Ideation

012 min

The problem ideation solves

Say a product team at a grocery delivery app learns that its couriers lose time in large apartment blocks. They reach the right building but cannot find the right door. In the first meeting about it, someone suggests a free-text notes field for customers, and the room agrees. The field ships two weeks later. Most customers leave it empty, and couriers still walk the halls.

The team did not pick a bad idea on purpose. It picked the only idea it had looked at. Nobody asked what else could work: a photo of the entrance, a pin on a floor plan, a call button that rings the customer, or a building code the courier gets on arrival. Each of these fixes a different part of the problem, and the team never saw them side by side.

A step that produces that list before anyone commits is what the design field calls ideation. The Nielsen Norman Group, a usability research firm, tells designers not to settle for their first idea, or for an idea a manager hands them. The reason is simple: you cannot choose the best option from a set of one.

The cost of skipping the step stays hidden. A team that builds its first idea sees only that idea's results. It never learns how the entrance photo would have done, because nobody tried it. For a company, this is why ideation work exists at all: it is how firms find new products, fix weak ones, and keep up with competitors who face the same problems.

023 min

How ideation works: wide first, then narrow

So how does a team get past the first idea? Ideation splits the work into two modes that run one after the other.

In the first mode, people produce as many options as they can and judge none of them. Psychologists call this divergent thinking: moving outward from one question to many possible answers. The ban on judging matters because criticism stops people from saying half-formed ideas, and half-formed ideas are often where good ones start. The goal here is quantity, not quality.

In the second mode, people sort, combine and cut the options down to a few worth testing. This is convergent thinking: moving from many answers back to one or two. Teams often group similar ideas first, a step called affinity diagramming. Then they score each group on three questions: do users want it, can the team build it, and can the business pay for it?

Keeping the two modes apart is the core of the method. When a team judges while it generates, the first idea that survives one comment becomes the answer. That is what happened to the notes field.

Ideation can be done by one person, by a team, or by a crowd. A designer sketching options alone at a desk is ideating. So is a company that asks thousands of employees or customers to send in ideas through an online platform.

The ideas themselves come in a few kinds. Some improve an existing product a little. Some join two existing ideas into one. A few are new to the field, and some turn up by chance while someone works on something else.

Where ideation sits in the design process

Ideation needs a clear problem to work on, so it comes after research. The delivery team needed to talk to couriers first, to learn whether the hard part was finding the building, the entrance or the floor. Without that research, a team generates answers to a problem it has guessed.

Many teams turn the problem into "How might we…" questions before they start. "How might we help a courier find the right door without calling the customer?" is one. Each question opens one part of the problem, and a session can work through several of them.

When ideation ends, the ideas that survive become rough prototypes. Real users try them, and what the team learns may send it back to generate more ideas. The process loops; it is not a straight line.

Why later ideas tend to be better

Quantity matters for a second reason: the more original ideas tend to come late. Creativity researchers call this the serial order effect. In 2012, the psychologists Roger Beaty and Paul Silvia asked 133 young adults to spend ten minutes listing unusual uses for an everyday object, and three raters scored every answer. Their answers grew much more creative as the minutes passed, while the number of new answers peaked in the first minute and then fell quickly. Beaty and Silvia linked the rise to deliberate effort: once the obvious answers are used up, people have to search their memory on purpose and rework what they know.

For the delivery team, the notes field was the first and most obvious answer. The meeting ended before any less obvious answer had time to appear.

032 min

Where ideation comes from

The rule of generating first and judging later is older than the design use of the word "ideation". It comes from Alex Osborn, an American advertising executive. In his book Applied Imagination, first published in 1953, Osborn set out four rules for group brainstorming: the more ideas the better, the wilder the better, build on and combine other people's ideas, and do not criticize. He also claimed that a person following these rules in a group would produce twice as many ideas as when working alone. Researchers tested that claim within a few years, and it did not hold.

Design teams took up the method in the following decades. The design firm IDEO and Stanford's d.school, founded in 2005, made it a named step in design thinking, the process they taught to students and companies. Design thinking has five modes: empathize, define, ideate, prototype and test. The Stanford d.school's design thinking guide calls ideate the mode in which a team puts its effort into producing ideas. The guide stresses range over correctness: early in a project, the aim is the widest set of options, and the best one is found later through user testing.

The word "ideation" itself draws complaints. Critics call it jargon for "thinking of ideas". Others point out that in medicine it is best known from "suicidal ideation", which means thoughts of suicide. The two uses are unrelated.

The practice is not limited to product design. Advertising agencies, architects, marketers, engineers and strategy teams run the same kind of sessions, often with the same methods.

041 min

Ideation vs. brainstorming, design thinking and innovation

Because ideation grew out of brainstorming and spread through design thinking, people often use the three words as if they meant the same thing. They do not. The axis that separates them is scope: a single method, a single stage, or a whole process.

052 min

Ideation methods and when each one helps

If brainstorming is only one method, what are the others? Teams choose from dozens, and each one fixes a different problem in a session.

There are more structured methods for special cases. Morphological analysis lists the parts of a product and every option for each part, then combines them. A gallery walk pins ideas on the walls so people can read and add to them. Six Thinking Hats has a group look at one idea from six set angles, one at a time.

For the delivery team, two of these fit well. In bodystorming, one person plays the courier and walks a real apartment block with a phone, while another plays the customer at home. The walk shows where time is lost. Analogies help too: the team could ask how hotel staff find a guest's room, or how ambulance crews find a flat at night.

072 min

Where group ideation falls short

The IDEO team was trained and had a practised facilitator. Most groups do not, and the research on ordinary group sessions is not kind to them.

Groups produce fewer ideas, not more

In 1958, Donald Taylor, Paul Berry and Clifford Block at Yale University tested Osborn's claim. People brainstormed for 12 minutes, either alone or in groups of four. The researchers then pooled the ideas of four people who had worked alone, and counted each repeated idea once. Psychologists call these pooled sets "nominal groups", because they are groups in name only. Together, these pooled individuals came up with almost double the number of distinct ideas that the real groups did. Later studies found the same gap again and again.

This is the most common misreading of ideation: that putting people in a room multiplies their ideas. In a spoken session it usually does the opposite.

Groups have social problems as well. The most senior or loudest person often speaks first, and the others drift toward that idea. In the worst cases a group slides into groupthink, where the wish to agree overrides honest judgment. This is part of what happened to the delivery team: the first idea came from the person who ran the meeting.

None of this means teams should stop meeting. It means people should produce ideas alone first and combine them as a group. Brainwriting does exactly this, because everyone writes at the same time and nobody waits.

092 min

Where ideation breaks down: choosing and AI ideas

A good session ends with a wall of options. The harder step comes next: picking one.

Choosing the best idea is harder than producing it

In 2006, Eric Rietzschel, Bernard Nijstad and Wolfgang Stroebe at the University of Groningen gave groups a task that included both producing ideas and choosing the best. Some people worked alone and some in interacting groups. The people who worked alone produced more ideas, as earlier studies predicted. But the ideas chosen at the end were no better in any condition. Their final choices were not reliably better than picking at random.

So a long list of ideas does not lead to a better product by itself. The choosing step needs as much care as the generating step.

Each writer gained, but the group as a whole produced a narrower range. For a design team, AI suggestions can lift weak ideas and also pull everyone toward the same answers. Generating ideas without the tool first, then using it to add to the list, keeps the range wide.

Signs a session went wrong

• The first idea named in the room is the one that gets built.
• The same two or three people produced most of the ideas.
• The list stops growing within the first few minutes, and nobody pushes past it.
• Every idea on the wall looks like a competitor's product.
• The team cannot say why the chosen idea beat the others.

102 min

How to run an ideation session

The evidence points to a session that looks different from an open group discussion. Here is how the delivery team ran its second attempt.

In the delivery team's next session, six people each wrote ideas on their own notes for ten minutes, and then all the notes went up on one wall before anyone spoke. The group then read the wall, built on what it saw, and only after that started to sort.

1. 1

   Prepare.

   Share the research and the problem statement before the session, and write three to five "How might we…" questions. Book a quiet room or set up a shared online whiteboard for remote teams. Fix the time and plan short breaks.

2. 2

   Invite a mix of people.

   Bring in design, engineering, support and, where possible, people who live with the problem, such as couriers. Different backgrounds produce different ideas.

3. 3

   Set ground rules.

   A facilitator explains that nobody's idea is judged during generation, keeps time, and asks the most senior person to speak last.

4. 4

   Generate alone, then together.

   Start with brainwriting or Crazy Eights, then share. When the flow slows, keep going for a few more minutes, because later ideas tend to be more original.

5. 5

   Capture everything.

   One idea per note, and a photo of every wall before anyone tidies it.

6. 6

   Choose with stated criteria.

   Cluster the notes, then score each cluster on user value, ease of building and business cost, with originality as its own column. Vote, and keep at least one bold option.

7. 7

   Measure what happens next.

   Count how many people took part and how many ideas came in. Then track how many reached a prototype and how those prototypes did with users.

?4 questions

Questions people ask

§8 sources

Sources