How to Ideate: Techniques That Beat Brainstorming

In 1987 Michael Diehl and Wolfgang Stroebe compared two kinds of group. In one, four people brainstormed together out loud. In the other, four people worked alone in separate rooms and their lists were pooled afterward. The pooled lists had roughly twice as many ideas, and judges rated more of them as good. A 1991 meta-analysis by Brian Mullen and colleagues found the same gap across twenty studies. Alex Osborn, the advertising executive who popularized brainstorming in the 1950s, had promised the opposite.

The reason most teams still brainstorm is that it feels productive. The room is loud and the whiteboard fills up, and everyone leaves having said something. The ideas that make it onto the board, though, are mostly the ones the most confident people thought of first.

The format matters more than the intent, and a few well-tested formats fix the specific ways a brainstorm breaks.

What it means to ideate

To ideate is to generate possible solutions to a problem you've already defined, before deciding which ones are any good. In the design thinking process, ideation is the third stage: it takes the problem statement written in the define stage and turns it into a spread of competing approaches for the prototype stage to make concrete.

Two rules separate ideation from an ordinary meeting about solutions:

1. Generation and evaluation happen at different times. Nobody judges an idea while ideas are still being produced. Criticism, even mild, makes people hold back the strange ones.
2. Quantity comes before quality. Early ideas are usually obvious ones. Teams that stop at five ideas pick from the most predictable five.

The input matters as much as the method. If the define stage produced a vague goal like "improve onboarding," any technique will produce vague ideas. A problem statement that names who is stuck and where ("new admins leave during workspace setup because they don't know which settings they can change later") gives people something specific to push against.

Why group brainstorming underperforms

Researchers have identified three mechanisms, and each technique below targets at least one of them.

Production blocking. Only one person can talk at a time. While you wait for a gap, you either forget your idea or spend the time rehearsing it instead of thinking of the next one. Diehl and Stroebe found this was the largest single cause of the gap.

Evaluation apprehension. People censor ideas they expect others to judge, especially with a manager present. The rule "no criticism" doesn't remove the fear; it only removes the spoken criticism.

Anchoring on early ideas. The first few ideas said aloud set the frame. Later contributions cluster around them, which is a version of what researchers call design fixation, covered in more detail in functional fixedness in design.

A fourth problem is less studied but easy to see in practice: social loafing. In a group of eight, it's easy for three people to contribute almost nothing and still feel they took part.

Brainwriting and the 6-3-5 method

Brainwriting removes the talking. Everyone writes ideas silently, then the ideas are shared and built on in writing.

The most structured version is 6-3-5, published by Bernd Rohrbach in 1969. Six people each write three ideas on a sheet in five minutes. Each sheet then passes to the next person, who adds three more, either new ones or extensions of what's already there. After six rounds, the group has up to 108 ideas in about half an hour, and every idea has been seen and extended by several people.

It works because it removes production blocking entirely (everyone writes at once) and reduces evaluation apprehension (nobody watches you write). The sheet-passing keeps the benefit groups do offer, which is building on other people's ideas.

For remote teams, a shared document with one column per person does the same job. Set a timer, keep cameras on but microphones off, and rotate columns between rounds.

Crazy 8s

Crazy 8s is a sketching exercise from the Google Ventures design sprint. Each person folds a sheet of paper into eight panels and sketches eight different ideas in eight minutes, one minute per panel.

One minute is too short to draw anything well, which is the point. The time limit pushes people past their first two or three ideas, and the sketch format forces a decision about what the interface would actually show. A written idea like "make the pricing clearer" can mean anything; a sketch has to commit to a layout.

Crazy 8s works best on a problem with a visible interface, such as a screen, a flow, or a notification. It's less useful for service or policy problems, where brainwriting fits better.

How Might We questions

A "How Might We" question reframes a problem statement into an open prompt. "New admins leave during workspace setup" becomes:

• How might we let admins skip setup and come back to it?
• How might we show which settings are safe to change later?
• How might we make the first workspace useful before it's configured?

The format came out of Procter & Gamble in the 1970s, through the creativity researcher Min Basadur, and IDEO later made it standard practice in design teams. Each wording points toward a different family of solutions, so writing ten HMW questions before generating any ideas is itself a way to widen the range.

The risk is writing questions that already contain the answer ("How might we add a setup checklist?"). A useful HMW leaves room for at least three unrelated solutions.

SCAMPER

SCAMPER is a checklist for changing something that already exists. Bob Eberle assembled it in 1971 from an earlier list of Osborn's. Each letter is a prompt:

• Substitute: replace a component, material, or step.
• Combine: merge two features or steps into one.
• Adapt: borrow a pattern from somewhere else.
• Modify: make something bigger, smaller, faster, or more frequent.
• Put to another use: use an existing feature for a different job.
• Eliminate: remove a step entirely.
• Reverse: change the order, or have the other party do the step.

Applied to the workspace-setup problem, "Eliminate" suggests creating the workspace with defaults and no setup at all; "Reverse" suggests letting a teammate invite the admin into an already configured workspace. SCAMPER suits redesign work, where there's an existing flow to take apart.

Analogous inspiration

Analogous inspiration means studying how a different field solved a structurally similar problem. A team designing a hospital check-in might observe how airline gates or hotel lobbies handle queues and uncertainty. A team working on account recovery might look at how banks verify someone over the phone.

The analogy helps because it bypasses the conventions of your own category, which is where a team that studied only its competitors ends up. A useful starting prompt is to name the core tension in your problem ("people need to start before they understand the whole system") and list three places outside software where people face it.

The worst possible idea

Ask the group for the worst ways to solve the problem: the onboarding that loses every user, the error message that makes people cry. People find bad ideas easy and funny to produce, so evaluation apprehension mostly disappears. Then invert each one. "Ask for a credit card before showing anything" inverts to "show a working example before asking for anything." It makes a good warm-up for a group that includes senior stakeholders, because everyone talks before ideas start carrying status.

Choosing which ideas survive

Generating a hundred ideas is easy to overdo if nothing narrows them.

Cluster first. Group similar ideas so the team votes on directions, not on near-duplicates that split the vote.

Vote silently. Dot voting, where each person gets three to five dots, works if everyone votes at the same time. Sequential voting lets the first voters anchor everyone else, which brings back the brainstorm's original problem.

Score against the problem statement. For each surviving cluster, ask whether it addresses the specific user and moment named in the define stage. Ideas that only address a general goal drop out here.

Keep one wild card. Carry one unusual idea into prototyping alongside the safe ones. It's cheap to test at the paper-prototype stage.

A 60-minute ideation session

For a group of four to eight with a written problem statement:

1. 5 min: Read the problem statement and any research summary aloud. Answer clarifying questions only.
2. 10 min: Each person writes How Might We questions silently. Cluster them and pick two or three to work on.
3. 5 min: Worst possible idea, out loud, as a warm-up.
4. 15 min: Brainwriting on the chosen HMW questions, rotating sheets every five minutes.
5. 8 min: Crazy 8s on the most promising directions, if the problem has an interface.
6. 10 min: Cluster everything on a wall or board and dot-vote silently.
7. 7 min: Pick two or three directions to prototype and one wild card. Assign owners.

Facilitators should hold back their own ideas until the brainwriting rounds, since the person running the room has the most anchoring power in it.

FAQ

What does ideate mean?

To ideate means to generate ideas, usually many of them, before judging any. In design work it refers to the stage after a problem has been defined, when the team produces competing solutions for later prototyping.

What is the ideation stage of design thinking?

It's the third of five stages: empathize, define, ideate, prototype, and test. Its job is to turn a defined problem into a range of possible approaches, so the team isn't stuck testing the first solution anyone thought of.

Is brainstorming the same as ideation?

Brainstorming is one ideation technique, the one where a group generates ideas out loud. Ideation is the broader activity and includes silent and individual methods like brainwriting, Crazy 8s, and SCAMPER, which tend to produce more ideas per person.

Why is brainwriting better than brainstorming?

Because everyone generates at once instead of waiting to speak, and nobody watches anyone else write. That removes production blocking and most of the fear of judgment, the two main reasons groups produce fewer ideas than the same people working alone.

How many ideas should an ideation session produce?

There's no fixed number, but a session that ends with fewer than about twenty ideas has probably stopped at the obvious ones. Techniques like 6-3-5 can produce over a hundred in half an hour, and most of the value comes from narrowing that set well.

Can you ideate alone?

Yes, and research on nominal groups suggests individuals working separately and pooling their ideas outperform the same people brainstorming together. Solo ideation misses the chance to build on others' ideas, which is why brainwriting combines silent individual work with passing ideas around.

Where to go next

The design thinking hub collects the guides on every stage, from the empathize stage to an overview in the complete guide to design thinking.

Sources: Diehl, M. & Stroebe, W. (1987), "Productivity loss in brainstorming groups: Toward the solution of a riddle," Journal of Personality and Social Psychology · Mullen, B., Johnson, C. & Salas, E. (1991), "Productivity loss in brainstorming groups: A meta-analytic integration," Basic and Applied Social Psychology · Osborn, A. (1953), Applied Imagination · Rohrbach, B. (1969), "Kreativ nach Regeln: Methode 635" · Eberle, B. (1971), SCAMPER: Games for Imagination Development.