The Role of Prototyping in PM Product Discovery: Methods and Implementation

Prototyping serves as a cornerstone activity in the pm-product-discovery toolkit, enabling teams to test assumptions quickly, gather concrete user-behavior data, and reduce risk before committing to full development.

In the phuryn/pm-skills repository, prototyping is embedded throughout the discovery workflow as a low-effort validation method. This open-source framework treats prototypes not as optional design assets, but as mandatory experiment types that accelerate learning while keeping resource investment minimal.

Why Prototyping Matters in Product Discovery

Product discovery fails when teams commit to solutions based on opinions rather than evidence. Prototyping bridges this gap by creating tangible artifacts that generate real interaction data without requiring full engineering effort.

According to the pm-product-discovery source code, prototypes allow teams to validate hypotheses through methods like first-click testing and task completion studies. These techniques reveal actual user behavior patterns that surveys and interviews cannot capture, ensuring that product decisions rest on observed actions rather than stated intentions.

How Prototyping is Embedded in the PM-Skills Framework

The framework integrates prototyping into multiple command workflows, ensuring it appears at critical decision points.

Low-Effort Validation in Brainstorm-Experiments-Existing

In pm-product-discovery/skills/brainstorm-experiments-existing/SKILL.md, the system explicitly calls for "First-click testing or task completion with a prototype" as a primary experiment option. This skill treats prototypes as equal validation methods alongside A/B tests and wizard-of-oz experiments, emphasizing their role in rapid hypothesis testing.

Discovery Workflow Integration

The discover command, documented in pm-product-discovery/commands/discover.md, lists prototypes alongside A/B tests, fake-door tests, and spikes as lightweight validation methods for existing products. This placement ensures that teams consider interactive mockups before committing to production code.

Rapid Feasibility Triage

When evaluating incoming requests, the triage-requests command (in pm-product-discovery/commands/triage-requests.md) recommends a "prototype" action specifically for requests needing rapid feasibility checking. This triggers immediate validation rather than lengthy specification documents.

Ideation Scaffolding

The brainstorm command, defined in pm-product-discovery/commands/brainstorm.md, surfaces "prototypes" as a primary experiment type during ideation sessions. This ensures that clickable mockups remain part of the conversation from the earliest planning stages.

Practical Implementation: Running Prototype Experiments

To execute prototype-based discovery, use the /brainstorm command to generate specific experiment plans:

/brainstorm experiments existing "Revamp checkout flow for mobile app"

The system executes this workflow:

1. Context gathering: Prompts for product details, user segments, and known constraints.
2. Experiment generation: Returns a structured table of validation options.

The output typically includes prototype experiments such as:

| # | Experiment | Tests Hypothesis | Method | Effort | Timeline |

|---|------------|------------------|--------|--------|----------| | 1 | Clickable prototype | 30% of users will complete checkout faster | Interactive mock-up in Figma/Proto.io | Low | 1 week | | 2 | A/B test of CTA copy | 15% increase in conversion | Deploy on staging, split traffic | Low | 2 weeks | | 3 | Wizard-of-Oz test | Validate perceived value of new payment option | Manual backend processing | Medium | 2 weeks |

You can then request detailed execution plans for any specific experiment (e.g., "detail experiment 1") to receive step-by-step guidance on building and testing the prototype.

Summary

• Prototyping is a mandatory validation method in pm-product-discovery, not an optional design phase.
• The brainstorm-experiments-existing skill explicitly mandates prototype testing for first-click and task-completion validation.
• Three command files (discover.md, triage-requests.md, brainstorm.md) embed prototyping into standard workflows.
• Low effort, high learning: Prototypes generate concrete user-behavior data with minimal engineering investment.
• Real interaction data: Unlike surveys, prototypes reveal actual user navigation patterns and decision points.

Frequently Asked Questions

When should I use a prototype versus an A/B test in product discovery?

Use a prototype when you need to validate navigation flows, information architecture, or user comprehension before writing production code—typically during the brainstorm-experiments-existing phase. Use an A/B test when you have a functional system and need to optimize specific elements like copy or button placement. The pm-product-discovery framework explicitly positions prototypes as the appropriate tool for early-stage validation, while A/B tests belong in later optimization cycles.

What specific prototype methods does the pm-skills framework recommend?

According to pm-product-discovery/skills/brainstorm-experiments-existing/SKILL.md, the framework recommends first-click testing and task completion studies using interactive mockups. These methods focus on behavioral metrics rather than aesthetic preferences, ensuring that prototypes answer functional questions about user flow rather than visual design opinions.

How does the triage-requests command utilize prototyping?

The triage-requests command, located in pm-product-discovery/commands/triage-requests.md, automatically recommends a "prototype" action when incoming requests require rapid feasibility checking. This triggers an immediate validation workflow rather than lengthy requirements gathering, ensuring that teams build only what they can validate through user interaction.

Can I run prototype experiments on existing products, or only new features?

The framework specifically supports prototyping for existing products through the brainstorm-experiments-existing skill. The discover command lists prototypes as a discovery technique for existing products alongside other lightweight methods like fake-door tests and spikes, making them applicable for both iterative improvements and major overhauls.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}