Idea Generation & Brainstorming

Brainstorming was popularised by advertising executive Alex Osborn in the 1940s and 1950s, who proposed that groups should defer judgement and aim for quantity. Later research has been less kind to the group version. Studies by Diehl and Stroebe in 1987 and a meta analysis by Mullen, Johnson and Salas in 1991 found that individuals working alone often produce more ideas than the same number of people brainstorming together.

Several effects explain this. Production blocking means only one person can speak at a time, so others wait and forget ideas. Evaluation apprehension makes people censor what they say. Social loafing means individuals expend less effort in a group. Anchoring means the first ideas shape everything after.

This does not mean group ideation is useless. Groups are good for building on and combining ideas, for shared understanding and for buy in. The failure is in the usual format: a room, a whiteboard and a request to shout ideas.

A better design is to mix individual and group work. Let people generate alone first, then share, combine and build together. We cover that in a later module.

Divergent techniques deliberately widen the option space. Rather than waiting for inspiration, they create conditions in which unusual connections are likely. The common feature is that they supply something to push against: a constraint, a prompt, a role or a reversal.

Constraints help more than they hurt. Limiting materials, time, budget or form often leads to more inventive solutions than an open brief. Ask what you would do with only two colours, only one hour or only paper.

Other techniques include forced associations with a random word or image, role play such as thinking as a child or a competitor, worst idea inversion, and crazy eights, where you sketch eight ideas in eight minutes by folding a sheet into eight panels.

Record everything and keep it visible. Do not discard ideas during the session. The aim is to make a large and varied pool, from which patterns and promising seeds can be drawn later.

SCAMPER is a checklist of seven prompts for varying an existing idea or product. The letters stand for Substitute, Combine, Adapt, Modify, Put to another use, Eliminate and Reverse. The checklist builds on questions proposed by Alex Osborn and was arranged into the SCAMPER acronym by Bob Eberle in the early 1970s.

Each prompt is a question. Substitute: what could be replaced, such as a material, a person or a step. Combine: what could be merged with it. Adapt: what else is like this, and what could be borrowed. Modify: what could be magnified, minimised or changed in shape. Put to another use: who else could use it, or for what. Eliminate: what could be removed. Reverse: what if the order, roles or direction were flipped.

Apply it to something concrete. Take a water bottle and run each prompt in turn, writing at least two answers for each. You will find that the checklist forces coverage and prevents stopping at the first obvious idea. Many answers will be poor, but a few will open new directions.

SCAMPER suits improvement of existing things more than invention from nothing. Pair it with other methods such as analogy when a project needs a new concept rather than an iteration.

Analogical thinking borrows structure from one domain to solve a problem in another. A shape from nature, a process from a hospital, a rule from a game or a technique from another century can supply a pattern you would not find by looking only at your own field.

Famous cases show the idea. The hook and loop fastener was inspired by the burrs that stuck to a dog's coat, and the front of some trains was reshaped after the beak of the kingfisher to reduce noise. These are examples of biomimicry, though the stories vary in detail so treat popular tellings with care.

To use analogy deliberately, abstract the problem first. Instead of asking how to make a better queue, ask how things move through a constrained space in an orderly way. Then look for other domains that solve that abstract problem, such as water in pipes or traffic on roads.

Then translate back. What in the analogy does the equivalent job in your context. Not every transfer works, so test the useful parts and discard the rest.

Research suggests that individuals working alone generate more and more varied ideas, while groups are better at developing, combining and evaluating ideas and at creating shared ownership. The best design uses both in sequence.

A common format is brainwriting. Everyone writes ideas silently for several minutes, passes sheets to build on each other's ideas and only then discusses. This removes blocking and evaluation apprehension, and gives quieter people an equal voice.

Another is the diverge alone, converge together pattern. Spend a fixed time on individual idea generation. Then share all ideas without comment, cluster as a group, and select together. This pairs the variety of individual work with the social benefits of group decision making.

Keep sessions short and well bounded. Fatigue reduces quality, and a long session tends to drift. Set the question clearly at the start, so that the group knows what they are generating for.

Convergence is a separate skill from divergence. It needs its own methods and its own mindset. The aim is to reduce a large pool to a small set worth developing, without losing the interesting oddities too soon.

Start with clustering. Group similar ideas to see the underlying themes and to remove duplicates. Name each cluster by what it does. Then apply criteria. Typical criteria include desirability for users, feasibility to build and viability for the organisation, along with fit to the problem statement.

Dot voting is a quick way to show preferences, but it favours familiar ideas. Combine it with a structured comparison, such as a matrix of ideas against criteria, and with a deliberate wild card pick to keep some risky direction alive.

Finish by writing why the chosen ideas were selected and what was set aside. This record helps if direction changes later, and it makes the reasoning visible to people who were not in the room.

A facilitator protects the process. They set a clear question, explain the rules, keep time, manage dominant voices and keep divergence and convergence separate. They do not need to be the most senior person, and often it helps if they are not.

Preparation matters more than performance. Write a one sentence challenge, prepare prompts and materials, decide the sequence of activities and set realistic time boxes. Share the question in advance so people arrive with thoughts.

During the session, watch the energy and the participation. Use silent writing to give everyone a voice and rotate who presents. Reframe when the group gets stuck, for example by adding a constraint or switching technique, and call a break when attention falls.

Close properly. Summarise what was generated, confirm the next steps and owners, and share outputs promptly. A session without follow up trains people to expect that ideation is for show.