Back to Course

Class 3: Customer Discovery

Table of Contents

ASU SES 498/598

Customer Discovery & Market Validation

Space Business & Entrepreneurship

Fred von Graf, PMP

Find customers before building products

Today's Mission

What You'll Learn Today

🎯

Customer Discovery Process

AI-enhanced interview techniques

📊

Market Validation Methods

Testing assumptions quickly

🔍

Problem-Solution Fit

Ensuring you solve real problems

💡

MVP Planning

Minimum viable product strategy

The #1 Startup Killer

Building Something Nobody Wants

CB Insights: Top Reasons Startups Fail

  1. No market need (42%)
  2. Ran out of cash (29%)
  3. Not the right team (23%)
  4. Got outcompeted (19%)
  5. Pricing/cost issues (18%)

Space Industry Specific

  • • Long development cycles
  • • High capital requirements
  • • Limited customer base
  • • Complex requirements

Solution: Validate before you build

Customer Discovery Framework

Steve Blank's Four Steps

1. Customer Discovery

Goal: Find problem-solution fit

  • • Identify customer segments
  • • Discover their problems
  • • Test problem hypotheses

2. Customer Validation

Goal: Find product-market fit

  • • Test MVP with customers
  • • Validate business model
  • • Verify sales process

3. Customer Creation

Goal: Scale demand

4. Company Building

Goal: Scale organization

AI-Enhanced Customer Research

Before You Talk to Anyone

Market Segmentation Prompt

Identify all potential customer segments for [space product/service].
For each segment provide:
- Organization types and examples
- Budget range and procurement process
- Current solutions they use
- Key decision makers
- Pain points specific to space operations

Persona Development Prompt

Create a detailed persona for a [role] at [organization type]
who would purchase [solution]. Include:
- Daily responsibilities
- KPIs they're measured on
- Biggest frustrations
- Technology comfort level
- Decision-making authority

The Golden Rule ⭐

"Talk about their life, not your idea"

Instead of pitching your solution and asking for opinions, you dig into their current reality, problems, and behaviors. This gives you facts instead of politeness.

🚀 Why It Works for Space Ventures

In the space industry, this is especially critical because:

  • • Long development cycles - you can't afford to build the wrong thing
  • • High capital requirements - mistakes are expensive
  • • Limited customer base - every potential customer conversation matters
  • • Complex requirements - you need to understand the real operational challenges

Example in Action:

Instead of:

"Would satellite operators buy our collision avoidance software?"

Ask:

"Tell me about the last time you had to deal with a potential collision. What was that process like? How much time did your team spend on it? What tools did you use? What was the most frustrating part?"

The first question gets you a polite "maybe." The second gets you real data about their actual pain points, current solutions, and the true cost of the problem.

Customer Interview Script

AI-Generated but Human-Refined

Prompt Template

Create a customer interview script for [target customer]
about [problem space]. Include:

Opening:
- Introduction and permission to record
- Context setting (not pitching)

Problem Discovery (5 questions):
- Current workflow
- Pain points
- Frequency and severity
- Current solutions
- Cost of problem

Closing:
- Next steps
- Referral request

Live Demo - Customer Discovery

Example: Satellite Operators

Step 1: Identify Segments (Perplexity)

"List all types of satellite operators with fleet size, typical budgets, and operational challenges"

Step 2: Find Contacts (LinkedIn + AI)

"Generate LinkedIn search parameters to find [role] at [company type]"

Step 3: Prepare Questions (Claude)

"Based on this company's website [paste], what specific operational challenges might they face?"

Step 4: Synthesize Learnings (ChatGPT)

"Analyze these 5 interview transcripts and identify common patterns"

Problem Validation Matrix

Quantifying Customer Pain

Problem Frequency Severity Current Solution Willingness to Pay
Example: Satellite collision risk Daily Critical Manual tracking $100K/year
Your Problem #1 ? ? ? ?
Your Problem #2 ? ? ? ?

Scoring:

• Frequency: Daily (5) → Never (1)

• Severity: Critical (5) → Nice-to-have (1)

• Current Solution: Nothing (5) → Perfect (1)

• WTP: High (5) → Zero (1)

Target Score: >15/20

Market Sizing with AI

TAM, SAM, SOM Calculation

TAM (Total Addressable Market)

"What is the total global market size for [solution category]
including all possible applications and geographies?"

SAM (Serviceable Addressable Market)

"Within the TAM for [solution], what portion is accessible to
a [company type] based in [location] with [constraints]?"

SOM (Serviceable Obtainable Market)

"What realistic market share could a new entrant capture in
[market] within 3-5 years given [competitive landscape]?"

MVP Definition

What to Build First

MVP Is NOT

  • • A broken product
  • • Half of your features
  • • Just a landing page
  • • A perfect product

MVP IS

  • • Fastest way to test core hypothesis
  • • Solves #1 customer problem
  • • Generates learning
  • • Path to next iteration

Space Industry MVPs

  • • SpaceX: Falcon 1 (not Starship)
  • • Planet: 2 demo satellites (not constellation)
  • • Spire: AIS tracking (not weather)

Validation Experiments

Testing Without Building

Landing Page Test

  • • Value prop clarity
  • • Email capture rate
  • • Traffic sources

Concierge MVP

  • • Manual service delivery
  • • Direct customer feedback
  • • Process refinement

Wizard of Oz

  • • Automated frontend
  • • Manual backend
  • • Scale testing

Crowdfunding

  • • Market demand
  • • Price validation
  • • Community building

Case Study - Planet Labs

From Garage to Global Imagery

Customer Discovery (2010-2011)

  • • Hypothesis: Cheap satellites could work
  • • Interviews: 50+ potential customers
  • • Finding: Frequency > Resolution

MVP Evolution

  • • v1: 2 CubeSats (Dove 1-2)
  • • v2: 4 satellites (Flock-1)
  • • v3: 28 satellites
  • • Today: 200+ satellites

Key Insight

Customers didn't need better pictures—they needed more recent pictures

Hands-On Exercise

15-Minute Discovery Sprint

Your Task:

  1. Define your target customer segment
  2. Write 5 problem discovery questions
  3. Create problem validation matrix
  4. Calculate rough TAM/SAM/SOM
  5. Design simplest validation experiment

Pair & Share:

  • • Partner interviews partner
  • • Use "Mom Test" rules
  • • Document responses
  • • Share key insights

Common Discovery Mistakes

What Not to Do

❌ Avoid:

  • • Leading questions
  • • Talking more than listening
  • • Pitching your solution
  • • Ignoring negative feedback
  • • Only talking to friends

✅ Instead:

  • • Stay curious
  • • Listen 80%, talk 20%
  • • Focus on problems
  • • Embrace rejection
  • • Talk to strangers

Remember: You're not selling, you're learning

From Discovery to Validation

Evidence-Based Decision Making

Green Light Signals

  • • Multiple customers same problem
  • • Quantifiable pain ($$ or time)
  • • Existing budget allocated
  • • Failed previous attempts
  • • Urgency to solve

Red Flag Signals

  • • "That's interesting..."
  • • No current workaround
  • • Committee decisions
  • • "Not my department"
  • • Future problem

Assignment

Customer Discovery Report (Due Thursday)

Requirements:

  1. Interview 3+ potential customers (15-20 min each)
  2. Document using Mom Test methodology
  3. Create problem validation matrix
  4. Calculate TAM/SAM/SOM with sources
  5. Design MVP validation experiment
Deliverables:
  • • Interview transcripts/notes
  • • Problem validation matrix
  • • 1-page market sizing
  • • MVP experiment design
Grading:
  • • Interview quality (40%)
  • • Problem validation (30%)
  • • Market analysis (30%)

Resources & Tools

Your Discovery Toolkit

Interview Tools

  • • Calendly (scheduling)
  • • Otter.ai (transcription)
  • • Typeform (surveys)

Analysis Tools

  • • Miro (affinity mapping)
  • • Notion (CRM)
  • • Dovetail (insights)

AI Prompts Library

github.com/[class-repo]/prompts

Reading

  • • "The Mom Test" - Rob Fitzpatrick
  • • "Talking to Humans" - Giff Constable

Next Class Preview

Thursday: Financial Modeling & Unit Economics

You'll Learn:

  • • Space business economics
  • • Unit economics modeling
  • • Financial projections with AI
  • • Investor-ready financials

Come Prepared:

  • • Customer discovery complete
  • • Basic pricing assumptions
  • • Cost structure ideas

Questions?

Key Takeaway:

"Fall in love with the problem, not the solution"

Office Hours:

Tuesday/Thursday 3:00-4:00 PM

Contact:

Fred@W4M.ai

🚀 Get out of the building and talk to customers!