market-sizing

Produce a rigorous, sourced TAM/SAM/SOM market sizing for any product or business. Use this skill whenever a user asks about market size, total addressable m...

Security Scan

The skill's instructions, required resources, and examples are coherent with its stated purpose (TAM/SAM/SOM market sizing); it requests no credentials or installs and stays within its domain.

Purpose & Capability

Name/description match the SKILL.md: the skill explains step-by-step how to produce TAM/SAM/SOM analyses and provides example templates for B2B SaaS, B2B physical, and B2C. It requires no binaries, env vars, or installs — all of which are appropriate for an instruction-only market-sizing skill.

Instruction Scope

The SKILL.md instructs the agent to load local example files, classify the product, perform bottom-up and top-down searches, and explicitly collect and cite external data sources. This scope is appropriate for market sizing, but the agent will need web access and may encounter paywalled sources; the guidance to 'collect enough reliable data' is broad and gives the agent discretion in source selection, so users should verify cited sources and assumptions in outputs.

Install Mechanism

No install spec and no code files — lowest-risk form. Nothing is downloaded or written to disk by the skill itself.

Credentials

The skill requests no environment variables, credentials, or config paths. All required data comes from public sources or the included example files, which is proportionate to the described functionality.

Persistence & Privilege

always is false and the skill does not request special privileges or persistent system changes. Autonomous invocation is allowed (platform default) but not combined with elevated privileges or secret access.

Assessment

This skill appears internally consistent and low-risk: it only contains prose instructions and example templates for doing market sizing and asks for no credentials or installs. Before relying on outputs, check the agent's cited sources (some recommended sources are paywalled or subscription-based) and validate the skill's key assumptions (ACV anchors, penetration rates, filters) against primary data. If you will run market sizing on confidential internal products or use organization credentials, confirm the agent/tooling used to fetch data is allowed to access those resources — the skill itself does not request any secrets.

TAM / SAM / SOM Market Sizing

Produce both a bottom-up and a top-down analysis for every engagement. Both methods are required — they serve different purposes and must be reconciled. The goal is a defensible, sourced output — not a fast estimate.

---

When to use this skill

Trigger on any prompt that contains:

• "TAM", "SAM", "SOM", or "market size"
• "How big is the market for X"
• "Total addressable market", "serviceable market", "obtainable market"
• "Market sizing for [product/company]"
• Investor deck context requiring market opportunity quantification

---

Loading examples

Before calculating, check the examples/ directory for a file matching the product's category. Load the closest match to calibrate data sources, filter logic, and ACV anchors before you begin.

File	Use when product is...
examples/b2b-saas.md	Software / AI tool sold to businesses
examples/b2c-consumer-brand.md	Consumer packaged goods, DTC brands, subscriptions
examples/b2b-physical.md	Physical goods or materials sold to businesses

If no exact match exists, load the closest file and adapt.

---

Pre-flight: Classify the product before doing anything else

Before searching or calculating, classify the product along these two axes. The classification directly determines which data sources to use and which pricing method applies.

Axis 1 — Business model

Code	Type	Revenue unit	Pricing anchor
B2B-SaaS	Software sold to businesses	Annual contract (ACV)	10–20% of economic value created
B2B-Physical	Physical goods sold to businesses	Per-unit price × annual volume	Gross margin benchmarks by industry
B2C-Brand	Consumer product (any category)	Revenue per customer per year	Avg. purchase price × purchase frequency
B2C-Subscription	Consumer subscription	Monthly/annual subscription fee	Stated price or comp pricing
Marketplace/Platform	Takes % of GMV	Take rate × GMV	Industry take rate benchmarks

Axis 2 — Market geography

• US-only: Use US Census, NAICS codes, US industry associations
• Global: Use global market reports, then apply regional share (US ~25–30% of global GDP; NA ~32%; APAC ~38%)
• Single city/region: Use metro area population data + category penetration rates

Step-by-step workflow (universal)

Step 0 - Data Gathering

Since accuracy and reliability are key of this task, make sure you collect enough reliable data from reliable sources before the calculation. Limited estimation is fine with enough evidence support, but any kind of data making up or hallucination without evidence during the calculation and output phase is a fraud.

STEP 1 — Define the revenue unit

Before any market data search, lock down exactly what the product sells, to whom, and at what price.

Required inputs:

• Who is the buyer? (job title / consumer demographic)
• What is the unit of sale? (per seat, per SKU, per kg, per subscription, per transaction)
• What is the price? (stated, or estimated by comp analysis)
• What is the purchase frequency? (one-time, monthly, annual, recurring)

If price is unknown: Search "[product category] average price" OR "[closest competitor] pricing" and anchor to the median. For B2B-SaaS, also apply the 10–20% value rule: price = 10–20% of annual economic value the product creates for the customer.

Output of Step 1: A single sentence — "The revenue unit is [X] sold to [Y] at [Z]/year."

STEP 2 — TAM (Total Addressable Market)

TAM = the maximum theoretical revenue if every potential buyer purchased the product.

2A — Bottom-up method (PREFERRED for all product types)

For B2B products:

1. Search for the number of firms in the target industry
   • US: Use NAICS code lookup — search "NAICS [code] number of firms US" or site:siccode.com NAICS [code]
   • Global: Use IBISWorld, Statista, or national business registries
2. Estimate the % that are genuinely relevant (apply product-fit filter at this stage only if very obvious — e.g. residential-only firms for a commercial B2B product)
3. TAM = Total relevant firms × ACV

For B2C products:

1. Find the total consumer population in the target geography (US Census, APPA, Statista, trade associations)
2. Find the % in the target demographic/behavioral segment
3. TAM = Segment population × annual spend per person

Search queries to use:

• "[industry] number of [businesses/firms/brands] US 2024"
• "NAICS [code] firm count employees revenue"
• "[consumer category] number of [households/users/owners] US"
• "[category] market size 2024" (for cross-reference only — not the primary method)

2B — Top-down method (SECONDARY cross-check)

1. Find total category market size from analyst reports (Grand View, IMARC, Mordor, Precedence, IBISWorld)
2. Apply funnel: Total market → relevant segment → product-type share → geography
3. Search: "[category] software/product market size 2024" or "[category] industry revenue US 2024"

Rule: Bottom-up is the primary method. Top-down is the cross-check. Both results should be reported. If they diverge significantly (>2×), explain why (e.g. top-down only counts current spenders; bottom-up counts total potential).

TAM output format:

• Single dollar figure (round to nearest $100M for large markets, $10M for mid-size)
• Conservative range (low–high)
• The math shown explicitly: N firms × $X ACV = $Y or N consumers × $Z annual spend = $Y

STEP 3 — SAM (Serviceable Addressable Market)

SAM = the subset of TAM that the specific product can actually serve today, given its business model, geography, language, and product fit.

Universal filter checklist — apply all that are relevant:

Filter	Typical reduction	How to estimate
Geography	Varies	If US-only product, filter out non-US. If city-level, apply metro population fraction.
Product-fit segment	30–60% reduction	Remove segments the product doesn't serve (e.g. residential-only for commercial SaaS; solo operators who can't justify the cost)
Tech/channel readiness	10–30% reduction	% of target customers with internet access, cloud tools, or relevant distribution channel access
Language/regulatory	Varies	Only relevant for global products
Willingness to pay tier	20–40% reduction	Remove segments priced out of the product's tier

Formula: SAM = TAM × Filter₁ × Filter₂ × ... × Filterₙ

Blended ACV for SAM: If there are multiple customer segments (SME vs. mid-market, or mass vs. premium), calculate a weighted average ACV: Blended ACV = (Segment_A_ACV × weight_A) + (Segment_B_ACV × weight_B)

Search queries for filter data:

• "[industry] percentage [commercial/residential/enterprise/SMB]"
• "[industry] small business cloud software adoption rate"
• "[consumer category] premium vs mass market share"

SAM output format:

• Single dollar figure with conservative range
• Every filter listed with its % reduction and data source
• Blended ACV (if applicable)

STEP 4 — SOM (Serviceable Obtainable Market)

SOM = the portion of SAM the product can realistically capture in a defined timeframe (typically 3 years / Year 1–3 ramp).

Penetration rate benchmarks by product type:

Product type	Year 1	Year 3	Source basis
New B2B SaaS (no category)	0.1–0.3% of SAM	1.5–3% of SAM	Andreessen Horowitz, OpenView benchmarks
New B2B SaaS (established category)	0.3–0.8%	3–7%	Category incumbents' early growth data
New B2C consumer brand (retail)	0.05–0.2% of SAM	0.5–2%	Nielsen new brand launch data
New B2C DTC subscription	0.1–0.5%	1–4%	DTC cohort benchmarks
New B2B physical product	0.2–0.5%	2–5%	Trade distribution ramp benchmarks

Cross-validation method: Find 2–3 comparable companies (same category, similar stage) and anchor SOM to their reported early ARR/revenue. Search: "[similar company] [Series A / early revenue / ARR] raised [year]"

Formula:

• SOM (Year 1) = SAM × Year 1 penetration rate
• SOM (Year 3) = SAM × Year 3 penetration rate
• Customer count = SOM ÷ Blended ACV (sanity check — does this number of customers feel achievable given go-to-market?)

SOM output format:

• Year 1 and Year 3 ARR/revenue targets
• Customer count at each stage
• Named comparable company validations
• Penetration rate used (with justification)

STEP 5 — Reconcile and flag conflicts

After running both methods:

1. If bottom-up TAM > top-down TAM: Usually because the product creates new demand (not just shifting existing spend). State this explicitly.
2. If bottom-up TAM < top-down TAM: Often because the product is a niche within a large category. State which definition is being used.
3. Always present a range, not a single number, for TAM and SAM. Single numbers imply false precision.
4. Flag the 3 most sensitive assumptions — the ones where a change of ±20% would materially shift the output. These are the assumptions to validate with primary research.

Source hierarchy (use in this order, stop when satisfied)

Tier 1 — Primary/official sources (highest credibility)

• US Census Bureau / Economic Census — firm counts, employment, revenue by NAICS code
• NAICS/SIC code databases (siccode.com, census.gov) — industry firm counts
• Industry trade associations (APPA for pets, CFMA for construction, NMMA for marine, etc.)
• SEC filings / public company 10-Ks — addressable market disclosures
• Government labour/business statistics (BLS, SBA, ONS)

Tier 2 — Reputable market research

• IBISWorld, Grand View Research, Mordor Intelligence, IMARC Group, Precedence Research, Statista
• Use these for: category market size, CAGR, segment shares
• Cross-reference at least 2 sources — market research firms often disagree by 20–40%

Tier 3 — Comparable company evidence

• Funding announcements (TechCrunch, Crunchbase, Construction Dive, etc.)
• Public company revenue disclosures at IPO/Series B
• Use for: SOM cross-validation, ACV benchmarks, growth rate validation

Tier 4 — Pricing/ACV benchmarks

• Competitor pricing pages (direct)
• G2, Capterra, Trustpilot reviews mentioning price
• Stripe, Monetizely, OpenView pricing guides (for B2B SaaS)

ACV / pricing benchmarks by product type

Product type	Typical ACV/ASP	Notes
SMB B2B SaaS (1–50 employees)	$3,000–$15,000/yr	10–20% of value created
Mid-market B2B SaaS (50–500 employees)	$15,000–$60,000/yr
Enterprise B2B SaaS (500+ employees)	$60,000–$500,000+/yr
Consumer subscription (premium)	$100–$600/yr
Consumer packaged goods (premium)	$30–$150/order, 6–12×/yr
B2B physical product (SMB)	$5,000–$50,000/yr	Depends heavily on industry
B2C DTC brand (avg. LTV proxy)	2–3× first-order value

Common errors — check before finalising

Error	How to catch it
"1% of a $10B market" without proof	Always verify with bottom-up customer count
Top-down only, no bottom-up	Always run both methods
Using global TAM when product is US-only	Apply geography filter explicitly
Including non-buyers in TAM (e.g. residential firms in a commercial B2B)	Apply product-fit filter before TAM finalisation
SOM penetration rate not benchmarked	Always cite at least one comp company
ACV not grounded in value or comp pricing	Always show the value math or comp reference
Treating TAM as "market size from report" uncritically	Check the report's definition matches your product's scope
Single number without range	Always show low–high range

Industry-specific data source cheat sheet

Industry	Best firm-count source	Best market-size source	Key trade association
Construction / Trades	NAICS via siccode.com or Census	IBISWorld, CFMA Benchmarker	CFMA, AGC, NECA
Pet / Consumer packaged goods	Census NAICS 311	APPA, Grand View, IMARC	APPA, PFI
Food & Beverage	Census NAICS 311/312	Mordor, Grand View, Mintel	FMI, GMA
Beauty / Personal care	Census NAICS 325620	Euromonitor, Grand View	PBA, CEW
Fashion / Apparel	Census NAICS 315	McKinsey Fashion Report, Statista	AAFA
SaaS / Software	Crunchbase, G2 categories	Gartner, IDC, Forrester	Varies
Automotive	NAICS 441/336	Ward's Auto, IHS Markit	NADA, SEMA
B2B Packaging	Census NAICS 322/326	Grand View, Mordor	PMMI, FPA

---

Output format

The output file is where you calculations are made. After collecting necessary data, make a step-by-step analysis & calculation on the file, finally yielding a summary and a comparison. Produce a structured markdown document containing:

• Both methods in full, each covering TAM → SAM → SOM completely before moving to the next method. Each method include 3 sections (TAM, SAM, SOM), each with: headline figure + range, full calculation shown line by line, and key assumptions&sources listed
• A standalone section for all used figures' sources & assumptions.
• A summary: table with TAM / SAM / SOM figures (bottom-up and top-down side by side); a method comparison note explaining divergence if bottom-up and top-down differ by >2×

The following is an example markdown output with a structure you should follow:

"""

Market Sizing: BuildFlow AI

AI-powered invoicing automation for HVAC and Electrical subcontractors — United States

---

Method 1: Bottom-up

TAM

Input	Value	Source
HVAC firms (NAICS 238220)	88,738	siccode.com / US Census
Electrical firms (NAICS 238210)	55,951	siccode.com / US Census
% doing GC-facing commercial work	55%	SBA subcontractor mix data
ACV — HVAC SME tier	$7,500/yr	$1.5M avg. revenue × 0.5%; CFMA 2024
ACV — Electrical SME tier	$10,000/yr	$2.0M avg. revenue × 0.5%; CFMA 2024
ACV — mid-market tier (>$5M rev)	$24,000/yr	Value-based; comp: Siteline pricing

HVAC SME (48,806 × 82%):           40,021 × $7,500  = $300M
HVAC mid-market (48,806 × 18%):     8,785 × $24,000  = $211M
Electrical SME (30,773 × 82%):     25,234 × $10,000  = $252M
Electrical mid-market (30,773×18%): 5,539 × $24,000  = $133M
Value-based ceiling (5% leakage × 10–20% SaaS capture): ~$1.87B

→ TAM: $1.87B (range: $1.5B–$2.2B)

SAM

Filter	Reduction	Remaining	Source
Remove residential-only operators	−40%	47,748 firms	SBA / IBISWorld industry mix
Remove solo / owner-only operators	−35%	31,037 firms	Census nonemployer statistics
Remove cloud / tech-unready firms	−25%	23,278 firms	Construction tech adoption surveys 2024

Blended ACV:
  SME tier (70%):     $7,500  × 0.70 = $5,250
  Mid-market (30%):   $18,000 × 0.30 = $5,400
  Blended:                            = $10,650/yr

23,278 firms × $10,650 + mid-market uplift ≈ $560M

→ SAM: $560M (range: $430M–$690M)

SOM

Year	% of SAM	ARR	Customers
Year 1	0.27%	~$1.5M	~140
Year 2	0.85%	~$4.8M	~450
Year 3	1.5–3.0%	$8.4M–$16.8M	790–1,580

Benchmark: 1.5–3% of SAM by Year 3 — a.H./OpenView SaaS benchmarks for new B2B entrant in established category

Comp	Signal
Payra	Raised $15M (Edison Partners); AP automation for construction subcontractors
Siteline	Raised $18M Series A; subcontractor billing / SOV management, comparable ICP
Adaptive Construction Solutions	Bootstrapped to ~$5M ARR in 3 years; construction AR automation

→ SOM: $1.5M ARR (Year 1) / $8.4M–$16.8M ARR (Year 3), ~790–1,580 customers

---

Method 2: Top-down

TAM

Global construction accounting software (2025):  $2.64B  [Precedence Research]
× US share (76%):                                 $2.0B
× AP/AR segment (54.4%):                          $1.09B

→ TAM: $1.09B

SAM

Top-down TAM:                        $1.09B
× Specialty trade sub-share (~28%):  $305M
× HVAC + Electrical only (~50%):     $152M

→ SAM: $152M

SOM

SAM (top-down):              $152M
× 3% Year 3 penetration:     $4.6M

→ SOM: $0.4M ARR (Year 1) / $4.6M ARR (Year 3)

---

Sources

Input	Value assumed	Source
HVAC firms (NAICS 238220)	88,738	siccode.com / US Census
Electrical firms (NAICS 238210)	55,951	siccode.com / US Census
% GC-facing commercial work	55%	SBA subcontractor mix data
Avg. HVAC subcontractor revenue	$1.5M/yr	CFMA 2024 Benchmarker
Avg. Electrical subcontractor revenue	$2.0M/yr	CFMA 2024 Benchmarker
Subcontractor net profit margin	2.2–3.5%	CFMA 2024 Benchmarker
Profit lost to unbilled/errors	~5% of revenue	Industry estimate
Global construction accounting software market	$2.64B (2025)	Precedence Research
AP/AR segment share	54.4%	Precedence Research
Cloud/tech readiness of small construction firms	~75%	Construction tech adoption surveys 2024
SaaS penetration benchmark (Year 3)	1.5–3% of SAM	a.H./OpenView SaaS benchmarks
Comp: Payra raise	$15M	Edison Partners announcement
Comp: Siteline raise	$18M Series A	Public announcement

---

Summary

	Bottom-up	Top-down
TAM	$1.87B	$1.09B
SAM	$560M	$152M
SOM — Year 1 / Year 3	$1.5M / $8.4M–$16.8M ARR	$0.4M / $4.6M ARR

Bottom-up preferred — top-down counts only existing software spend and misses the large paper/spreadsheet segment. Use top-down as the floor / downside scenario. """

Quality checklist before delivering output

• Both bottom-up AND top-down methods calculated
• Every input number has a named source
• TAM and SAM shown as ranges, not single numbers
• Blended ACV used where multiple customer segments exist
• SOM penetration rate benchmarked against 1–2 comp companies
• At least 3 sensitivity risks flagged
• Part 1 markdown follows the Output Example structure exactly
• Prose narration follows all 6 workflow steps
• Top-down vs bottom-up divergence explained if >2×