Double Diamond process
A visual map of the design process that separates finding the right problem from finding the right solution, using alternating cycles of divergent (open, explore) and convergent (close, decide) thinking.
Created by the British Design Council in 2005 after studying how design teams at Sony, Starbucks, LEGO and others actually worked. Refreshed in 2019 as the Framework for Innovation, adding engagement, leadership and continuous feedback loops around the diamonds.
1. The core mental model
Two diamonds, four phases:
flowchart LR
A(( )) --> B[Discover
diverge]
B --> C[Define
converge]
C --> D[Develop
diverge]
D --> E[Deliver
converge]
E --> F(( ))| Diamond 1 — Problem space | Diamond 2 — Solution space | |
|---|---|---|
| Question | Are we solving the right thing? | Are we solving it the right way? |
| Phases | Discover → Define | Develop → Deliver |
| Output | A problem statement | A validated solution |
| Failure mode | Solving a symptom | Shipping an unusable answer |
2. Phase by phase
2.1 Discover - diverge on the problem
What´s really going on?
Understand the situation as it actually is, not as the brief describes it. You are collecting evidence, not opinions.
This is the time to clear your head. You still don’t fully understand what the real problem is, so you do some research. You talk to users, look at the competition, read data, and observe behaviors.
2.2 Define - converge on the problem
What exactly is the problem?
Synthesise the mess into one sharp, agreed problem worth solving.
After all that exploration, it’s time to close out the first diamond. You take everything you’ve learned and distill it into a single thing: a clear definition of the problem worth solving.
This is known as a Problem Statement or POV (Point of View). It sounds academic, but it’s the most important moment in the process. If you get this wrong, everything that comes after will be off track - no matter how beautiful the UI turns out to be.
2.3 Develop — diverge on the solution
What paths?
Generate many possible answers to the now-fixed problem. Quantity before quality, multiple concepts before refining one.
The second diamond begins, and it starts with another round of exploration. Now that you have a clear understanding of the problem, you explore possible solutions. You sketch, prototype, and discard ideas. You test both wild and conservative ideas. You make mistakes quickly so you can learn faster.
If in Phase 1 you explored to understand the problem, here you explore to find paths to a solution. And you’re going to come up with many. Most won’t work and that’s exactly what needs to happen.
Force at least one deliberately "wrong" concept — the cheapest possible version, or the one that removes the feature entirely. It calibrates the team on what value the other concepts are actually adding.
2.4 Deliver — converge on the solution
Does this solution, well executed, solve the problem?
Test, refine, kill what fails, ship what works, and measure.
You complete the second diamond. You choose the solution that best addresses the problem, refine it, validate it with users, and prepare it for the development team to build.
Here, the handoff with the developers takes place, final details are adjusted, and the workflows are documented. This is the time for UI kits, design systems, and high-fidelity prototypes.
3. The four Design Council principles
- Be people-centred. Start with the humans using it.
- Communicate visually and inclusively. Shared understanding beats individual brilliance.
- Collaborate and co-create. Design with, not for.
- Iterate. Treat every phase as reversible, so learning is the deliverable.