"The Double Diamond is a visual representation of the design and innovation process. It's a simple way to describe the steps taken in any design and innovation project." — British Design Council
Activation Conditions and Trigger Words
- Direct invocation: "Use the Double Diamond to analyze...", "discover define..."
- Semantic trigger: the user faces a vague, complex problem and needs a structured process, and is unsure how to go from problem to solution
- Combined invocation: "First use the Double Diamond to run the full process, then use the empathy method of design thinking for Discover"
Method Framework Overview
The core of the Double Diamond is two rounds of structured exploration through divergence and convergence. By first exploring the problem space and then the solution space, and by diverging and then converging in each space, it ensures that the right problem is solved and that it is solved in the right way.
Core Principles (3-7, each must be accompanied by ≥2 pieces of cross-domain evidence)
Principle 1: Diverge first, then converge is the basic rhythm for solving complex problems
One-sentence definition: Don't rush to find an answer—first discover as many possibilities as possible, then focus your energy on selecting the optimal solution.
Cross-domain evidence：
- Creative design: At the start of a project, IDEO does not draw but instead broadly observes user behavior (diverge), then converges on the most inspiring insights (converge) (Source: Kelley, The Art of Innovation, 2001)
- Scientific research: Scientific discovery follows the cycle of "generating hypotheses (diverge) → designing experiments to confirm/falsify (converge)" (Source: Popper, The Logic of Scientific Discovery, 1934)
How to apply: Before each stage begins, explicitly ask "Are we now in the divergence or convergence stage?" During divergence, evaluation is prohibited; during convergence, divergence is prohibited.
Limitation: Divergence takes time. In extremely time-constrained scenarios, divergence may be skipped and convergence may be done directly (that is, judging by experience and intuition).: Divergence takes time. In extremely time-constrained scenarios, divergence may be skipped and convergence may be done directly (that is, judging by experience and intuition).
Principle 2: The problem space and the solution space need to be explored separately
One-sentence definition: Spending time finding the right problem is more valuable than quickly solving the wrong problem.
Cross-domain evidence：
- Corporate transformation: BBC's digital transformation first spent 6 months understanding "what audiences truly need in the digital age" (problem space), then designed digital products (solution space) (Source: BBC, Digital Transformation Case, 2010)
- Product strategy: Before redesigning, Vettery spent a great deal of time studying that "the most painful thing for hiring managers is not finding resumes but screening them" (redefining the problem), and then redesigned the product (Source: Vettery Product Redesign, 2017)
How to apply: Before you start thinking about solutions, first confirm that you have truly identified the right problem. Rewrite the problem in HMW format ≥3 times.
Limitation: Some scenarios (such as urgent bug fixes) cannot afford two full rounds of exploration.: Some scenarios (such as urgent bug fixes) cannot afford two full rounds of exploration.
Principle 3: Redefining the problem is more valuable than hastily giving a solution
One-sentence definition: Most companies fail not because they solved the problem incorrectly, but because they worked on the wrong problem.
Cross-domain evidence：
- Transportation: Uber did not solve "how to get more people to take taxis" but instead redefined it as "how to let anyone obtain reliable transportation at any time" (Source: Kalanick interviews, 2012-2015)
- Accommodation: Airbnb does not solve "how to make hotels cheaper" but redefines it as "how to let people find a place to stay anywhere" (Source: Chesky, Airbnb Origin Story, 2016)
Application method: Before you provide any solution, redefine the problem in three different ways. Choose the boldest definition that has logical support.
Limitation: Over-redefinition may lead to procrastination. A "perfect definition of the problem" is unattainable; after sufficient definition, one needs the courage to act.
Principle 4: Each of the four stages has a different mode of thinking and output
One-sentence definition: Discover→Define→Develop→Deliver is not an assembly line, but a switch between cognitive states.
Cross-domain evidence：
- Design consulting: IDEO uses different criteria to evaluate outputs at the end of each stage (the Discover stage uses depth of insight, the Define stage uses clarity of the problem, the Develop stage uses breadth of solutions, and the Deliver stage uses feasibility) (Source: IDEO, Human-Centered Design Toolkit, 2011)
- Software development: Agile's "discovery-delivery" dual-track system is essentially an engineering implementation of the Double Diamond model (Source: Sy, Dual-Track Agile, 2016)
Application method: Before entering each stage, first ask, "What is my output at this stage? How do I judge that this output is good enough?"
Limitation: In reality, there is a great deal of jumping back and overlap between stages. The Double Diamond is a framework, not an operation manual.
Operational Protocol (Agentic Protocol)
Step 1: Problem classification
Applicable signals：
- Facing a complex, vaguely defined problem
- There are multiple possible directions but uncertainty about which is best
- The team needs structure between "doing the right thing" and "doing things right"
Not applicable signals：
- The problem is simple and direct and only requires execution (such as fixing a known bug)
- Extremely urgent crisis handling
- A task with an already clear solution
Step 2: Double Diamond analysis
Research dimensions(each dimension must be traced back to a core principle):
-
Discover stage (Discover): Diverge to explore the problem space
- Source principle: Principle 1 (divergence-convergence) + Principle 2 (problem vs. solution space)
- Method of operation: broadly collect user data, conduct interviews, observe behavior, research the market—do not evaluate, only collect
-
Define stage (Define): Converge to lock onto the real problem
- Source principle: Principle 3 (redefine the problem)
- Method of operation: induce the core problem from the insights found in Discover, and define it in one sentence
-
Develop stage (Develop): Diverge to explore the solution space
- Source principle: Principle 1 (divergence-convergence) + Principle 2 (solution space)
- Method of operation: generate as many solutions as possible for the defined problem (20+), without evaluating
-
Deliver stage (Deliver): Converge to lock onto the final solution
- Source principle: Principle 4 (four-stage modes of thinking)
- Method of operation: screen solutions using user feedback and feasibility, choose the best, and conduct small-scale testing
Step 3: Double Diamond output
- Discover takeaways: 3-5 core user insights
- Define outcome: one-sentence problem definition (HMW format)
- Develop takeaways: Top 5 solution concepts
- Deliver plan: testing plan and iteration plan
Applicability/non-applicability judgment
| Problem type | Applicability | Description |
|---|---|---|
| Complex product definition | High | Classic scenario |
| Service design innovation | High | The Double Diamond is essentially a service design tool |
| User experience redesign | High | Suitable for a UXR-driven process |
| Simple execution of a known solution | Low | No need for two rounds of exploration |
| Urgent operational problem | Low | The Double Diamond is too slow |
Classic case library
Success cases
-
BBC digital transformation(Media/2010)
- Problem: How to keep the BBC relevant in the digital age
- Method application: Discover (audience research) → Define (lock in "personalized news experience") → Develop (multiple digital product prototypes) → Deliver (products such as iPlayer)
- Result: The BBC became one of the most successful traditional media organizations in digital transformation
- Source: BBC Digital Transformation Case Study (2010)
-
Vettery product redesign(Recruiting tech/2017)
- Problem: Poor user experience on the recruiting platform
- Method application: The Define stage spent a large amount of time redefining the problem → discovered that the real pain point was "screening"
- Result: After the redesign, user engagement increased significantly
- Source: Vettery Product Redesign Case Study (2017)
Failure cases
- The cost of skipping the Define stage(Product design/general)
- Problem: The team jumped directly from Discover to Develop
- Misuse method: After returning from research, they started drawing prototypes directly without any convergence
- Lesson: A Double Diamond without a Define stage equals a single diamond. The convergence of the first diamond (Define) is the most underestimated but most critical step
- Source: Design Council, Eleven Lessons (2015)
Misuse detector (≥3 misuse patterns)
| Misuse signal | Detection logic | Warning message | Recommended action |
|---|---|---|---|
| Method does not match the problem | Users have clear needs and only need execution | "You have a clear task and do not need Double Diamond exploration. Using agile iteration directly is faster." | Recommend agile |
| Skipping key steps | Entering Develop directly from Discover and skipping Define | "You skipped Define—the convergence of the first diamond. Developing solutions without clearly and correctly defining the problem is the highest-risk design error." | Guide back to Define |
| Stage confusion | Starting to evaluate solution feasibility during the Discover stage | "Evaluation is prohibited during the Discover stage. Your task is only to collect information. Evaluation happens in the Define and Deliver stages." | Guide correct stage behavior |
Honest boundaries (≥3 specific limitations)
- Not suitable for simple linear problems: fixing bugs and routine operational optimization do not require two rounds of divergent-convergent thinking
- Long time cycle: a complete Double Diamond takes weeks to months, making it unsuitable in scenarios with high iteration speed requirements
- High team requirements: the Double Diamond requires the team to possess multiple capabilities such as user research, insight synthesis, and prototyping
- Stages are not strictly linear: in reality, it loops back multiple times. Treat the Double Diamond as a framework rather than a strict Gantt chart
Research sources (primary sources must account for >50%)
| # | Source | Type | Primary/Secondary | Weight |
|---|---|---|---|---|
| 1 | British Design Council, The Double Diamond (2005) | Original model | Primary | High |
| 2 | IDEO, Human-Centered Design Toolkit (2011) | Method guide | Primary | High |
| 3 | Design Council, Eleven Lessons (2015) | Research report | Primary | High |
| 4 | BBC Digital Transformation Case (2010) | Case | Primary | Medium |
| 5 | Sy, Dual-Track Agile (2016) | Article | Primary | Medium |
| 6 | Kelley, The Art of Innovation (2001) | Book | Secondary | Medium |
This Skill was generated by Forge Skill — Forging Thinking Tools Generated