Generate innovative ideas by creating new dependencies between previously independent variables. Part of SIT (Systematic Inventive Thinking) ideation toolkit. Map internal variables (controlled) and external variables (uncontrolled), then link pairs that were never connected before to discover non-obvious innovations. Can be used standalone or as part of ideation-sit orchestration.
SIT: Variable Dependency - Create New Relationships Between Variables
Purpose
Generate innovative ideas by discovering what happens when you create new dependencies between variables that are currently independent. This technique produces "smart" products and services that adapt, respond, and behave differently based on changing conditions—innovations that customers could never have articulated wanting.
When to Use This Skill
Use this skill when:
You want to make a product/service that adapts to context, users, or conditions
You're looking for "intelligent" product behaviors
Current product treats all users/situations the same (one-size-fits-all)
You want innovations that feel magical ("how did it know to do that?")
You need ideas beyond what SCAMPER modification provides
Part of a comprehensive SIT session (use ideation-sit orchestrator)
The Variable Dependency Technique
相关技能
Core question: What if [Variable A] changed as a function of [Variable B]?
Philosophy: Most products have variables that are fixed or independent of each other. By creating new dependencies between them, you create products that respond to their environment in ways no one has seen before. The innovation isn't adding something new—it's creating a new relationship between things that already exist.
Understanding Variables
Internal Variables (Controlled by the Maker)
These are attributes of the product/service that you set and control:
List external variables (aim for 8-12):
Go through external components and environmental factors.
Example for a lending app:
User's income pattern, time of day, user's repayment history, local market conditions, user's phone type, connectivity level, user's location, weather (affects market traders), user's age, user's business type
Step 2: Build the Dependency Matrix
Create a matrix with internal variables as rows and external variables as columns. Each cell represents a potential new dependency.
You don't need to fill every cell. Scan for interesting pairs.
Step 3: Create Virtual Products
For each promising pair, create a "virtual product" by stating the dependency:
Formula: "[Internal variable] changes as a function of [External variable]"
Then apply Function Follows Form:
What would this product look like?
How would it behave?
Who would benefit?
What problem does it solve?
Step 4: Also Consider Internal ↔ Internal Dependencies
Some of the most interesting innovations come from linking two internal variables:
Does loan amount affect interest rate? (Maybe it should, but inversely to what's expected)
Does notification content change based on notification frequency?
Does the approval process change based on loan size?
Step 5: Consider Breaking Existing Dependencies
Look at current dependencies and ask: "What if these were independent?"
What if price didn't depend on cost?
What if availability didn't depend on inventory?
What if access didn't depend on location?
Step 6: Generate 8-12 Ideas
Push beyond obvious dependencies. The most innovative ideas come from unlikely variable pairs.
Step 7: Capture and Document
Record each idea with:
The dependency created (or broken)
The variables involved and their types
The value discovered
Prompting Questions
Use these to stimulate thinking for each variable pair:
Creating dependencies:
What if [product attribute] automatically adjusted based on [user condition]?
What if [feature intensity] responded to [environmental factor]?
What if [business term] varied with [user behavior]?
What if [content] adapted to [context]?
What if [process step] changed based on [time factor]?
Breaking dependencies:
What if [currently linked attribute] was completely independent of [its current driver]?
What if users could control [something currently automatic]?
What if [something that varies] became fixed?
What if [something restricted by X] was freed from that constraint?
Unusual pairs:
What's the most unlikely pair on my matrix? What would happen if I linked them?
What if a customer-facing attribute depended on an operational variable?
What if an operational parameter adapted to an emotional state?
Examples from Case Studies
FairCredit (Nigeria Micro-Lending)
Dependency #1: Interest rate ↔ repayment streak
Type: Internal-Internal
Virtual Product: Interest rate automatically decreases after consecutive on-time repayments. Resets after a late payment.
Value: Self-rewarding system. Borrowers are motivated to maintain streaks. Loyal borrowers get better rates without negotiating. Risk-adjusted pricing happens automatically.
Dependency #2: Loan amount available ↔ time of day
Type: Internal-External
Virtual Product: Maximum available loan is higher during early morning hours (5-7 AM) when market traders buy stock, lower during evening.
Value: Matches real cash-flow needs. Traders get capital when they need it most. Reduces frivolous borrowing during off-hours.
Dependency #3: SMS notification language ↔ user's phone language setting
Type: Internal-External
Virtual Product: System auto-detects phone language and sends loan terms, reminders, and tips in Yoruba, Igbo, Hausa, or English accordingly.
Value: Non-English speakers understand their obligations. Reduces disputes. Builds trust with wider population.
Dependency #4: Repayment flexibility ↔ market day schedule
Type: Internal-External
Virtual Product: Repayment due dates automatically align with local market days (when traders have cash) rather than fixed calendar dates.
Value: Payments fall on high-cash-flow days. Default rates drop. System feels designed for their reality.
Virtual Product: When team members are absent from a meeting, the recording automatically generates a more detailed summary with action items highlighted for those who missed it.
Value: Absent team members (common with 6+ hour time zone spreads) get richer catch-up content. Present members get brief notes. Same meeting, different outputs.
Dependency #3: Interface language ↔ content being discussed
Type: Internal-Internal
Virtual Product: When discussing a document in Thai, the interface tools and prompts appear in Thai. When switching to an English document, interface follows.
Value: Multilingual teams work in their most comfortable language per task. No global language setting needed.
SolarSeva (Rural India Home Solar)
Dependency #1: Payment amount ↔ energy generated that month
Type: Internal-External
Virtual Product: Monthly payment varies with actual solar energy generated. Monsoon months (less sun) = lower payment. Summer months (more sun) = slightly higher payment.
Value: Feels fair—pay less when you get less. Reduces complaints during monsoon. Aligns cost with value received.
Dependency #2: Maintenance alert priority ↔ season
Value: Maintenance guidance that matches actual conditions. Prevents seasonal damage. Feels locally relevant.
Dependency #3: Battery charge reserved for emergency ↔ local grid reliability
Type: Internal-External
Virtual Product: In areas with frequent grid outages, battery automatically reserves 20% charge as emergency backup. In reliable-grid areas, it uses full capacity.
Value: Adaptive resilience. Households in unreliable-grid villages always have emergency light. System learns local conditions.
Unbroken (Sweden Disaster Logistics)
Dependency #1: Data sync frequency ↔ connectivity quality
Type: Internal-External
Virtual Product: Full real-time sync on good connections. Compressed periodic batch sync on poor connections. Store-and-forward mode when offline.
Value: System never fails—just degrades gracefully. Volunteers don't lose data when connectivity drops. Automatically recovers when connection improves.
Dependency #2: Resource allocation priority weights ↔ time since disaster
Value: No manual reprioritization needed. System evolves with disaster phase. Coordinators can override but defaults are smart.
Output Format
# VARIABLE DEPENDENCY Ideas: [Product/Service Name]
## Variable Inventory
**Internal Variables (controlled):**
1. [Variable] - [brief description of range/options]
2. [Variable] - [brief description]
[List 10-15]
**External Variables (environment/user):**
1. [Variable] - [brief description of range/conditions]
2. [Variable] - [brief description]
[List 8-12]
## Dependency Matrix Highlights
[Note the most interesting variable pairs from scanning the matrix]
## New Dependency Ideas
### Idea 1: [Short Name]
**Dependency:** [Internal variable] varies with [External/Internal variable]
**Type:** [Internal-Internal / Internal-External / Breaking existing]
**Virtual Product:** [Describe what the product does with this dependency]
**Value Discovered:**
- [Benefit 1]
- [Benefit 2]
**Who benefits most:** [Target user/segment]
**Feasibility:** [High/Medium/Low]
### Idea 2: [Short Name]
[Same structure]
[Continue for 8-12 ideas]
## Broken Dependency Ideas
### Idea B1: [Short Name]
**Current dependency:** [Variable A] currently depends on [Variable B]
**What if independent:** [Describe the product without this link]
**Value Discovered:** [What new capability or freedom this creates]
[Continue for 2-3 ideas]
## Summary
**Total dependency ideas:** [count]
**Most promising:** [List top 3-4]
**Most surprising:** [Which ideas emerged that no one would have requested?]
**Ready for:** Evaluation phase or continue with other SIT techniques
Tips for Success
Do:
Spend time mapping variables thoroughly—the richness of your variable list determines idea quality
Try unlikely pairs—the most innovative dependencies come from variables no one thought to connect
Consider the user's emotional and contextual variables, not just physical ones
Think about breaking dependencies too, not just creating them
Include "time" as a variable—many powerful innovations involve time-based adaptation
Don't:
Stop at obvious dependencies (price ↔ demand is known; go further)
Dismiss a dependency because "it would be hard to implement"—you're discovering value, not building yet
Forget external variables—the most powerful dependencies often link internal attributes to external conditions
Only consider linear relationships—sometimes the dependency is a threshold, a curve, or an inversion
Common Pitfalls
Only mapping obvious variables: Dig deeper. "User's mood" and "social context" are valid external variables, not just "time of day."
Creating trivial dependencies: "Price varies with quantity" already exists. Look for genuinely new relationships.
Forgetting Function Follows Form: After creating the dependency, you MUST ask "what value does this create?" The dependency alone isn't the innovation—the discovered value is.
Skipping the matrix: The matrix forces you to consider pairs you'd never think of intuitively. Don't skip it.
Only creating, never breaking: Some of the best innovations break existing dependencies (streaming broke the schedule-dependency of TV watching).
Integration Points
Input from solution-definition:
Closed World internal components map to internal variables
Closed World external components map to external variables
PRFAQ customer pains suggest which dependencies would be most valuable
Output to idea-evaluation:
Variable dependency ideas with clear value propositions
Feasibility notes (some dependencies are easy to implement, others need sensors/data)
Target user segments for each dependency
Used with other SIT techniques:
Variable Dependency + Subtraction: Remove a component, then create a dependency to compensate
Variable Dependency + Task Unification: A component with a new task whose behavior depends on context
Variable Dependency + Multiplication: Copies that differ based on a variable dependency
Next Steps
After generating variable dependency ideas:
Continue SIT: Use ideation-sit orchestrator for Subtraction, Multiplication, Division, Task Unification
Or continue SCAMPER: Add to your SCAMPER ideas for a larger portfolio
Evaluate: Use idea-evaluation skill to rate dependency ideas
Validate: Test assumptions with critical-validation skill
Variable Dependency is often the most distinctive SIT technique—it produces "smart" products that feel like magic to users.