8. The Double Diamond

§ 8.1What it gave us#

§The Design Council drew it in 2005 and it has been redrawn a thousand times since: two diamonds side by side, each widening then narrowing. The first is the problem. The second is the solution. Within each, a divergent phase where the team opens up, gathering, exploring, generating, and a convergent phase where it narrows to a definition or a choice. Discover, Define, Develop, Deliver.

§Its gift is rhythm. Most methods are lists of activities. The Double Diamond is a shape, and the shape says something true that lists cannot: you must widen before you narrow, and you must do it twice, once for the problem and once for the solution, and the narrowing is where the work is.

§The separation of the two diamonds is the other gift. It makes visible the mistake Part I of this whitepaper was written to prevent: jumping to solutions before the problem is defined. In the picture, that is a team leaping from the left edge of the first diamond to the middle of the second, skipping the definition that would have told them which problem they were solving.

§ 8.2The closes are the point#

§Here is the part most retellings get backwards. The divergent phases are enjoyable and they look like work: notes on walls, ideas everywhere, energy in the room. The convergent phases are uncomfortable, because converging means choosing, and choosing means leaving things out that people in the room cared about.

§So teams linger in the opens and rush the closes, and a rushed close is not a decision, it is a vote. The Double Diamond's real instruction is the opposite: spend the discomfort on the closes. The first close produces a problem statement someone could photograph. The second produces a solution with alternatives it was chosen over. Both are documents, both exclude, and both can be argued with, which is what makes them useful.

§ 8.3What it assumes#

§The original model assumes phases with room to breathe: weeks of discovery, a proper definition, exploration with time for wrong turns. It assumes that the team can afford to fully diverge before it converges. And, like Design Thinking, it assumes the organization will implement what the second diamond delivers.

§ 8.4Where it breaks at one to five, and how to bend it#

§At a startup the diamonds get compressed, and the original guide's adaptation is the right one. Run the phases as focused sprints rather than open-ended stages. Prototype early inside the second diamond rather than at its end. Set a timebox for each phase and honor it. And make the closes on sufficient information, not complete information, because complete information arrives after the runway does.

§What must not be compressed is the number of closes. A startup that skips the first close, the problem definition, because it is in a hurry has removed the one thing that stops the second diamond from being an argument about features. The shape can be small. It cannot be missing a corner.

§The Double Diamond is cited by name in Product Sprint, which runs it, and in Product Strategy, which is the final close of the whole book. Part II is done. Part III is about what holds discovery up when the method is stripped to its minimum.