure on this page runs on three warehouse sites.

Benjamin Tregoe Benjamin Tregoe, Kepner's partner, started the firm with him in his garage in Pacific Palisades, California, in 1958, and wrote both books with him. The New Rational Manager (1981) added situation appraisal, the process for sorting a tangle of concerns into separate issues and ranking them before any analysis begins. It is the step the distributor took when its drifting meetings were split into separate sessions, one question to each.

How to run it

  1. Appraise the situation. List every concern on the table, split the tangled ones apart, and rank them by seriousness, urgency, and how fast they are growing.
  2. Analyze a deviation where there is one. State what should be happening and what is, then compare where the problem occurs and where it does not to narrow its cause.
  3. Write the decision statement. One line naming the choice and its purpose, such as select a third distribution center to serve the holiday season.
  4. Separate musts from wants. A must is measurable and absolute; anything negotiable is a want, weighted from 1 to 10 before any option is scored.
  5. Screen, then score. Drop every option that fails a must, then score the survivors on each want and multiply by its weight.
  6. Test the winner for risks. List what could go wrong with the top choice, how likely and how serious each risk is, and whether any changes the ranking.
  7. Plan preventive and contingent actions. For each serious risk, name a step that makes it less likely and a fallback with the trigger that sets it off.

Where it goes wrong

Musts multiply. A team that writes a must for every preference screens out every option, or leaves only the favorite standing; spot it when a must cannot be measured, or when someone would trade it for enough of something else, and move it to the wants.

The second is weighting after scoring. Once the scores are in, weights drift toward whatever puts the favorite on top. Agree on the weights and record them before anyone scores an option.

The third is stopping at the ranking. The top score with an unexamined risk can be a worse choice than the runner-up, and the risk session is the one teams skip when a deadline presses. Schedule it separately, with its own owner.

When to use something else

  • Weighted Decision Matrix. When there is one choice to make and no tangle of problems around it, so a scored matrix alone is faster.
  • Analytic Hierarchy Process. When the weights themselves are in dispute and the group needs a check on whether its judgments agree.
  • Premortem. When the choice is made and the team needs a sharper look at how it could fail than a list of risks gives.

Where it sits in SOLVE eX

Go to the source

  • The Rational Manager, 1965
  • The New Rational Manager, 1981

Back to the catalog