Generating ideas with structured techniques

The classic, and the one most often done badly. Five to twelve people, a moderator, somebody who writes everything down where all can see it, and about fifteen to thirty minutes. Four rules carry it:

Two weaknesses are worth naming, because they decide when you use a different technique. Only one person can speak at a time, so ideas are lost while waiting — Bewertungsangst and Produktionsblockierung are the German terms for the two effects. And a group with a strong personality in it produces that person’s ideas. Both are exactly what 6-3-5 removes.

Six people, three ideas each, five minutes per round, six rounds. Every round the sheet moves one seat on, and you read what is there before you write.

Why this one rather than a discussion:

The output is not 108 good ideas. It is perhaps five worth pursuing, and the other hundred are the price of finding them.

The topic in the centre, main branches radiating outwards, one keyword per branch, sub-branches for detail. Colour and small drawings are not decoration: they are what makes the map re-readable a week later.

It is the technique for a single person or a pair, and its strength is that it generates and orders at the same time — you can see which area is thin and which is overgrown. Its weakness follows from the same property: the branch you started with attracts the most detail, whether or not it deserves it. In a group it degenerates quickly, because the person holding the pen decides the structure.

In this course you meet mind maps a second time as a representation rather than a technique — the curriculum overview on the start page is one.

The systematic one, and the only one of the six that aims at completeness rather than inspiration. Fritz Zwicky’s idea: break the problem into its independent characteristics, list the possible values of each, and then read combinations across the rows.

Each row is a decision that can be taken independently of the others, so every path across the rows is a possible solution: member + phone + before training
club database is one, trainer + tablet + during + club database is another. Four rows with four, four, three and four values give 4 · 4 · 3 · 4 = 192 combinations — far more than a group would ever propose freely, and that is the point: the box makes you see the combinations nobody thought of.

Two rules keep it usable. The characteristics must be genuinely independent — if choosing a value in one row forces a value in another, they are one characteristic, not two. And five to seven rows is the practical limit; beyond that the number of combinations stops being informative.

Nature has been running a very long optimisation process, and its results are free to look at. Bionik transfers them: abstract your problem until it is a functional question, find a living system that answers that question, and transfer the principle back — not the shape.

The standard examples are worth knowing because each shows a different kind of transfer: the burdock burr became the hook-and-loop fastener (a mechanism), the lotus leaf became self-cleaning surfaces (a surface property), shark skin became riblet films that reduce drag on aircraft (a texture), termite mounds became passive building ventilation (a system principle).

The step people skip is the abstraction. Copying the shape of a bird’s wing gives you a bad wing; asking "how is lift produced with little energy" gives you something usable. In a school project the same move works on a smaller scale: "how does an anthill distribute work without a central plan" is a real question about a scheduling problem.

The odd one out: it does not produce ideas out of nothing, it sharpens an assessment that no single person can make reliably — how long something will take, how a technology will develop, whether an approach carries.

Two properties do the work. Anonymity removes the effect that the loudest or most senior voice decides — nobody knows whose estimate is whose. And the fed-back summary lets each expert revise in the light of the others' reasons rather than their status, which is why the spread usually narrows over two or three rounds.

The price is time: written rounds take days, not minutes, and the method needs people who actually know something about the question. For a school project this matters mostly as a model — the milestone estimates in this course use the same two ideas in miniature, when everybody writes an estimate down before they are discussed.