/research
Conduct structured market research with web search and produce an evidence-based landscape document that downstream planning commands can build on.
When to use it
Run this first, before /bootstrap-product, when you are not yet
confident you know the market. The command runs you through a
questionnaire — what the product is, what industry, who the
competitors might be, what audience signals you’ve seen — then uses
web search to flesh out a competitive landscape, audience
characteristics, and pricing patterns. The output lands in
docs/MARKET-RESEARCH.md and becomes the substrate the other Plan
phase commands draw from.
If you already have a clear competitive map from prior work, you can
skip straight to /jtbd or
/positioning. Those commands check for the
research file’s presence at the start and pre-fill from it when it
exists; without it, they still work but ask you to supply the same
information by hand. The research command is the leverage point —
one focused session here saves the same questions being asked three
more times in later commands.
Don’t use this for narrow technical lookups (“what does library X do”) — that’s day-to-day web search work, not a market research session. Use this when the question is “who else is in this space, who are they serving, and where are the gaps.”
How it fits
The research document becomes upstream context for the rest of the
Plan phase. Three later commands (/jtbd, /positioning,
/bootstrap-product) read it at start and pre-fill the relevant
sections.
What it does (quick)
- Asks a structured questionnaire (product, industry, known competitors, target audience signals)
- Runs web searches to expand the competitive landscape and surface pricing patterns + audience characteristics
- Synthesizes findings into
docs/MARKET-RESEARCH.mdwith sections for competitor analysis, target audience insights, pricing patterns, and underserved segments - Surfaces gaps it could not answer from web search alone — those become items for follow-up
Related commands
/jtbd— natural next step; reads the research file and pre-fills competitor pain points/positioning— uses competitor analysis as the “competitive alternatives” input/bootstrap-product— the PRD draft pulls audience + competitive sections directly
Canonical spec: .claude/commands/research.md