Agent skill

thesis-market-mapping

Build and maintain a thesis-driven market map and roadmap: segment taxonomy, company list, wedge hypotheses, and 'must be true' claims. Use when exploring a new sector, refreshing coverage, or generating high-conviction sourcing targets.

Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/data/thesis-market-mapping

Metadata

Additional technical details for this skill

author
evalops
version
0.2

SKILL.md

Thesis market mapping

When to use

Use this skill when you need to:

  • Build an internal point of view on a sector (new thesis or coverage refresh)
  • Produce a market map (companies + segmentation) that drives sourcing and diligence
  • Turn "interesting" into "actionable" (who to meet, what to ignore, what to bet)

Inputs you should request (only if missing)

  • Sector / theme and a tight wedge to start with (e.g., "AI eval tooling for regulated industries")
  • Stage focus (pre-seed/seed/A/B+) and check size range
  • Any existing notes, prior memos, or firm theses
  • Time horizon (what "now" means: 6, 12, 24 months)

Outputs you must produce

  1. Market map database (CSV/Sheet-style, not prose)
  2. Thesis one-pager (max ~1 page) with falsifiable claims
  3. Top 10 force-ranked list with:
    • One-sentence "must be true" per company
    • Wedge hypothesis
    • Why now
    • What would change your mind
  4. Kill criteria (explicit signals that would cause you to abandon the thesis)
  5. Contrarian hypotheses:
    • 3 reasons this thesis is right when others disagree
    • 3 reasons this market might be fake
  6. Sourcing routing (who internally should see what, and why)

Recommended templates:

  • assets/market-map.csv
  • assets/thesis-one-pager.md
  • assets/top-targets.md

Procedure

1) Set the aperture (do this first)

Write a 3-5 line problem statement:

  • Who is the buyer?
  • What job are they trying to get done?
  • What is broken about the current workflow?
  • What changed that makes a new approach possible now?

Then define the wedge:

  • Narrow enough that you can know it better than the internet in 30-60 days.
  • Broad enough that it can expand to a big market if you're right.

2) Create a segmentation taxonomy (avoid "misc")

Create categories that reflect how buyers buy, not just product features. Use 5-9 primary segments max. For each segment, add:

  • ICP / buyer role
  • Budget source (security, IT, product, compliance, etc.)
  • Adoption trigger event (regulatory change, platform shift, cost curve, breach, etc.)
  • "Why now" drivers

3) Populate the market map (treat it like a database)

Fill the table with:

  • Company / product
  • Wedge + ICP
  • Pricing / GTM motion (PLG, sales-led, channel, services-assisted, etc.)
  • Proof of pull (referenceable customers, OSS adoption, hiring, usage signals)
  • Differentiation (one sentence)
  • Switching costs / moat hypothesis
  • Notable risks (one sentence)
  • Investors / financing (if known)
  • Status (meet / watch / pass)

Rules:

  • Keep entries short; you're optimizing for scan speed.
  • Record unknowns explicitly; don't guess.

4) Force-rank and write "must be true" claims

Pick the top 10 companies that matter and write:

  • "This becomes huge if ______ is true."
  • Evidence you have today (links, conversations, data)
  • What would falsify it (what you'd need to see to stop believing)
  • Fastest test to validate or kill (what's the 1-day test?)

Then write 3-7 "must be true" claims for the thesis overall, such as:

  • Buyer behavior ("CISO will fund this from X budget because...")
  • Distribution ("This category will be won by teams with...")
  • Product leverage ("Model improvements reduce cost by...")

5) Write contrarian hypotheses and kill criteria

Contrarian hypotheses (why you might be right when others disagree):

  • 3 reasons this thesis is correct that aren't consensus
  • What do you believe that most investors in the space would disagree with?

Why this market might be fake (steelman the bear case):

  • 3 reasons this market doesn't exist or won't work
  • What would make you abandon this thesis entirely?

Kill criteria (explicit exit conditions):

  • List 3-5 signals that would cause you to stop pursuing this thesis
  • Examples: "If [buyer type] doesn't fund this from [budget], abandon." "If cycle time exceeds X months, abandon."

6) Identify archetypes and failure modes

For each segment, list the 2-4 most common failure modes:

  • "Nice-to-have" that dies in procurement
  • High integration cost with unclear ROI
  • Incumbent bundles it "good enough"
  • Founder misunderstands buyer / cycle time

This becomes your future screening rubric.

7) Convert the map into a sourcing plan

Create a weekly plan:

  • 10 "new names" to investigate
  • 5 outreach targets
  • 3 operator calls (buyer-side truth)
  • 1 deep dive (new sub-segment)

Deliverable: a 10-bullet "what changed this week" update.

Salesforce logging (optional but recommended)

If Salesforce is the system of record, log the map so the firm can route and measure work.

Option A (simple):

  • Create a Salesforce Campaign called Thesis: <topic> (YYYY-MM) and add target companies/people as Campaign Members.

Option B (low-dependency):

  • Create/update Leads for founder contacts and Accounts for companies.
  • Store Thesis=<topic> in a tagging field if you have one; otherwise prepend it to Description.

If you need API workflows, use the salesforce-crm-ops skill.

Examples

  • Input: "Map the AI code review / AI SDLC tooling market for seed deals."
  • Output: market-map.csv with 120 companies, taxonomy, top 10 force-ranked targets with must-be-true + fastest test, 5 falsifiable claims, 3 contrarian hypotheses, 3 bear cases, kill criteria, weekly sourcing plan.

Edge cases

  • If the wedge is too broad: propose 2 narrower wedges and proceed with the one that has the clearest buyer + budget.
  • If you can't find buyers: switch to operator calls (the map is empty without buyer truth).

Expand your agent's capabilities with these related and highly-rated skills.

Didn't find tool you were looking for?

Be as detailed as possible for better results