asb-interview-hypotheses

Elicits and records falsifiable customer hypotheses mapped to interview goals in HYPOTHESES.md.

Updated Nov 15, 2025
One-click install
npx skills add https://github.com/evanish/reusable-PM-templates --skill asb-interview-hypotheses-evanish
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: asb-interview-hypotheses
Source: https://github.com/evanish/reusable-PM-templates/tree/main/Templates%20to%20reuse/Work/asb-customer-discovery/asb-interview-hypotheses
Command: npx skills add https://github.com/evanish/reusable-PM-templates --skill asb-interview-hypotheses-evanish

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Before customer interviews, founders and PMs hold unwritten assumptions that confirmation bias silently protects. This Skill draws out those beliefs, sharpens them into numbered, falsifiable hypotheses (H1, H2, ...) mapped to goal questions, and preserves them in a HYPOTHESES.md file so interviews can confirm or contradict them. ## Core Features & Use Cases - Goal-driven elicitation: Reads a GOALS.md goal-question list and walks it one goal at a time, drawing out the user's own beliefs rather than generating answers for them. - Falsifiability rubric: Critiques draft hypotheses for specificity, testability, single claims, goal coverage, and whether resolving each one would change a real decision. - Resumable artifact: Writes HYPOTHESES.md incrementally with an in-progress status note, so sessions can resume exactly where they stopped. - Use Case: A founder with a list of eleven goal questions about a hosting market uses this Skill to record eighteen numbered hypotheses with thresholds like "$50/mo" and "100,000 page-views," each tagged to the goals it addresses, before writing any interview questions. ## Quick Start Run the asb-interview-hypotheses skill on my GOALS.md to help me write down my customer hypotheses before interviewing.

Frequently Asked Questions about asb-interview-hypotheses

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I write customer interview hypotheses before user research?

Start from a numbered list of goal questions, then state your current best guess for each as a specific, falsifiable claim with numbers or thresholds. Record them in a HYPOTHESES.md file with goal mappings so each interview question traces back to a decision.

What makes a good falsifiable customer hypothesis?

A good hypothesis contains a quantity, threshold, segment, or named behavior that evidence could contradict, such as "bloggers with 100,000 page-views will pay $50/mo for uptime." Vague claims like "customers care about speed" cannot lose an argument with reality and fail the test.

Can the AI generate the hypothesis list for me?

The Skill declines to author the list outright because unowned hypotheses teach nothing when reality contradicts them. It offers candidate hypotheses as templates only when you are stuck, and requires you to pick, correct, or reject each one individually.

What input does the interview hypotheses skill need?

It expects a GOALS.md file containing numbered goal questions, or a pasted equivalent. If no goal list exists, it helps draft a minimal one in chat first, since hypotheses without goals produce interviews without direction.

When should I not use this hypothesis-writing step?

Skip it when interviews have already been conducted and you need post-interview analysis, or when you want the interview questions themselves written. Those are separate steps; this Skill only records beliefs before interviewing.