/client-discovery — Customer & Market Discovery

Run structured customer and market discovery using Jobs to Be Done, Opportunity Solution Trees, and PMF measurement frameworks.

When to Use

• User says "customer research", "discovery", "JTBD analysis", "understand the customer"
• Starting a new project and need to understand the target audience
• Evaluating product-market fit for an existing product
• Mapping the customer journey to find opportunities

Before Starting

Check for existing project context:

1. Read projects/ directory for any project onboarding context
2. Check docs/strategy/ for prior strategy or discovery work
3. If a project name is mentioned, read projects/<project>/onboarding.md for context gathered by /onboarding

Process

Step 1: Intake — What Are We Discovering?

If project context exists from /onboarding, use it as the starting point. Otherwise, gather essentials:

AskUserQuestion:
  question: "What product or business are we researching? And what's the key question you need answered?"
  header: "Discovery"
  options:
    - label: "Who should we build for?"
      description: "Identify and prioritize target segments"
    - label: "Why do people buy/switch?"
      description: "Understand the forces driving adoption and churn"
    - label: "Where's the opportunity?"
      description: "Map the customer journey to find gaps and friction"
    - label: "Do we have PMF?"
      description: "Measure product-market fit and identify the ideal segment"

Then gather:

• The product/business — What does it do? Who is it for today?
• The hypothesis — What do you believe about the customer that you want to validate or challenge?
• Existing data — Any customer interviews, NPS scores, churn data, or survey results?
• Competitive context — What alternatives do customers use today?

Summarize back and confirm before proceeding.

Step 2: Research — Parallel Market Intelligence

Launch 2 agents IN PARALLEL:

Agent 1 — Market Landscape

Task(subagent_type: "general-purpose", description: "Research market landscape")
prompt: Research the market landscape for [PRODUCT/BUSINESS].
  - Who are the main competitors and alternatives (including "do nothing")?
  - How do competitors position themselves? What do they promise?
  - What are customers saying in reviews, forums, social media? (Look for struggling moments and unmet needs)
  - What are the key trends and shifts in this space?
  Return structured findings with sources.

Agent 2 — Customer Behavior Patterns

Task(subagent_type: "general-purpose", description: "Research customer behavior")
prompt: Research customer behavior patterns for [TARGET AUDIENCE] in [SPACE].
  - What jobs are these customers hiring products to do?
  - What are common struggling moments that trigger searching for solutions?
  - What alternatives do people use today? What do they love/hate about each?
  - What does the buying journey look like? (First thought → active looking → deciding → first use)
  Return structured findings organized by JTBD framework.

Step 3: Synthesize — Build the Discovery Map

Using research results + intake context, build:

1. Jobs to Be Done Analysis
For each job identified:

• The job statement: "When [situation], I want to [motivation], so I can [outcome]"
• Push forces (what's driving change)
• Pull forces (what's attracting them to solutions)
• Anxiety forces (what's holding them back)
• Habit forces (what keeps them in the status quo)

2. Opportunity Map
Using Teresa Torres' Opportunity Solution Tree format:

• Business outcome we're targeting
• Customer opportunities (unmet needs / pain points / desires) — ranked by frequency and severity
• Potential solutions per opportunity (at least 2-3 per opportunity)
• Riskiest assumptions to test

3. Customer Segments

• Who are the "very disappointed" users (or would-be users)? — The Rahul Vohra PMF segment
• What do they have in common? (Demographics, behaviors, context)
• What language do they use to describe their problem and the value they get?

4. Competitive Landscape

• How customers currently solve this job (including "do nothing")
• Strengths and weaknesses of each alternative
• White space — where no current solution serves well

Step 4: Validate — Check with the User

Present the discovery map and check:

AskUserQuestion:
  question: "Does this match your understanding? What surprises you or feels off?"
  header: "Validation"
  options:
    - label: "Spot on"
      description: "This matches what I've seen — let's move forward"
    - label: "Partially right"
      description: "Some things resonate but others feel off — let me clarify"
    - label: "Missing context"
      description: "I have customer data or insights that should inform this"

Iterate based on feedback.

Step 5: Recommendations

Based on the discovery, provide:

1. Target segment recommendation — Who to focus on and why (the PMF segment)
2. Top 3 opportunities — Ranked by impact and feasibility
3. Key assumptions to test — What needs to be true, and how to test it cheaply
4. Interview guide — 5-7 questions to ask in customer interviews, anchored in past behavior (not opinions)
5. Competitive positioning implications — How the discovery should inform positioning

Methodology

This skill synthesizes frameworks from leading practitioners. See references/discovery-frameworks.md for detailed methodology.

Key sources: Teresa Torres (Opportunity Solution Trees), Bob Moesta (JTBD / 4 Forces), Rahul Vohra (PMF Engine), Gia Laudi (Customer-Led Growth).

Output

Save discovery brief to: projects/<project>/discovery.md

If no project context exists, present findings inline.

Next Steps

• Have a target segment? → /positioning to define how you show up
• Need pricing? → /pricing-packaging to design monetization
• Ready for GTM? → /gtm-strategist to plan your go-to-market
• Need to formalize? → /product-prd to write the requirements doc