---
name: product-discovery
description: Expert product discovery guidance for user research and problem validation. Use when conducting user interviews, validating problems, applying jobs-to-be-done framework, sizing opportunities, customer segmentation, competitive analysis, prototype testing, usability testing, designing surveys, or synthesizing research insights. Covers discovery sprints, continuous discovery, and research operations.
---

# Product Discovery

Strategic user research and problem validation expertise — from interview techniques and JTBD to opportunity sizing and insight synthesis.

## Philosophy

Great products start with great problems. Discovery is how you find problems worth solving for people who will pay.

The best product discovery:
1. **Talk to users, not stakeholders** — Customers know their problems, not solutions
2. **Validate problems before solutions** — Build the right thing, then build it right
3. **Quantify and qualify** — Numbers tell you what, conversations tell you why
4. **Continuous over batched** — Weekly habits beat quarterly projects

## How This Skill Works

When invoked, apply the guidelines in `rules/` organized by:

- `research-*` — User interview techniques, survey design, research ops
- `discovery-*` — Problem discovery, JTBD framework, validation
- `analysis-*` — Synthesis, segmentation, competitive analysis
- `testing-*` — Prototype testing, usability testing

## Core Frameworks

### Discovery Process

| Phase | Activities | Outputs |
|-------|------------|---------|
| **Explore** | Interviews, observation, data mining | Problem space map |
| **Validate** | Problem interviews, surveys, experiments | Validated problems |
| **Prioritize** | Opportunity scoring, segmentation | Prioritized roadmap |
| **Test** | Prototype testing, usability studies | Solution validation |

### Jobs-to-be-Done Framework

```
                    ┌─────────────────────┐
                    │    FUNCTIONAL JOB   │
                    │   (What they do)    │
                    └──────────┬──────────┘
                               │
              ┌────────────────┼────────────────┐
              │                │                │
              ▼                ▼                ▼
       ┌──────────┐     ┌──────────┐     ┌──────────┐
       │ EMOTIONAL│     │  SOCIAL  │     │ CONTEXT  │
       │   JOB    │     │   JOB    │     │ (When/   │
       │ (Feel)   │     │ (Appear) │     │  Where)  │
       └──────────┘     └──────────┘     └──────────┘
```

### Opportunity Scoring (OST)

| Factor | Weight | Description |
|--------|--------|-------------|
| **Importance** | 40% | How important is this job to the customer? |
| **Satisfaction** | 30% | How satisfied are they with current solutions? |
| **Frequency** | 20% | How often do they encounter this problem? |
| **Willingness to Pay** | 10% | Will they pay to solve this? |

**Opportunity Score = Importance + max(Importance - Satisfaction, 0)**

### Research Method Selection

```
┌─────────────────────────────────────────────────────────────┐
│                   GENERATIVE RESEARCH                       │
│              (Discover unknown unknowns)                    │
│  ┌───────────┐  ┌───────────┐  ┌───────────┐               │
│  │Contextual │  │ Discovery │  │ Diary    │               │
│  │ Inquiry   │  │ Interviews│  │ Studies  │               │
│  └───────────┘  └───────────┘  └───────────┘               │
├─────────────────────────────────────────────────────────────┤
│                   EVALUATIVE RESEARCH                       │
│              (Validate known hypotheses)                    │
│  ┌───────────┐  ┌───────────┐  ┌───────────┐               │
│  │ Usability │  │  A/B      │  │ Prototype │               │
│  │ Testing   │  │  Testing  │  │ Testing   │               │
│  └───────────┘  └───────────┘  └───────────┘               │
├─────────────────────────────────────────────────────────────┤
│                   QUANTITATIVE RESEARCH                     │
│              (Measure and prioritize)                       │
│  ┌───────────┐  ┌───────────┐  ┌───────────┐               │
│  │  Surveys  │  │ Analytics │  │ Card      │               │
│  │           │  │  Review   │  │ Sorting   │               │
│  └───────────┘  └───────────┘  └───────────┘               │
└─────────────────────────────────────────────────────────────┘
```

### Customer Segmentation Matrix

| Dimension | Consumer (B2C) | Business (B2B) |
|-----------|----------------|----------------|
| **Demographics** | Age, income, location | Company size, industry, revenue |
| **Behavior** | Usage patterns, purchase history | Buying process, tech stack |
| **Psychographics** | Values, lifestyle, attitudes | Company culture, risk tolerance |
| **Needs** | Problems, goals, aspirations | Business outcomes, KPIs |

### Continuous Discovery Cadence

```
Weekly:
├── 2-3 customer interviews
├── Review analytics/feedback
└── Update opportunity backlog

Monthly:
├── Synthesis session
├── Prioritization review
└── Stakeholder alignment

Quarterly:
├── Deep-dive research sprint
├── Competitive analysis refresh
└── Segment review
```

## Interview Quick Reference

| Interview Type | When to Use | Key Questions |
|---------------|-------------|---------------|
| **Discovery** | Exploring problem space | "Tell me about the last time..." |
| **Problem** | Validating specific pain | "How painful is this 1-10? Why?" |
| **Solution** | Testing concepts | "Would this solve your problem?" |
| **JTBD** | Understanding motivation | "What were you trying to accomplish?" |
| **Usability** | Testing interfaces | "What do you expect to happen?" |

## Anti-Patterns

- **Solution-first discovery** — Falling in love with solutions before validating problems
- **Leading the witness** — Asking questions that suggest desired answers
- **Confirmation bias** — Only hearing what supports your hypothesis
- **Sample of one** — Making decisions from a single interview
- **Proxy research** — Asking salespeople instead of customers
- **Feature requests as research** — Users ask for features, not problems
- **Analysis paralysis** — Researching forever, never deciding
- **HiPPO-driven** — Highest Paid Person's Opinion overriding data