Skip to content
Structured Decision Making | Part 3 of 6

Generate Alternatives: At Least Three, and Always Doing Nothing

Why a decision needs at least three alternatives including the option of doing nothing, five prompts to find them, and how the must check sorts them.

When the plant finally wrote down its question, how to cover 420,000 laser-cut blanks per year, exactly two answers existed: keep buying, or buy a large laser cell. A comparison of two options, one of which nobody wants, is an approval with extra steps. The method earns its keep in the third and the fourth option.

Why the first option is not enough

The first option is usually the one somebody already wants. It arrives finished, with a quotation and a layout. Whoever questions it seems to question the person. The second best option is then never costed, and the savings are measured against nothing.

Generating alternatives is therefore a separate step with its own time, before anything is evaluated. Techniques for producing ideas in a group are a subject of their own. For a decision, five prompts are usually enough.

Five prompts

  • Do nothing. Write it first and cost it. It is the baseline against which every saving is measured.
  • The obvious one. The proposal that already exists. It becomes one option among several.
  • A smaller version. Half the asset, part of the scope.
  • Someone else owns it. A contract, a service or consignment stock instead of capital.
  • The ambitious one. The version that shows where the limits are.

The five options of the plant

Option What it is Investment EUR
A Stay as today Two suppliers, same call-off rules 0
B 6 kW laser cell All 420,000 blanks cut in house 740,000
C 3 kW laser cell 310,000 thin sheet blanks in house, thick plate purchased 310,000
D Framework and store One larger supplier, call-off store at the plant 60,000
E 10 kW cell and tower All blanks in house, automated sheet tower 1,340,000

Option C did not exist before the prompts. It came from the question about a smaller version, and from one fact supplied by planning: 310,000 of the 420,000 blanks, 74 percent, are thin sheet that a small cell can cut. Option D came from purchasing, who knew a larger supplier willing to commit capacity under a framework contract.

The same depth for every option

An option described on one line cannot compete with one that has a quotation and a layout. Each option therefore gets the same data: investment, annual cost with its parts, floor space, capacity at two shifts, people needed, months to the first part. Where a number is missing, it is requested with a name and a date.

The must check

Before any scoring, every option is held against the four musts. The answer is yes or no, with evidence.

  • Option A fails M1. Neither supplier will commit 420,000 blanks plus 15 percent in writing. This is what made the matter a decision in the first place.
  • Option E fails M2 and M4. The sheet tower needs 14 meters of clear height, and the hall has 9.5. The cell also reaches its cost per blank only on three shifts, which would need a new shift agreement.
  • Options B, C and D pass all four musts and go into the comparison.

Excluded is not deleted

Both excluded options stay in the record, each with the must it failed and the evidence. Six months later somebody will ask why the tower was not bought, and the answer has to be findable without the people who were in the room.

Option A has a second job. It fails as a solution, but it remains the cost reference: EUR 1,757,100 per year, the purchase spend plus the carrying cost of 24 days of stock. A saving without a baseline is a number without a meaning.

What the step produced

Two options became five, and two of the five were excluded for reasons anyone can check. Of the three that remain, one did not exist a week earlier. The next part scores the three options against the weighted criteria and puts the money beside the scores.

Newsletter

One Lean idea.
Every two weeks.