Innovation Engine

Input: $ARGUMENTS

Overview

A systematic search for non-obvious strategies using cross-domain mapping, unexplored gap search, and strategy inversion.

This procedure is mandatory for all goals to prevent the system from simply recommending the “default” or “passive” path.

Steps

Step 1: Map the Strategy Space

Define the dimensions along which strategies can vary:

1. What are the key decision dimensions? (e.g., target audience, delivery method, pricing model, timing, scale)
2. What options exist along each dimension?
3. What combinations are currently being used? (Conventional strategies)
4. What combinations are NOT being used? (Gaps = innovation opportunities)

STRATEGY SPACE:
Dimension 1: [name] — options: [A, B, C, D]
Dimension 2: [name] — options: [X, Y, Z]
Dimension 3: [name] — options: [1, 2, 3]

Used combinations: [A-X-1, B-Y-2, ...]
Unused combinations: [A-Z-3, C-X-1, ...] ← INVESTIGATE THESE

Step 2: Cross-Domain Search

Look for strategies that work in OTHER domains:

Source domains to search:

1. Nature/biology — What does evolution do here?
2. Military/warfare — What would a strategist do?
3. Games/sports — What do winners do?
4. Technology — What does an engineer do?
5. Markets/economics — What does an economist do?
6. Art/design — What does a creative do?
7. Science — What does a researcher do?

For each source domain:

1. What’s the analogous problem?
2. What strategies have been tried?
3. Which strategies won?
4. What’s the mechanism? (WHY did it work?)
5. Can the mechanism transfer?

Step 3: Strategy Inversion

For the conventional/obvious strategy:

1. What is the obvious approach?
2. What if you did the OPPOSITE on each dimension?
3. For each inversion: Is there ANY merit? (Not “is it better” — just “is there a non-zero case?”)
4. Which inversions reveal something interesting?

Step 4: Gap Analysis

From the strategy space mapping:

1. Which unused combinations are most promising?
2. WHY are they unused? (Bad idea, or just unexplored?)
   • If bad idea → understand why (may be bad for them but good for you)
   • If unexplored → investigate further
3. Which gap has the best mechanism story? (Not just “different” — different AND causally sound)

Step 5: Evaluate Novel Strategies

For each promising non-obvious strategy found:

Red flags for novel strategies:

• No clear mechanism (“it just might work” = gamble, not strategy)
• Untestable assumptions
• Easy for competitors to copy once proven
• Requires conditions that don’t exist in your context

Green flags:

• Clear causal mechanism
• Testable with small investment
• Hard to copy (requires capability, position, or insight others lack)
• Evidence from analogous domains

Step 6: Design Minimum Viable Test

For the top novel strategy:

1. What’s the smallest test that would validate/invalidate this?
2. What’s the cheapest test?
3. What result would confirm the mechanism works?
4. What result would kill the idea?
5. How long would the test take?

Step 7: Report

INNOVATION ENGINE:
Domain: [where we're innovating]
Conventional approach: [what everyone does]

Strategy space: [N] dimensions × [options] = [total combinations]
Currently used: [N] combinations
Gaps found: [N] unexplored combinations

Cross-domain strategies:
1. From [domain]: [strategy] — mechanism: [why it works]

Inversions with merit:
1. [inversion] — insight: [what it reveals]

Top novel strategy:
Name: [label]
Source: [where found]
Mechanism: [causal chain]
Key assumption: [what must be true]
Competitive advantage: [why hard to copy]

Minimum viable test:
Action: [what to do]
Cost: [resources needed]
Success criterion: [what proves it works]
Kill criterion: [what proves it won't]

Verification

• At least 3 source domains searched for cross-domain strategies
• Strategy space mapped with dimensions identified
• Empty cells (gaps) explicitly identified
• Each novel strategy has mechanism explanation
• Novel strategies validated before commitment
• Minimum viable test designed for top candidate

When to Use

• During strategy discovery for any new goal
• When the current plan feels obvious or conventional
• When you want to find a competitive advantage
• When you feel stuck in a local optimum
• → INVOKE: /cda (cross-domain analogy) for deeper cross-domain analysis
• → INVOKE: /va (variation analysis) for systematic inversion
• → INVOKE: /ma (morphological analysis) for exhaustive combination generation