process & strategy
Divergent & convergent thinking
What are divergent and convergent thinking?
Divergent and convergent thinking are the two complementary modes design swings between. Divergent thinking opens up - generating many ideas, options and possibilities without judging them. Convergent thinking narrows down - evaluating, comparing and choosing among them. Good process deliberately separates the two, because trying to create and critique at the same time does both badly.
Also known as: divergent and convergent thinking, divergent thinking, convergent thinking
Prefer to watch?
Watch the recap 1:05
The demo
Design swings between two modes. First diverge - pile up ideas with no judgement - then converge - narrow to a choice. Run the rhythm, then see what happens if you try to do both at once.
What this demo shows (text version)
The demo runs the two modes in sequence. In the divergent phase you generate ideas freely - each tap adds another, with no judging - and the field visibly widens, rewarding quantity and breadth. In the convergent phase you cluster and choose, narrowing the many ideas down to one direction. A final illustration shows what happens when you try to do both at once: judging ideas as they're created suppresses them, so you end up with a few timid options and a stalled decision.
That's divergent and convergent thinking: the complementary modes of creating and choosing, which work best when kept separate in time. Diverge to explore widely, then converge to decide, and repeat the rhythm at each stage - it's the engine inside the double diamond and most design processes. Mix them and you do both badly.
Creating and judging are different jobs - do them separately. Divergent thinking widens the field: brainstorm freely, defer judgement, chase quantity and wild ideas, explore the whole problem and solution space. Convergent thinking narrows it: critique, group, prioritise and decide. The classic mistake is mixing them - judging ideas as they're proposed kills the divergence (people self-censor and you get a few safe options), while never converging leaves you with a pile of ideas and no decision. So sequence them explicitly: diverge first to find a rich set of possibilities, then converge to choose - and repeat the rhythm at each stage (it's the engine inside the double diamond).
Watch the rhythm: first the field opens out as ideas pile up with no judgement (divergent), then it narrows as you cluster and choose (convergent). Try to do both at once and you get the worst of each - timid ideas and no decision. Separating them is the whole trick: diverge to explore widely, then converge to decide - that alternation is how good design actually moves forward.
Divergent thinking generates possibilities: it values quantity, novelty and breadth, defers judgement, and explores widely (brainstorming, "how might we", sketching many options, considering the whole problem). Convergent thinking selects: it evaluates, compares against criteria, groups and prioritises, and commits to a direction (dot-voting, decision matrices, critique). They use different mindsets - one expansive and non-judgemental, the other critical and decisive.
The core discipline is keeping them apart in time. Judging ideas as they emerge suppresses divergence: people stop offering bold or half-formed ideas, so you converge prematurely on a few obvious, safe options and miss better ones. Equally, endless divergence with no convergence produces lots of ideas but no progress. Naming which mode you're in - "we're diverging now, hold the critique" - protects each.
This alternation is the engine of design process models. The double diamond is literally two diverge-converge cycles (explore the problem, define it; explore solutions, deliver one); design sprints, design thinking and ideation methods all build in the rhythm. Use it consciously: open up to explore the problem and the solution space, then narrow to define and decide, repeating at each phase - so you neither settle too early nor wander forever.