Market Mapping

Builds the picture of a sector before you commit a sourcing run to it. An hour of mapping saves a week of discovering the wrong 200 companies, and it tells you early whether three PE platforms already got there first.

Read first:

• references/sourcing-directories.md — the same sources, read for structure
• references/target-taxonomy.md — ownership types and what each means
• references/scoring-rules.md — the anti-fabrication rules

Step 0 — Load the thesis profile

Read thesis-profile.md from the user's workspace.

If it does not exist: stop. Copy templates/thesis-profile.md into the workspace and walk the user through sections 2, 3, and 4 at minimum. A market map with no thesis behind it is a school report, not a sourcing tool.

Confirm which sector from section 4 is being mapped. If the profile lists several, ask which one, or map them one at a time. Never blend two sectors into one map.

Step 1 — Confirm the scope

Restate back to the user:

• The sector and subsector being mapped
• Geography, or No constraint
• The size box the map should be read against
• What decision this map is meant to inform (usually: which segments target-discovery should prioritize, and whether to enter at all)

Wait for confirmation before researching.

Step 2 — Research the structure

Work through these six questions in order. For each, capture quotes and source URLs, not impressions.

1. Segments. How does the sector actually divide — by service line, customer type, geography, license class? Name each segment and what a typical company in it does.
2. Value chain. Who sits upstream and downstream. Where the margin sits, if stated anywhere quotable.
3. Fragmentation. How many companies operate in each segment, from association member counts, license registries, or census data. Label every count with its source. If you estimate, label it an estimate and state the basis. Never present a modeled count as a counted one.
4. Typical owner profile. Per segment: founder-owned, family-owned, franchisee-heavy, already consolidating? Use target-taxonomy.md language.
5. Buyer competition. Which PE platforms, holdcos, or strategics are already consolidating each segment. Evidence is their own portfolio pages and announced add-ons, quoted with URLs. A platform's existence is a fact about the market, not a reason to quit — but the user decides that.
6. Where the thesis fits. Which segments match the size box and deal type in the profile, and which are already picked over.

Rules while researching

1. Not found is correct and expected. A segment with no fragmentation data gets Not found, not a confident round number.
2. Every claim carries a source URL. Market-shaped folklore is the failure mode of this skill. Quote the association page, the registry count, the portfolio page.
3. Recency matters. A 2019 industry report may describe a market that no longer exists. Record the date of every piece of evidence.
4. Do not infer market size from vendor marketing. "The $80B home services market" from a software vendor's landing page is a sales line, not a datum. If used at all, attribute it as such.

Step 3 — Draft the map

Assemble a single markdown document with these sections:

# Market Map — [sector], [geography], [date]

## Segments            (table: segment, what they do, est. company count
                        + count basis, typical owner profile)
## Value chain         (who feeds whom; where margin sits, if quotable)
## Fragmentation       (the evidence, quoted and sourced, per segment)
## Buyer competition   (table: platform, sponsor, segment, add-ons observed,
                        source URL, evidence date)
## Thesis fit          (which segments match the size box and deal type;
                        which are picked over; which are open)
## Recommended focus   (2 to 3 segments for target-discovery, with reasons)
## What I could not establish   (the honest gaps, named)

The last section is mandatory. A map that claims full coverage of a private market is wrong by definition.

🚧 HARD GATE — before writing

Show the user the full draft in the conversation. Ask specifically:

• Does the segmentation match how operators in this market describe it?
• Is the buyer-competition list complete as far as they know?
• Do the recommended focus segments match their read?

The user usually knows things about their market that public sources do not. Fold corrections in. Get approval. Then write the map to the workspace as market-map.md (one file per sector; suffix the sector name if mapping more than one).

Step 4 — Report honestly

Market map complete — commercial landscaping, US Southeast

Segments mapped:        5 (maintenance, design-build, arbor, irrigation, snow)
Fragmentation evidence: 3 of 5 segments have sourced counts
                        (2 marked Not found — no registry, association gated)
Buyer competition:      4 active platforms found, 11 add-ons since 2023,
                        all sourced to portfolio pages
Thesis fit:             maintenance and irrigation match the size box;
                        design-build mostly runs above it
Recommended focus:      maintenance contracts $2M–$10M revenue; irrigation
                        as a secondary angle

Weakest section:  Value chain margin data — nothing quotable found
Not established:  Company counts for arbor and snow segments

Written to: market-map.md

Next: run target-discovery against the recommended segments.

Report what did not work. Gaps stated plainly are useful; gaps papered over cost real money later.

Notes

• Map before bulk discovery, not after. Running discovery first and mapping second means the map inherits the list's blind spots.
• Heavy buyer competition is information, not a verdict. Some of the best entries are into consolidating markets, earlier or smaller than the platforms go. Present the facts; the user makes the call.
• One sector per map. Blended maps hide the differences that matter.
• Revisit the map every few months if sourcing runs long. Platforms announce add-ons constantly, and the competition section goes stale first.