About Zwicky Box
A tool for structured problem exploration using General Morphological Analysis
What is General Morphological Analysis?
General Morphological Analysis (GMA) is a method developed by the Swiss astrophysicist Fritz Zwicky in the 1960s for systematically exploring all possible solutions to a complex problem. Rather than brainstorming in an unstructured way, GMA breaks a problem down into its fundamental parameters (dimensions), lists the possible values (options) for each one, and then examines every combination.
The key insight is that most real-world problems have far too many combinations to evaluate one by one. GMA makes this manageable through Cross-Consistency Assessment (CCA) — a pairwise check that marks certain value combinations as incompatible or risky, dramatically reducing the solution space to only the internally consistent configurations worth exploring.
GMA has been used in fields as diverse as defence strategy, product design, policy analysis, and business model innovation. Zwicky Box brings this technique into a modern, AI-assisted interface.
How to use Zwicky Box
Each box follows a four-step workflow. You can move freely between steps at any time — your work is saved automatically.
Define the problem you want to explore. Set a clear goal, provide background context, specify the scope, and list the key stakeholders. This context is used by the AI when making suggestions later, so the more specific you are here, the better the results.
Break your problem into its key dimensions. Each parameter represents one independent aspect of the solution (e.g. "Revenue Model", "Target Audience", "Technology Stack"). For each parameter, list the possible values it could take. You can add these manually or use Suggest with AI to generate a starting set, then refine from there. Use the AI+ button on any parameter to expand its values.
The Cross-Consistency Assessment step shows every pairwise combination of values across different parameters. Click a cell to ask the AI to assess whether that pair of values is compatible. Press and hold a cell to manually cycle through statuses: allowed (green), flagged (yellow), or forbidden (red). You can also use Assess all with AI to evaluate every unassessed pair automatically.
Build concrete solution configurations by selecting one value per parameter. The builder warns you in real time if your selection includes forbidden or flagged pairs. Once saved, use Summarise with AI to get an analysis of each configuration's pros, cons, risks, and next steps.
Tips
- Start broad, then narrow. Begin with 4–7 parameters and 3–5 values each. You can always add more later.
- Parameters should be independent. Each parameter should represent a genuinely separate dimension of your problem. If two parameters always change together, merge them.
- Use AI suggestions as a starting point. The AI is good at generating an initial structure, but you know your problem best. Edit, remove, and add values freely.
- Don't skip the CCA step. The real power of GMA comes from systematically checking compatibility. Even a partial CCA dramatically narrows the solution space.
- Build multiple configurations. Compare 3–5 different combinations to surface trade-offs you might not have considered.