Novelty Space Search
Overview
Creativity = novelty + value. This procedure separates the two:
- Generate novelty (systematic combination, variation, import)
- Filter for value (criteria-based selection)
By separating, we can be exhaustive in generation and rigorous in selection.
Goal
Systematically generate novel ideas/possibilities, then search for those that are valuable, feasible, and timely.
Steps
Step 1: Define the Innovation Space
Specify:
- What domain/problem
- What hard constraints
- What would make an idea valuable
Output: Innovation brief
Step 2: List Existing Solutions
What already exists in this space? These are the baseline to be more novel than.
Output: Existing solution inventory
Step 3: Apply Combinatorial Generation
List key elements/components. Systematically combine in new ways.
Output: Combinatorial ideas
Step 4: Apply Attribute Variation
For existing solutions, vary each attribute. Use SCAMPER or similar.
Output: Variation ideas
Step 5: Apply Analogy Import
Search other domains for analogous problems. Import their solutions.
Output: Imported ideas
Step 6: Apply Constraint Removal
List assumed constraints. Generate ideas that violate them.
Output: Constraint-free ideas
Step 7: Apply Extreme Parameters
Push key parameters to extremes. What does this enable?
Output: Extreme parameter ideas
Step 8: Compile All Ideas
Combine all generated ideas. Remove duplicates. State each clearly.
Output: Master idea list
Step 9: Quick Novelty Filter
For each idea, quick check:
- Already exists? → Remove
- Obvious? → Flag but keep
Output: Novel ideas
Step 10: Score on Value Criteria
For each novel idea, score on criteria:
- Solves real problem
- Better than alternatives
- Technically feasible
- Resource feasible
- Right timing
- Sustainable advantage
Output: Scored ideas
Step 11: Rank and Select
Rank by total score. Select top ideas for further development.
Output: Promising ideas
When to Use
- Innovation/ideation
- Finding new approaches
- Breaking out of local optimum
- Exploring possibility space
- Product/feature development
Verification
- Multiple generation methods were applied
- Ideas are genuinely novel (not already done)
- Ideas address real problems
- Feasibility was honestly assessed
- Top ideas have clear next steps