Design Thinking + Lean Startup

Input: $ARGUMENTS

Overview

Two complementary methodologies for innovation:

Design Thinking: Discover user needs through empathy, generate creative solutions, and prototype iteratively. Good for finding the right problem.

Lean Startup: Test business assumptions through minimum viable products and validated learning. Good for testing whether solutions work.

Combined: Design Thinking discovers what to build, Lean Startup validates whether you should build it.

Steps

Step 1: Determine Phase

Where are you in the innovation process?

If you…	Use	Phase
Don’t know the problem	Design Thinking	Empathize + Define
Know the problem, need solutions	Design Thinking	Ideate + Prototype
Have a solution, need validation	Lean Startup	Build-Measure-Learn
Have validation, need scale	Neither — this is execution	→ INVOKE: /action

Step 2: Design Thinking — Empathize

Understand the user’s actual experience:

Observe: Watch users in their natural context. What do they DO (not just say)?
Interview: Ask open-ended questions. “Tell me about the last time you…”
Experience: Try being the user yourself. Feel the friction.
Map the journey: What steps do they go through? Where is the pain?

EMPATHY MAP:
Says: [what the user tells you]
Does: [what you observe them doing]
Thinks: [what they might be thinking — inferred]
Feels: [emotional state — frustrated, anxious, bored]

Key pain points:
1. [pain] — severity: [H/M/L]
2. [pain]

Unmet needs:
1. [need they expressed]
2. [need they didn't express but you observed]

Step 3: Design Thinking — Define

Turn empathy into a problem statement:

Point of View (POV) format: “[User] needs to [need] because [insight].”

Quality checks:

Is the need real (observed, not assumed)?
Is it specific enough to solve?
Does it frame the problem WITHOUT implying a solution?
Is it big enough to matter?

Step 4: Design Thinking — Ideate

Generate many possible solutions:

Diverge: Generate as many ideas as possible (quantity over quality)
- No judging during divergence
- Build on others’ ideas
- Encourage wild ideas
- Go for volume (aim for 20+ ideas)
Converge: Select the most promising
- Which ideas address the core need?
- Which are feasible?
- Which are novel?
- Select 3-5 to prototype

Step 5: Design Thinking — Prototype

Build the simplest version that can be tested:

Prototype Type	Speed	Fidelity	Best For
Paper sketch	Hours	Low	Concept testing
Clickable mockup	Days	Medium	Flow testing
Wizard of Oz	Days	Medium	Service testing (human behind the curtain)
Concierge	Days	High	High-touch service testing
Landing page	Days	Medium	Demand testing
Functional MVP	Weeks	High	Full solution testing

Rule: Use the LOWEST fidelity that can answer your question.

Step 6: Lean Startup — Build-Measure-Learn

Now validate with real users:

Identify the riskiest assumption:

What must be TRUE for this solution to work?
Which assumption, if wrong, kills the idea?
Test THAT assumption first

Build: Create minimum viable test

What’s the smallest thing you can build to test the assumption?
It should be embarrassingly simple

Measure: Define success criteria BEFORE testing

What metric will tell you if the assumption is true?
What number is “good enough”?
How many data points do you need?

Learn: Analyze results honestly

Did the assumption hold?
YES → test next riskiest assumption
NO → pivot (change approach) or persevere (more data needed)

Step 7: Pivot or Persevere

After each Build-Measure-Learn cycle:

Signal	Action
Metrics meet success criteria	Persevere — test next assumption
Metrics close to criteria	Iterate — refine, don’t restart
Metrics far from criteria	Pivot — change approach
No one uses the test	Pivot — problem may not be real

Pivot types:

Customer pivot: Same product, different audience
Problem pivot: Same audience, different problem
Solution pivot: Same problem, different solution
Channel pivot: Same product, different distribution
Revenue model pivot: Same product, different monetization

Step 8: Report

DESIGN THINKING + LEAN:
Phase: [empathize / define / ideate / prototype / validate]

Design Thinking findings:
User: [who]
Need: [what]
Insight: [why]
Solution concept: [what we're testing]

Lean Startup validation:
Riskiest assumption: [what must be true]
Test: [what we built/did]
Metric: [what we measured]
Result: [what happened]
Decision: [persevere / iterate / pivot]

Next step: [what to do next based on learnings]

When to Use

Developing new products or services
Solving ambiguous problems
When unsure what users actually need
Validating business assumptions
Reducing waste by learning before building
→ INVOKE: /cd (customer development) for deeper customer insight
→ INVOKE: /prd (product design) for product specification

Verification

User needs discovered through observation (not just assumption)
Problem defined without implying solution
Multiple solutions ideated before selecting one
Prototype is minimum fidelity needed
Riskiest assumption identified and tested first
Success criteria defined before testing
Pivot/persevere decision based on data

dtl