Wallas's Four Stages of Creativity

Thinking Tools4.2 Expansive Idea FormationAlso known as Preparation, incubation, illumination, and verification

Developed by Graham Wallas (The Art of Thought, 1926).

A four-stage model of creative work (preparation, incubation, illumination, and verification) that counts stepping away from a fully loaded problem as part of the work.

A stuck pricing problem taken through preparation, incubation, illumination, and verification

At each stage pick the pricing lead's move; the result shows, with whether the move fits the stage. The stage that looks like doing nothing is part of the method. Illustrative.

Pricing lead

What came of it

Pricing lead

What came of it

Select a stage to read it, or play the sequence.Preparation. The new add-on has to be priced, and three weeks of direct attempts have failed. Pick her move.Pick a move to see the reply.She ends with one sentence: per-call pricing punishes the customers who use it most.Fits the stage: the problem is loaded, and the place it sticks is named.The walk is pleasant, and nothing comes of it.Too early: there is nothing loaded yet for time away to work on.The sticking point. Preparation ends with the difficulty stated sharply, which is what the later stages work from.Incubation. Direct approaches have run out. Pick her move.Pick a move to see the reply.She does not think about pricing on purpose all weekend.Fits the stage: time away, after the problem is fully loaded.The fifth model fails where the first four did.More of the same: the direct approaches were already exhausted.Time away. Nothing visible happens in this stage, and that is the point of it.Illumination. On a Sunday run an idea arrives: charge per active seat, with calls included. Pick her move.Pick a move to see the reply.The note is waiting on Monday, reasons and all.Fits the stage: insights fade fast, so capture comes first.By Monday she remembers having an idea, and nothing of what it was.Lost: the flash is brief, and it does not come twice on request.The idea. The insight feels sudden, and it rests on the two days of preparation that named the sticking point.Verification. The idea feels right. Pick her move.Pick a move to see the reply.Ten accounts fit well; two heavy users would pay less than they cost to serve.Fits the stage: the feeling of insight is not evidence.Finance finds the two loss-making accounts on Tuesday, in public.Skipped: an untested flash is still a guess.The test. The two accounts that fail send her back to preparation with a smaller problem: a floor price for the heaviest users.

Moves that fit the stage: 0 of 4Moves that fit the stage: 1 of 4Moves that fit the stage: 2 of 4Moves that fit the stage: 3 of 4Moves that fit the stage: 4 of 4

  1. Preparation (Pricing lead). The new add-on has to be priced, and three weeks of direct attempts have failed. Pick her move.
    • Spend two days: six customer calls, every model tried, and where each one breaks. She ends with one sentence: per-call pricing punishes the customers who use it most. Fits the stage: the problem is loaded, and the place it sticks is named.
    • Skim the brief and take a long walk, waiting for inspiration. The walk is pleasant, and nothing comes of it. Too early: there is nothing loaded yet for time away to work on.
  2. The sticking point (What came of it). Preparation ends with the difficulty stated sharply, which is what the later stages work from.
  3. Incubation (Pricing lead). Direct approaches have run out. Pick her move.
    • Put the problem away for the weekend and spend Friday on the board deck. She does not think about pricing on purpose all weekend. Fits the stage: time away, after the problem is fully loaded.
    • Grind through more spreadsheet models on Saturday night. The fifth model fails where the first four did. More of the same: the direct approaches were already exhausted.
  4. Time away (What came of it). Nothing visible happens in this stage, and that is the point of it.
  5. Illumination (Pricing lead). On a Sunday run an idea arrives: charge per active seat, with calls included. Pick her move.
    • Stop and dictate it into her phone, with the two reasons it might work. The note is waiting on Monday, reasons and all. Fits the stage: insights fade fast, so capture comes first.
    • Keep running; she will remember it on Monday. By Monday she remembers having an idea, and nothing of what it was. Lost: the flash is brief, and it does not come twice on request.
  6. The idea (What came of it). The insight feels sudden, and it rests on the two days of preparation that named the sticking point.
  7. Verification (Pricing lead). The idea feels right. Pick her move.
    • Rerun last year's twelve largest accounts under the new price before telling anyone. Ten accounts fit well; two heavy users would pay less than they cost to serve. Fits the stage: the feeling of insight is not evidence.
    • Announce it at Monday's staff meeting. Finance finds the two loss-making accounts on Tuesday, in public. Skipped: an untested flash is still a guess.
  8. The test (What came of it). The two accounts that fail send her back to preparation with a smaller problem: a floor price for the heaviest users.

An example

A software company's pricing lead had three weeks to price a new add-on, and every model she tried failed the same review. She stopped trying fresh models and spent two days loading the problem: six customer calls, a list of every pricing scheme the team had considered, and a page on where each one broke. It came down to one sentence. Charging per call punished the customers who used the add-on most, and those were the ones most likely to leave.

Then she put the problem away for the weekend on purpose. On a Sunday run she had the idea of charging per active seat with calls included, and she stopped to dictate it into her phone. On Monday she tested it against last year's twelve largest accounts before telling anyone. Ten fit; two heavy users would have cost more to serve than they paid. She went back with that smaller problem and added a floor price, and the plan passed review that Friday.

Why it works

Hard problems often resist direct effort because the first framings keep coming back. Thorough preparation fills memory with the pieces and names exactly where they fail to fit; stepping away then lets the mind recombine them without the conscious habits that kept producing the same failed answers. The insight that arrives feels sudden, but it is built from what preparation loaded.

The model's value for a working team is that it makes the pause legitimate and puts limits around it. Time away counts as work only after the problem is fully loaded, and an idea counts only after it has been tested. Verification is a stage of its own because the feeling of having solved something is a poor guide to whether it holds.

The thinking behind it

Graham Wallas Graham Wallas, a social psychologist and a cofounder of the London School of Economics, described creative thought in four stages in The Art of Thought (1926): preparation, in which the problem is worked on directly; incubation, in which it is set aside; illumination, the moment the idea arrives; and verification, in which it is tested and made exact. His model remains the usual starting point for stage theories of creativity. It also makes a practical claim: the pause between hard work and the answer belongs to the work.

How to run it

  1. Work the problem until direct approaches run out. Gather the facts, try the obvious solutions, and keep going until new attempts repeat old ones.
  2. Write down where it sticks. End preparation with one or two sentences naming exactly what keeps the pieces from fitting.
  3. Step away on purpose. Sleep on it, walk, or switch to unrelated work for a set time, from a day to a week.
  4. Keep a way to capture ideas. Carry a notebook or phone, because insights tend to come during routine activity and fade within minutes.
  5. Test the idea against the problem. Check it against real cases or data before telling anyone it works.
  6. Loop back with what failed. Return to preparation with the narrower problem the test exposed, and run the stages again.

Where it goes wrong

Incubation starts too early. Stepping away from a problem that was never loaded produces rest and no insight. If the sticking point cannot be stated in a sentence, preparation is not done.

The pause has no end. Waiting for inspiration becomes a reason not to decide. Set a date to return to the problem, and go back to direct work if nothing has come by then.

The flash is trusted. Ideas that arrive suddenly feel certain, and teams ship them untested. Treat illumination as a hypothesis and let verification decide, even when the idea is your own.

When to use something else

  • Lateral Thinking. When the team cannot wait for an insight and needs deliberate techniques to break a fixed framing now.
  • First Principles Thinking. When the problem is stuck on inherited assumptions that can be listed and tested one at a time.
  • SCAMPER. When an existing product or process needs many variations quickly, from a list of prompts.

Where it sits in SOLVE eX

Go to the source

  • The Art of Thought, Jonathan Cape, 1926

Back to the catalog