# Shaper:market > Full-spectrum market research across problem space, competitive landscape, and market sizing. Use when validating a business idea against real evidence. - Skill: `rp7o/shaper-market` (Agent Skill) - Install (CLI): `npx skillmds@latest add rp7o/shaper-market` - Raw SKILL.md: https://api.skillmd.com/api/skills/rp7o/shaper-market/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Research & Search - Author: rp7o (https://skillmd.com/u/rp7o) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/rp7o/shaper-market --- # market — Full-Spectrum Market Research Runs up to three independent research modules — problem-space, competitive landscape, and market sizing — each with a user-agency choice at entry. Primary research is collected first, then the user selects which modules to run. Before running this command, read `../../references/shared-protocols.md` and `../../references/handbook-templates.md`. ## When to Run - After `shaper:canvas` to validate the Problem, Solution, and Revenue Streams blocks. - User explicitly says `shaper:market` or asks about competitors, market size, or problem validation. --- ## Workflow ### Step 1: Read the Canvas Read `.shaper/business_state.md`. If no canvas exists, ask: "I need to know what we are building first. Should we run `/shaper:kickoff` or `/shaper:canvas`?" Check the Session Log. If `shaper:market` has already run, apply the **Cascade Awareness** protocol from `shared-protocols.md`: note what changed and warn if downstream artifacts may be stale. ### Step 2: Ask for Primary Research Before doing any secondary research, ask the user if they have done primary research. Use the `AskUserQuestion` tool to ask: "Have you done any primary research — user interviews, surveys, usability tests, sales calls, or anything where you talked directly to potential customers?" with these single-select options: - **Yes, I have materials** — I can share transcripts, notes, survey results, or recordings. - **Yes, but nothing written down** — I did some informal conversations and can summarize. - **No** — We'll rely on secondary research for now. If **Yes, I have materials**: Ask the user to share them (paste text, provide file paths, or describe the key findings). Read and incorporate the primary data. If **Yes, but nothing written down**: Ask: "Give me the 3 most important things you learned from those conversations." Capture their summary. If **No**: Proceed. Note in the state that no primary research was done — this becomes a flag for `shaper:validate`. > **Primary research always takes precedence over secondary research.** When real user quotes, interview data, or survey results conflict with assumptions from web research, the primary data wins. ### Step 3: Select Research Modules Use `AskUserQuestion` (multi-select) to ask: "What market research do you want to run?" with these options: - **Problem space** — Find what real people say about this problem online: forums, Reddit, reviews, complaints. - **Competitive landscape** — Identify who already solves this, how they're priced, and where they're weak. - **Market sizing** — Estimate TAM/SAM/SOM for the opportunity. Run each selected module in sequence (A → B → C). Unselected modules are skipped; record them as `none` in state. --- ## Module A: Problem-Space Research _Runs only if "Problem space" was selected in Step 3._ Use `AskUserQuestion` (single-select) to ask: "How do you want to handle problem-space research?" with options: - **I have data on this** — I'll share what I've found: reports, posts, screenshots, or summaries. - **Research it for me** — Search for what people say about this problem online. _(Recommended if no primary research)_ - **Skip** — I'm confident in the problem hypothesis. **If "I have data on this":** Ask the user to paste or describe their findings. Structure the input into the top 3 pain points. Record with `confidence: user-provided`. Write output to `handbook/business/primary-research.md` under `## Secondary Research — Problem Space` using the template in Section 3 of `handbook-templates.md`. **If "Research it for me":** Tell the user: "I am deploying a Market Researcher to find what real people say about [Problem]. This will take a moment." Spawn a Market Researcher subagent with this **problem-focused** prompt: > "You are a Market Researcher. Search for discussions, complaints, reviews, and forum posts where people express frustration with [Problem] in the context of [Target Audience]. Sources to check: Reddit threads, forum posts, App Store/G2 reviews, Twitter/X complaints, job postings (as pain signals), and support ticket themes. > > Do NOT look for competitors or products — look for the raw pain expressed by real people. > > Return: > 1. Top 5 pain themes with representative quotes and source URLs. > 2. Frequency signals for each theme: is this a common complaint or a rare edge case? > 3. Assessment: which pain themes align with the canvas Problem block, and which challenge it? > > Do not invent data. If you cannot find meaningful signal for a theme, say so." Wait for the subagent to return. Present the top pain themes to the user. Write output to `handbook/business/primary-research.md` under `## Secondary Research — Problem Space`. **If "Skip":** Record `problem_research_secondary: none` in state. Proceed to the next module. **State update:** Set `problem_research_secondary: yes` (if run) or `none` (if skipped). --- ## Module B: Competitive Landscape _Runs only if "Competitive landscape" was selected in Step 3._ Use `AskUserQuestion` (single-select) to ask: "How do you want to handle competitive research?" with options: - **I know the landscape — let me describe it** — I'll walk you through each competitor. - **Research it for me** — Deploy a Market Researcher to find and analyze competitors. _(Recommended)_ - **Skip** — Not needed right now. **If "I know the landscape — let me describe it":** For each competitor, ask in sequence: 1. "Name and URL?" 2. "Core value proposition in one sentence?" 3. "Pricing model?" 4. "Their biggest weakness vs. your solution?" After each entry, ask: "Add another competitor, or done?" Repeat until the user is done. Assemble findings into the competitor table format from Section 2 of `handbook-templates.md`. Record with `confidence: founder-provided`. **If "Research it for me":** Tell the user: "I am deploying a Market Researcher to find existing solutions for [Problem] targeting [Audience]. This will take a moment." Spawn a Market Researcher subagent with this **competitor-focused** prompt: > "You are a Market Researcher. We are building a product to solve [Problem] for [Target Audience]. Our proposed solution is [Solution]. Find 3-5 direct or indirect competitors. For each, return: > 1. Name and URL > 2. Their core Value Proposition > 3. Their Pricing Model > 4. Their key weakness or feature gap compared to our proposed solution. > > Focus strictly on the competitive landscape. Do not invent a business model." Wait for the subagent to return. **If "Skip":** Record `competitors_analyzed: 0` in state. Proceed to the next module. **Output:** Write to `handbook/business/competitors.md` using the template in Section 2 of `handbook-templates.md`. **State update:** Set `competitors_analyzed` to the count found (or 0 if skipped). --- ## Module C: Market Sizing _Runs only if "Market sizing" was selected in Step 3._ Use `AskUserQuestion` (single-select) to ask: "How do you want to handle market sizing?" with options: - **I have existing data** — I'll share a report, deck, or estimates I've already worked out. - **Walk me through it** — Ask me questions and I'll work out the numbers with you. - **Research it for me** — Deploy a Market Researcher for a TAM/SAM/SOM estimate. - **Skip for now** — Not needed at this stage. **If "I have existing data":** Ask the user to share the data (paste figures, link a report, or summarize). Record TAM/SAM/SOM as stated by the user with `confidence: external estimate (user-provided)`. **If "Walk me through it":** Ask these four questions in sequence: 1. "Who is your primary customer — individual consumer, SMB, or enterprise?" 2. "How many of these customers exist in your target market? (Order of magnitude is fine.)" 3. "What does a customer currently spend annually to address this problem — on tools, time, services, or workarounds?" 4. "In years 1-3, what % of that market could you realistically capture given your team and distribution?" Compute: - **TAM** = (answer 2) × (answer 3) - **SAM** = TAM filtered to the specific segment and geography the user described - **SOM** = SAM × (answer 4 / 100) Present results with all raw assumptions visible so they are auditable. Label as `confidence: Founder estimate — treat as directional`. **If "Research it for me":** Tell the user: "I am deploying a Market Researcher focused on market sizing. This is a rough order-of-magnitude estimate only." Spawn a Market Researcher subagent with this **sizing-focused** prompt: > "You are a Market Researcher. Estimate the TAM (Total Addressable Market), SAM (Serviceable Addressable Market), and SOM (Serviceable Obtainable Market) for a product solving [Problem] for [Target Audience]. > > Use publicly available data — industry reports, analyst estimates, or census/demographic data. Prioritize named sources over extrapolation. > > Return: > - TAM: [estimate + source/reasoning] > - SAM: [estimate + source/reasoning] > - SOM (Year 1-3 realistic capture): [estimate + reasoning] > - Confidence: [High / Medium / Low] with a one-sentence explanation of why. > > Important: These are rough order-of-magnitude estimates only. Do not present them as precise figures. Flag any estimate with low confidence prominently." Wait for the subagent to return. Present the sizing to the user. > **Confidence caveat**: Always surface the confidence level. Low-confidence estimates should be treated as directional, not definitive. **If "Skip for now":** Record `market_sizing: none` in state. **Output:** Append to `handbook/business/competitors.md` under a `## Market Size Estimate` section. --- ## Step 4: Synthesis After all selected modules complete, synthesize across all available evidence: - **Primary research** (Step 2 — user interviews, if provided) - **Problem-space research** (Module A — if run) - **Competitive findings** (Module B — if run) - **Market sizing** (Module C — if run) Present a combined view: 1. **What the evidence confirms** — Where multiple sources agree. 2. **What the evidence challenges** — Where findings conflict with canvas assumptions. 3. **Open questions** — What remains unvalidated. If primary research was collected, lead with it. If no primary research: add note: > "This analysis is based entirely on secondary research. Real user interviews would strengthen the validation significantly." Ask: "Based on all of this, do we need to adjust any canvas blocks?" Wait for response. If yes, update `.shaper/business_state.md` and the canvas document. --- ## Output Write the competitive analysis to `handbook/business/competitors.md` (Modules B and C). Write problem-space research to `handbook/business/primary-research.md` under `## Secondary Research — Problem Space` (Module A). If primary research materials were provided in Step 2, save them to `handbook/business/primary-research.md` using the template in Section 3 of `handbook-templates.md`. ```markdown ## Market Research Complete Modules run: [list of modules that ran] Outputs saved: - `handbook/business/primary-research.md` — problem-space findings - `handbook/business/competitors.md` — competitive analysis and market sizing ### Suggested Next Step → `/shaper:validate` — Let's review the Canvas and the Market Data to make a Go/No-Go decision. → `/shaper:help` — See all available commands. ``` --- ## State Update - Update `.shaper/business_state.md` if the UVP or Solution changed. - Record primary research status: `primary_research: yes | informal | none`. - Record `problem_research_secondary: yes | none`. - Update `competitors_analyzed` count and `market_summary` in the Research section. - Record `market_sizing: yes | founder-estimate | external | none`. - **Session Log**: Append today's date, "shaper:market", "[modules run], primary research: [yes/informal/none]".