The Double Diamond

Theory & origin

The British Design Council published the Double Diamond in 2004 as a plain-language map of the design process, drawn from studying how leading design teams actually work. The four phases are Discover, Define, Develop, and Deliver, grouped into two diamonds. The first diamond is the problem space: Discover diverges to gather unfiltered insight about what is really going on, and Define converges to a single sharp problem statement worth solving. The second diamond is the solution space: Develop diverges to generate many possible answers, and Deliver converges to test, refine, and ship one. The shape carries the whole argument. Divergence is the deliberate widening, resisting the urge to decide too early. Convergence is the narrowing, the willingness to kill options and commit. The two pinch points are the model spine, a defined problem at the middle join and a delivered solution at the end. The honest caveat is that real projects are never this tidy. Teams loop back, the diamonds overlap, and discovery keeps happening late. Later versions wrapped the diagram in a set of principles, be people-centred, communicate visually, collaborate, and iterate, precisely because the clean shape tempted people into treating it as a rigid linear gate. Used as a lens rather than a gantt chart, it does one thing very well. It makes visible the phase teams most love to skip, which is defining the right problem before rushing to solve one.

Key components

The parts at a glance. Click any term for the full definition, a field example, and the common failure, in the model below.

How a consultant runs it

1. 01 Force the first diamond before anyone proposes a solution. The most expensive mistake is a beautifully delivered answer to the wrong problem, and it stays invisible until launch.
2. 02 Make divergence safe and convergence explicit. Name out loud when the team is widening, where no idea is bad yet, and when it is narrowing, where options get cut, because doing both at once produces mush.
3. 03 Write the Define statement as a single sentence the whole team can repeat. If it needs a paragraph, the problem is not defined yet, it is still being discovered.
4. 04 Treat the middle join as a real gate. Nothing enters the second diamond until the problem is agreed, or the solution work is building on sand.
5. 05 Expect to loop, and say so. Discovering a new insight during Develop is not failure, it is the model working, as long as the team goes back and re-checks Define.

Worked example

A multifinance lender is losing applicants partway through its new online loan application, and the first instinct is to redesign the whole flow. A Double Diamond reframes the work. Discover: the team stops guessing at the dashboard and sits with a dozen applicants who dropped out, watching where each one gave up. Define: the scattered friction resolves to one sentence, the KYC photo-upload step fails on low-end Android phones with weak cameras and weak signal, which is where most abandonment clusters. That single problem statement is the middle pinch, and nothing else gets solved until it is agreed. Develop: with the problem fixed, three fixes are prototyped and tested against a deliberately bad phone, on-device image compression, a retry queue for flaky uploads, and a fallback to manual review. Deliver: compression plus the retry queue ships, drop-off at the KYC step falls sharply, and the other prototypes are shelved with notes. The full-flow redesign the team almost built, of screens that were never the problem, is the second diamond they were saved from wasting.

Common pitfalls

1. 01 Skipping the first diamond, so the team delivers a strong solution to an unexamined problem
2. 02 Treating it as a rigid linear gate, when real projects loop back and the diamonds overlap
3. 03 Diverging and converging at the same time, producing neither real exploration nor a real decision
4. 04 A Define statement so vague or plural that the team is still solving several problems in the second diamond
5. 05 Delivering without validating against the defined problem, launching something polished and off-target

Sample deliverable

One real engagement, start to finish. Watch the numbers travel from raw input, onto the chart, into the finished artifact.