Double Diamond

Double Diamond Design Process

A structured facilitation guide for the Double Diamond — a four-phase framework that moves teams from fuzzy problems to validated solutions through two cycles of divergent and convergent thinking.

The Model at a Glance

The Double Diamond has two "diamonds" — each representing a diverge-then-converge cycle:

Diamond 1: Problem Space ("Design the right thing")

• Discover (diverge) — Explore the problem broadly. Research, observe, empathize.
• Define (converge) — Synthesize findings into a clear, actionable problem statement.

Diamond 2: Solution Space ("Design the thing right")

• Develop (diverge) — Generate many possible solutions. Brainstorm, sketch, prototype.
• Deliver (converge) — Test, refine, and ship the strongest solution.

The transition between divergent and convergent thinking — sometimes called the "groan zone" — is where the hard work happens. Teams must let go of weaker ideas to focus on what matters.

Core Principles

1. User-centered: Every phase roots itself in real user needs, not assumptions.
2. Diverge before you converge: Resist jumping to solutions. Explore widely first.
3. Iterate between diamonds: The process is not strictly linear — loop back when you learn something that changes the problem definition.
4. Show the work: Artifacts at each phase (research notes, problem statements, prototypes, test results) keep the team aligned.
5. Cross-functional collaboration: The best outcomes come from diverse perspectives — designers, engineers, product, domain experts, and users all contribute.

How to Use This Skill

When a user triggers this skill, figure out where they are and meet them there:

If they need help structuring a project:

1. Understand the challenge or opportunity they're facing
2. Help them plan which Double Diamond phases to run and in what depth
3. Produce a phase-by-phase project plan with activities, deliverables, and timing
4. Read references/overview.md for the full framework details

If they're in the Discover phase:

1. Help them plan research activities (interviews, observation, market research, data analysis)
2. Build interview guides, survey frameworks, or research plans
3. Read references/discover-define.md for detailed methods and tools

If they're in the Define phase:

1. Help them synthesize research into themes and insights
2. Facilitate problem statement creation (using "How Might We" or similar frames)
3. Read references/discover-define.md for synthesis methods

If they're in the Develop phase:

1. Help them plan ideation activities (brainstorming, Crazy 8s, design studios)
2. Guide prototyping strategy — what fidelity, what to test
3. Read references/develop-deliver.md for ideation and prototyping methods

If they're in the Deliver phase:

1. Help them plan usability testing and validation
2. Structure feedback collection and iteration cycles
3. Guide launch planning and success measurement
4. Read references/develop-deliver.md for testing and delivery methods

If they want a specific deliverable:

Produce the appropriate artifact — see Output Formats below.

Output Formats

Depending on the request, produce:

• Project plan — A phase-by-phase plan with activities, methods, deliverables, timeline, and team roles for all four phases
• Research plan — Discover-phase plan with research questions, methods, participant criteria, and logistics
• Problem statement — A well-framed problem definition synthesized from research, using HMW or similar formats
• Ideation workshop plan — A facilitation guide for a Develop-phase brainstorming session
• Prototype brief — What to build, at what fidelity, and what questions it needs to answer
• Test plan — Who to test with, what tasks, what success looks like, how to capture feedback
• Phase summary — A synthesis of any completed phase: key findings, decisions made, next steps

Adapting to Context

The Double Diamond is a framework, not a rigid process. Help users adapt it:

• Solo founder vs. large team: A solo founder might run the whole thing in a week with lightweight methods. A large team might spend months with dedicated research sprints.
• New product vs. feature improvement: New products need deeper Discover/Define. Feature improvements might start with existing data and move faster through Diamond 1.
• Technical vs. service design: The methods change but the structure holds. A technical product might prototype with code; a service design project might prototype with role-playing.
• Time-constrained: When time is tight, the user can run a compressed version — even a single-day design sprint follows the same diverge-converge-diverge-converge rhythm.

Always ground recommendations in the user's actual constraints — team size, timeline, budget, existing research, and domain.