TAM SAM SOM Explained: How Startups Size a Real Market

Our market is worth $10 billion.” Investors hear some version of this sentence in nearly every pitch, and most of them discount it immediately. A large industry number is not an opportunity. It describes how much money moves through a sector, most of which belongs to other companies, other customer types, other countries or other product categories that a young startup cannot serve.

TAM SAM SOM is the framework founders use to move from a big, impressive number to a defensible one. It takes you from the total market, to the relevant market, to the realistically obtainable market. Done well, it tells you whether a business can reach the revenue it needs to survive, raise capital and grow. Done badly, it is decoration on a pitch deck.

Realistic market sizing matters well beyond fundraising. It shapes startup validation, business planning, product strategy, go-to-market planning, revenue forecasting and competitive analysis. A founder who knows the serviceable market is 140,000 companies, not 40 million, will choose different channels, hire differently and price differently. This guide explains the reasoning behind each layer, shows how to calculate them with formulas and a worked example, and covers how to keep the numbers honest.

Quick answer: what are TAM, SAM and SOM?

• TAM (Total Addressable Market): the total annual revenue available if every potential customer for your type of product bought it.
• SAM (Serviceable Available Market): the portion of TAM that your product, pricing, geography and distribution can actually serve.
• SOM (Serviceable Obtainable Market): the portion of SAM you can realistically win within a defined time horizon, given your team, budget, competition and sales capacity.

What Is TAM?

Definition

Total addressable market (TAM) is the total annual revenue opportunity for a product or service category if you captured 100% of the demand, with no competitors, no geographic limits and no capacity constraints. It is a theoretical ceiling.

What TAM represents and what it does not

TAM represents the maximum revenue pool for the type of problem you solve. It answers: if everyone who could reasonably buy this did buy it, how large is the pool?

TAM does not represent:

• Your expected revenue
• The revenue of the industry you operate in
• The number of people who have the problem and would pay to solve it
• Your forecast for year five

The most common misuse is substituting an industry figure for a true addressable market. “The global healthcare industry is worth trillions” tells you nothing about a startup building scheduling software for dental clinics. The addressable market is dental clinics multiplied by what they would pay for scheduling software, not healthcare spending.

Industry vs. addressable market

An industry is defined by how analysts classify activity. An addressable market is defined by your product’s ability to replace spending or create new spending. A payments startup does not address the entire financial services industry, only the share of payment-related fees or software budgets it can displace. The test: whose budget line item does my product replace or create?

TAM across business models

• SaaS: potential accounts × annual contract value (ACV). A project management tool for construction firms counts construction firms, not all businesses.
• Services: potential clients × annual spend on that service. Services TAM is often bounded by the existing outsourced spend, because many buyers do the work in-house.
• Marketplaces: the total transaction value (GMV) in the category × your take rate. Many founders quote GMV as TAM, which overstates revenue by a factor of five to twenty, depending on commission.
• Consumer products: potential buyers × purchase frequency × price. Be careful with “everyone with a phone” assumptions.
• B2B businesses: accounts that match a specific profile (size, industry, tech maturity) × annual spend. B2B markets are usually smaller in count but larger in value per customer, which makes segmentation critical.

Practical TAM example

A startup builds invoice-automation software for independent veterinary clinics in the United States. If there are roughly N clinics (verify the count from an industry association or government data) and each would plausibly pay $2,400 a year, then TAM = N × $2,400. It does not include human medicine, pet insurance or the broader pet industry, even if a market report headline says “the pet industry is worth hundreds of billions.”

What Is SAM?

Definition

Serviceable available market (SAM) is the part of TAM that your current product, pricing model, geography and distribution can serve. It is the market you could address today or in the near term with the business you are actually building.

How SAM narrows TAM

Every filter you apply removes revenue that exists in theory but is out of reach in practice:

• Geography: a startup launching in the United States cannot serve Germany without localization, data residency and local sales.
• Customer segment: if your product suits 20–249 employee firms, enterprises and micro-businesses fall out of SAM.
• Product capability: a tool that integrates with only two accounting platforms cannot serve users of others.
• Pricing: if your price is $6,000 a year, businesses whose realistic budget is $500 are not serviceable.
• Industry restrictions: some sectors have procurement rules or certifications you lack.
• Regulatory restrictions: health data, financial data and children’s data trigger compliance requirements that determine who can buy from you.
• Distribution limitations: if your product only sells through channel partners, customers outside partner coverage are out of reach.
• Technology limitations: a cloud-only product excludes customers who require on-premise deployment.

Why SAM is more useful than TAM for planning

TAM is a ceiling. SAM is the playing field. Hiring plans, channel choices, product roadmaps and early revenue models all depend on SAM, because SAM describes real customers you can name, find and call. Investors who ask “who is your customer?” are usually probing your SAM. A founder who can describe it as “144,000 mid-sized firms in regulated industries that use cloud infrastructure and have no in-house security team” sounds credible. A founder who says “every business with a computer” does not.

What Is SOM?

Definition

Serviceable obtainable market (SOM) is the portion of your SAM that you can realistically capture within a stated time horizon, typically three to five years, given your resources and competition. It is the market-sizing number closest to a revenue plan.

Why SOM must be realistic

SOM is where the framework stops being theoretical. It forces you to confront the constraints that limit how fast a company can win customers:

• Sales capacity: a salesperson can only run so many deals per year.
• Distribution: channels take time to build and ramp.
• Competition: incumbents with brand trust, bundles and existing contracts defend their customers.
• Customer acquisition: paid channels have finite efficient spend before costs rise.
• Pricing: higher prices reduce conversion; lower prices reduce revenue per customer.
• Team size: support, onboarding and engineering capacity limit how many customers you can serve well.
• Geographic reach: you can only sell where you have language, compliance and presence.
• Brand awareness: unknown brands convert at lower rates, especially in security, finance and health.
• Capital constraints: the runway determines how much of the funnel you can fund.
• Time horizon: SOM for year one differs sharply from SOM for year five.

SOM is not a random percentage of SAM

“We only need 1% of the market” is not an assumption; it is a hope. SOM should be derived from operational assumptions: how many leads you can generate, how many convert, how many salespeople you can hire, how long deals take and how many customers stay. If you cannot trace SOM back to those inputs, it is not yet a SOM.

TAM vs SAM vs SOM

	TAM	SAM	SOM
Full name	Total Addressable Market	Serviceable Available Market	Serviceable Obtainable Market
Meaning	Total revenue if all potential customers bought	Portion of TAM your product and business model can serve	Portion of SAM you can realistically win in a set period
Market scope	Broadest	Narrower, defined by fit and reach	Narrowest, defined by capacity and competition
Main question	How big could this get in theory?	Who can we actually serve?	What can we realistically capture?
Typical use	Show the ceiling and category importance	Plan product, pricing, channels	Build revenue forecasts and hiring plans
Example	All SMBs needing cybersecurity software	SMBs in regulated industries reachable through cloud and MSP channels	Customers won within three years through a defined sales team and partner network

The relationship: TAM → SAM → SOM

Think of three nested filters. TAM asks whether the problem is worth solving at scale. SAM asks whether your approach can reach enough of it. SOM asks whether your company, with its people and money, can win enough of it to matter. Each step replaces optimism with a constraint. A healthy business plan shows how each layer was derived, so a reader can follow the logic and challenge any step.

TAM SAM SOM Example: AI-Powered Cybersecurity for Small and Mid-Sized Businesses

Illustrative example only. The numbers below are fictional assumptions chosen to show the method. They are not industry statistics, and a real analysis must replace each with sourced data.

The fictional startup, “ShieldLoop,” sells AI-assisted threat detection and automated response software to small and mid-sized businesses in the United States, with an average annual contract value (ACV) of about $6,000 for its core segment.

Step 1: TAM

• Assumed US employer businesses with 1–499 employees: 6,000,000 (to be verified from US Census Bureau business statistics)
• Assumed annual spend on this kind of security software per business: $3,000 (blended across sizes)
• TAM = 6,000,000 × $3,000 = $18.0 billion

Note that spend per business is blended. A 3-person firm spends far less than a 300-person firm, which is why SAM should use a different, segment-specific figure.

Step 2: SAM

Apply filters one at a time and write down each rationale:

Filter	Assumption	Remaining customers
Starting point	US businesses, 1–499 employees	6,000,000
Company size 20–249 employees (ShieldLoop’s fit)	12% of base	720,000
Data-sensitive industries (healthcare, legal, financial, professional services)	40% of that segment	288,000
Cloud-first IT and no in-house security team; reachable by sales or partners	50% of that group	144,000

• Expected ACV for this segment: $6,000
• SAM = 144,000 × $6,000 = $864 million

SAM is roughly 4.8% of TAM. That is a smaller number, but it describes customers ShieldLoop can name, segment and contact.

Step 3: SOM (three-year horizon)

Build from sales capacity, not percentages.

Direct sales:

• 8 account executives at full productivity by year three
• Each handles 125 qualified opportunities a year
• Win rate: 20%
• New customers per AE: 25
• Direct new customers: 8 × 25 = 200 per year

Partner channel (managed service providers):

• 25 active partners
• Each brings 8 new clients a year
• Partner new customers: 25 × 8 = 200 per year

New customers by year: Year 1: 120, Year 2: 280, Year 3: 400 (ramping as the team and partners mature).

Retention: assume 90% annual logo retention.

Year	Retained from prior year	New customers	Ending customers
1	0	120	120
2	108	280	388
3	349	400	~749

• Blended ACV (direct customers pay more, partner deals carry discounts): $5,500
• SOM (year-three ARR) = ~750 × $5,500 ≈ $4.1 million

That is about 0.5% of SAM and roughly 0.02% of TAM. The headline TAM of $18 billion tells you the problem is large. The SOM tells you what the company can build. Both are useful, but only the second belongs in the operating plan.

How to Calculate TAM

Top-down market sizing

Top-down starts with a large published figure and narrows it down.

Typical sources: industry reports, market research firms, government data, industry associations, public company annual reports and existing market datasets.

Example logic: national spending on a category × share applicable to your segment × share addressable by your product.

Strengths: fast, familiar to investors, useful as a sanity check.

Weaknesses: published figures often define categories loosely, differ between research firms, may be paywalled and obscure their assumptions. Top-down figures also tempt founders to apply a convenient percentage without evidence.

Bottom-up market sizing

Bottom-up builds the market from units:

TAM = number of potential customers × realistic annual spend per customer

Example: if 40,000 independent accounting firms could buy a document-automation tool at $1,800 a year, TAM is $72 million. Each element can be verified: firm counts from government or association data, price from competitor pricing pages and customer interviews.

Bottom-up is generally more useful for startups because every assumption is individually testable. If the number of firms is wrong, you fix one cell. If the price is too high, you test lower prices. A top-down number, by contrast, is a black box you cannot adjust.

Value-based (value-theory) approach

This method estimates market size from the economic value your product creates, then assumes customers will pay a fraction of it.

Example: if automating a process saves a mid-sized logistics firm $60,000 a year in labor and errors, and customers typically pay 10–20% of savings for software, the price ceiling is around $6,000–$12,000 per year. Multiply by the number of firms with that problem.

This approach makes sense when you are creating a new category with no direct spending to benchmark, or when your product replaces labor rather than software. Its limitations are real: customers do not pay based on value created, but on alternatives, budgets and perceived risk. Savings claims are often inflated, and procurement rarely accepts them unverified. Use value-based sizing to estimate pricing headroom, then check it against actual willingness to pay.

How to Calculate SAM

Start with your TAM customer count and apply filters in sequence:

1. Geography: which countries or regions can you legally and operationally serve in the next two to three years?
2. Customer type: consumers, SMBs, mid-market or enterprise?
3. Company size: employee count or revenue band.
4. Industry: which verticals fit, and which require certifications you lack?
5. Product compatibility: does it integrate with the systems your customers use?
6. Pricing: can the target segment afford the price?
7. Regulations: do privacy, financial or sector rules block certain buyers?
8. Distribution: can you reach them through channels you have or can build?
9. Technology requirements: cloud vs. on-premise, language support, data residency.

Formula:

SAM = (TAM customers × geography % × segment % × fit %) × annual revenue per relevant customer

Worked example: A bookkeeping automation app for UK freelancers. Assume 4.2 million potential self-employed users (illustrative; check UK Office for National Statistics data). If 60% invoice clients regularly, 70% use a compatible bank connection and 40% would consider paid software, the relevant customers number 4.2M × 0.6 × 0.7 × 0.4 ≈ 706,000. At £120 per year, SAM ≈ £84.7 million. Keep currency and year consistent throughout.

How to Calculate SOM

SOM is the section where a model earns credibility. Build it from the operating levers.

Sales capacity

Customers per year = number of salespeople × qualified opportunities per rep × win rate

Account for ramp time: new reps typically take months to reach full productivity.

Funnel math

For marketing-led or product-led models:

New customers = traffic or leads × lead-to-opportunity rate × opportunity-to-customer rate

If you generate 4,000 leads a year, 15% become qualified opportunities and 25% close, you win 150 customers.

Other inputs to include

• Average contract value (ACV): consistent with segment and pricing.
• Sales cycle: a nine-month enterprise cycle delays revenue compared with a two-week SMB cycle.
• Retention: ending customers = prior customers × retention + new customers.
• Geographic reach: only count regions where you will have sales or partners.
• Marketing budget: divide by realistic customer acquisition cost (CAC) to find how many customers the budget can buy.
• Distribution channels: resellers, marketplaces and partners each have capacity and ramp.
• Competitive intensity: adjust win rates down in crowded categories.
• Available capital: you cannot hire the sales team the model assumes if the runway doesn’t fund it.

Why the “1% of SAM” shortcut misleads

A 1% share sounds modest, but it hides the questions that matter: How many customers is that? How many salespeople would it take? What would it cost to acquire them? Can the company serve them? In a fragmented SAM of 500,000 customers, 1% is 5,000 customers; at 30 new customers per rep per year, that is 167 rep-years. A percentage is an output of the operating model, never an input.

Top-Down vs. Bottom-Up Market Sizing

	Top-Down	Bottom-Up
Method	Start with a large market, narrow by filters	Build from customers × spend
Starting point	Industry or category revenue	Individual customer units
Data required	Reports, government and association statistics	Customer counts, pricing, usage, interview data
Accuracy potential	Moderate; depends on report definitions	Higher, if inputs are verified
Common weakness	Category mismatch; unexamined percentages	Underestimating the customer count or overestimating willingness to pay
Best use case	Sanity checks, established categories	Startups, new products, SOM and forecasting

Why combine both

Each method catches the other’s errors. If bottom-up gives $90 million and top-down gives $4 billion, one of them is wrong, and the gap is where the useful questions are. Did the report include spending you cannot address? Did your customer count miss a segment? Convergence within the same order of magnitude is a good sign. A triangulated model, with bottom-up as the primary method and top-down as a cross-check, is more persuasive and more accurate than either alone.

How to Find Reliable Data for TAM SAM SOM

Sources to use:

• Government statistics: national statistical agencies publish business counts, employment, household data and industry spending. In the United States, the Census Bureau and Bureau of Labor Statistics are common starting points; in Europe, Eurostat.
• Industry associations: member counts, surveys and benchmark reports, often more segment-specific than general research.
• Public company annual reports: filings show revenue by segment, customer counts and average revenue per user.
• Regulatory agencies: licensing databases provide actual counts of regulated businesses.
• Investor reports: venture and analyst reports provide category context, though they reflect the author’s thesis.
• Reputable market research: useful, but always read the methodology and definitions.
• Company pricing pages: evidence of what customers actually pay.
• Customer surveys and interviews: direct evidence of budgets and urgency.
• Industry databases: business directories and datasets for building prospect lists.
• Search behavior: keyword volumes indicate demand for a problem, not market size.
• Competitor research: customer counts, funding, hiring and product pages reveal where money flows.

Distinguish the type of data:

• Primary data: collected by you directly (interviews, surveys, pilot results).
• Secondary data: published by others (reports, statistics, filings).
• Estimates: calculated figures derived from data (for example, a segment share).
• Assumptions: judgments not backed by data yet.

Label each input in your model. State the year, geography and currency for every number. If a figure comes from a report you cannot verify, treat it as an assumption until validated.

TAM SAM SOM for Different Business Models

All examples below are illustrative structures, not market data.

SaaS

Count target accounts, multiply by ACV, then filter by size, industry and integrations. SOM follows sales capacity and net revenue retention. Example: 90,000 dental practices × $2,000 ACV.

Marketplace

Start with GMV in the category, then multiply by take rate for revenue TAM. SAM is the portion of transactions that could move online in your launched geographies. SOM depends on liquidity: you need enough supply and demand in one local market before expanding. Counting national GMV from day one overstates what is obtainable.

Consulting and Professional Services

Potential clients × annual spend on the service. SOM is limited by billable hours: consultants × utilization × rate. A five-person firm cannot capture a market of thousands of clients; capacity, not demand, is the ceiling.

E-commerce

Target shoppers × purchase frequency × average order value, filtered by shipping regions and product category. SOM relies on traffic, conversion rate and repeat purchase rate. Marketplace and search competition affect acquisition cost heavily.

Mobile App

Distinguish downloads from paying users. Revenue TAM = potential users × conversion to paid × subscription price. A typical mistake is multiplying all smartphone owners by a subscription price. Paid conversion rates are usually small, so model them explicitly and test with early cohorts.

AI Startup

Size the workflow or labor budget being replaced or augmented, not “the AI market.” If an AI tool assists claims processors, the relevant budget is claims-processing labor and software spend. Also account for inference costs, which affect pricing and margins.

Cybersecurity Startup

Segment by company size, regulatory exposure and existing security stack. Buyers often purchase through MSPs or resellers, so channel coverage defines SAM. Trust and certifications (for example, SOC 2) can gate entire segments.

Local Business

Households or businesses within a service radius × adoption rate × spend. Local businesses are bounded by geography and capacity, so SOM is often a count of customers per week multiplied by price. A realistic local SOM beats a national TAM that you will never serve.

TAM SAM SOM in a Startup Pitch Deck

Where it appears

• Market opportunity slide: presents TAM, SAM and SOM with the logic behind each, ideally bottom-up.
• Revenue model: connects pricing and customer type to the SAM calculation.
• Go-to-market strategy: explains how you will acquire the SOM customers.
• Competitive landscape: shows who already serves the SAM and how you differentiate.
• Financial projections: should be consistent with SOM and not exceed what it supports.

What makes it credible

• Sources with year and geography
• Bottom-up logic that investors can recompute
• Clear segments and named customer types
• A SOM linked to hiring and channel plans
• Consistency between market slide and financials

What makes investors doubt the numbers

• Huge unsupported TAM: a big number with no derivation.
• Mixing global and local markets: global TAM with a local-only go-to-market plan.
• Confusing revenue with market size: quoting a competitor’s revenue as the market.
• Double counting: the same customer in two segments.
• Unsupported percentages: “we’ll take 2%” without a sales model.
• Outdated data: figures from several years ago in a changing market.
• Ignoring competitors: implying the market is empty.
• Assuming immediate global reach: no localization, compliance or distribution plan.

Common TAM SAM SOM Mistakes

1. Using an enormous industry number as TAM. Industry revenue is not the revenue your product can address.
2. Calling the entire internet the target market. “Everyone online” is not a customer definition.
3. Making SAM almost equal to TAM without justification. If no filter applies, you probably have not defined your product or customer yet.
4. Choosing an arbitrary SOM percentage. Derive it from capacity.
5. Using outdated statistics. Markets, prices and company counts change.
6. Mixing different currencies. Convert all figures consistently and state the rate and date.
7. Mixing different years. Compare same-year values or adjust with clear assumptions.
8. Ignoring geography. Regulation, language and distribution differ by country.
9. Ignoring customer segmentation. A 5-person firm and a 500-person firm are different markets.
10. Confusing users with paying customers. Free users are not revenue.
11. Confusing revenue opportunity with total industry revenue. Your opportunity is the portion you can displace, not the sum of everyone’s sales.
12. Ignoring competition. Competitors reduce your obtainable share and raise acquisition costs.
13. Ignoring distribution capacity. The market is not reachable just because it exists.
14. Overestimating willingness to pay. Interest is not budget; test real prices.
15. Double-counting customers. A business appearing in two filters is still one customer.

TAM SAM SOM vs. Market Validation

Market sizing does not prove that customers want your product. It answers how much money could exist. Validation answers whether this product earns it.

• Market size: how large the revenue pool is.
• Market demand: whether people actively want solutions in this area.
• Customer problem: how painful and frequent the issue is.
• Willingness to pay: whether they allocate a budget, not just express interest.
• Product-market fit: whether your product solves the problem well enough that customers adopt, pay and stay.
• Competitive demand: whether demand is already satisfied by existing alternatives.

A huge TAM can be a poor opportunity: the problem may be mild, incumbents may be entrenched, switching costs may be high or customers may be unwilling to pay. Conversely, a modest, well-validated market with urgent pain can support a strong company. Practical validation includes customer interviews, pricing tests, pilot programs, letters of intent, waitlists with deposits and early revenue. For a broader framework on separating strong ideas from weak ones, read How to Find a Good Startup Idea.

TAM SAM SOM and Unit Economics

Market size says how many customers might exist; unit economics says whether serving them is profitable. The connection runs through several metrics:

• CAC (customer acquisition cost): total sales and marketing spend ÷ new customers.
• ARPC / ARPA: average revenue per customer or account.
• Gross margin: revenue minus cost of delivery (hosting, support, inference costs for AI).
• Retention: how long customers stay.
• LTV (lifetime value): ARPC × gross margin ÷ churn rate (a simple approximation).
• Payback period: CAC ÷ (monthly revenue × gross margin).

Illustration: If the SAM is large but each customer costs $9,000 to acquire and yields $6,000 in annual revenue at 70% gross margin, payback exceeds two years. Growing faster only accelerates cash burn. A large TAM does not fix that, and a very large SAM can even tempt a company to chase customers it cannot acquire profitably.

Use unit economics to test SOM: if acquiring the number of customers in your SOM requires more capital than you can raise, the SOM is not obtainable. Business and finance resources at ValuFlash can help you extend this analysis into pricing, margins and growth planning.

TAM SAM SOM and Financial Modeling

Market sizing feeds the financial model, but it is not the model. The connections:

• Revenue projections: SOM sets a plausible ceiling for a time horizon.
• Customer growth: monthly or quarterly new customers from funnel assumptions.
• Pricing: ACV and discounting patterns.
• Conversion rates: from lead to customer.
• Cash flow: timing of invoices, annual vs. monthly billing, collection delays.
• Hiring: headcount to support sales, onboarding, engineering and support.
• Marketing investment: spend required to produce the funnel.
• Fundraising requirements: capital to reach milestones before revenue covers costs.

Market opportunity ≠ revenue forecast. The market is an upper bound on what is possible. A forecast is a time-based plan for the revenue you will actually book. A forecast that reaches 30% of SAM in three years should trigger a hard look at every assumption. Build scenarios (conservative, base, aggressive) that vary conversion, retention and sales ramp rather than pulling a single number from the market slide.

TAM SAM SOM for AI Startups

AI startups need extra caution because the ground shifts quickly:

• Rapidly changing markets: what was impossible last year may be a feature in an existing product this year.
• Infrastructure costs: compute, storage and monitoring reduce margins.
• Model and API costs: pricing from model providers can change, affecting gross margin and price.
• Pricing uncertainty: customers are still learning what to pay for per seat, per usage or per outcome.
• Fast-moving competitors: incumbents add AI features; new entrants copy workflows.
• Commoditization: capabilities that look unique can become standard quickly.
• Changing customer behavior: adoption and trust evolve, and so do procurement and compliance requirements.
• New categories: if a category does not exist yet, there is no established spending to size.

It is tempting to claim that AI expands every market. Sometimes it does, by making previously uneconomic services viable. Often it simply shifts budgets between existing vendors, or lowers prices as competition grows. Do not assume a growth premium; show where the new spending comes from.

Practical approach: define one specific workflow in one customer segment. Estimate the labor or software budget currently allocated to it. Interview buyers about willingness to pay. Model inference costs per task and test whether gross margin stays healthy at your price. Run pilots to measure real usage. Then size the market from verified customers and prices. To structure the operating side of an AI venture, see How to Build an AI-First Business System.

TAM SAM SOM and Technology Strategy

Technology decisions can expand or shrink your obtainable market:

• Product scalability: a product that needs custom setup for every client limits how many customers you can onboard.
• Infrastructure and cloud architecture: multi-region deployment enables sales in markets with data residency requirements.
• APIs and integrations: each integration opens a customer segment that uses that tool.
• Security: certifications and security design determine eligibility for regulated buyers.
• Geographic deployment: latency, language and legal requirements shape which countries are serviceable.
• Reliability: uptime commitments matter for enterprise and mission-critical use.
• Compliance: privacy and sector frameworks decide who can legally buy from you.

For example, adding a single integration with a widely used accounting platform might increase SAM by tens of thousands of customers, while lacking a specific compliance certification can remove an entire industry. These choices should be made with the market model in view. A practical walkthrough on matching tools to business goals is available in Choosing the Right Tech Stack.

From Market Size to Execution

Market sizing is the beginning of planning, not the end. The chain looks like this:

Market → Customer → Problem → Product → Pricing → Distribution → Sales → Retention → Revenue

Each link needs evidence:

1. Market: a defined segment, not “everyone.”
2. Customer: a named buyer and user with a budget.
3. Problem: a painful, frequent problem confirmed in interviews.
4. Product: the smallest solution that solves it.
5. Pricing: tested with real buyers.
6. Distribution: channels with measured conversion.
7. Sales: a repeatable process and capacity plan.
8. Retention: customers who stay and expand.
9. Revenue: the outcome of all earlier links.

Convert SOM assumptions into an operating plan: quarterly customer targets, hiring dates, marketing budgets, product milestones and infrastructure needs. Revisit the assumptions each quarter as real data replaces estimates. Delivering on this plan also requires disciplined engineering, release management and scaling practices, which The Real Engineering Process Behind Successful Startups explores in detail.

Skills Required to Do Market Sizing Well

Good market sizing is a blend of analytical and commercial skills:

• Market research: finding and evaluating sources.
• Excel or Google Sheets: building transparent models with separate assumptions.
• Financial modeling: linking market size to forecasts and cash flow.
• Data analysis: cleaning datasets, spotting outliers, segmenting customers.
• Competitive research: understanding positioning, pricing and customer bases.
• Customer interviews: extracting budget, urgency and buying process.
• Pricing analysis: comparing alternatives and testing willingness to pay.
• Business strategy: deciding which segment to enter first and why.
• Basic statistics: sample size, confidence, ranges and sensitivity.
• Industry research: knowing regulation, procurement norms and value chains.

None of these requires a specialist degree, but all benefit from practice. Founders and analysts who build these capabilities deliberately make better decisions across the company; see Building Skills That Create Long-Term Business Value for a practical perspective on developing them.

Practical TAM SAM SOM Worksheet

Copy this into a spreadsheet and fill in one column of sources for each line.

Scope

Product: ___
Customer definition: ___
Geography: ___
Currency and year: ___
Time horizon: ___

TAM

Potential customers: ___
Annual spend/customer: ___
TAM: ___
Source and year: ___

SAM

Relevant customers (after filters): ___
Filters applied (geography, size, industry, fit, regulation): ___
Annual spend/customer (for this segment): ___
SAM: ___
Source and year: ___

SOM

Salespeople / channels: ___
Qualified opportunities per year: ___
Win rate: ___
Reachable customers per year: ___
Retention rate: ___
Expected annual revenue/customer: ___
Estimated SOM: ___
Capital required to execute: ___

How to challenge every assumption

For each blank, ask:

• Where did this number come from? Is it primary, secondary, an estimate or an assumption?
• What happens to the total if this number is 30% lower?
• Is the year, geography and currency consistent?
• Could the same customer appear twice?
• What does a skeptical investor or customer say about it?
• Can I test it this month with an interview, a landing page or a pilot?

Run low, base and high cases and rank the assumptions by their influence on the result. Spend your validation effort on the top three.

TAM SAM SOM Checklist

• Defined customer: a specific buyer and user, not “everyone.”
• Defined geography: countries or regions you can serve in the time horizon.
• Defined product: what is included, and what is not.
• Defined pricing: tested price points and billing model.
• Reliable market data: sources identified with year and geography.
• Bottom-up calculation: customers × spend built from units.
• Top-down cross-check: comparison against published figures.
• Competitive analysis: incumbents, substitutes and switching costs mapped.
• Distribution assumptions: channels, partners and expected conversion.
• Sales capacity: headcount, ramp time and quota assumptions.
• Time horizon: SOM stated for a specific period.
• Revenue assumptions: ACV, retention and expansion justified.
• Customer validation: interviews, pilots or pre-sales supporting the inputs.

Frequently Asked Questions

What does TAM stand for?

TAM stands for Total Addressable Market, the total annual revenue opportunity if every potential customer for your type of product bought it.

What does SAM stand for?

SAM stands for Serviceable Available Market, the part of TAM that your product, pricing, geography and distribution can serve.

What does SOM stand for?

SOM stands for Serviceable Obtainable Market, the portion of SAM you can realistically capture within a defined time period.

What is the difference between TAM, SAM and SOM?

TAM is the theoretical maximum, SAM is the relevant segment you can serve and SOM is what you can realistically win given your resources, competition and timeline. Each is a subset of the previous one.

How do you calculate TAM?

Bottom-up: number of potential customers × annual spend per customer. Top-down: start with a published category figure and narrow it to your applicable segment. Using both and comparing results is the strongest approach.

How do startups calculate SOM?

By modeling operational capacity: salespeople × opportunities × win rate, plus channel and marketing-driven customers, adjusted for retention and ramp time, then multiplied by expected revenue per customer over a defined horizon.

Is SOM always a percentage of SAM?

No. A percentage can be an output used to sanity-check SOM, but it should not be the input. SOM should result from sales capacity, funnel conversion, pricing and constraints.

What is a good TAM for a startup?

There is no universal number. Venture-backed companies generally need a market large enough to support a very large company, so investors often look for markets with substantial revenue potential. But the quality of the derivation matters more than the size. A smaller, well-defined TAM with strong economics can be a better business than a vague, huge one.

Can TAM change over time?

Yes. New technology, regulation, pricing changes and customer behavior can enlarge or shrink a market. Review your estimates regularly and update inputs with fresh data.

Why do investors care about TAM?

Because it indicates whether a company can grow into a large business and whether the upside justifies the risk. They also read your approach to TAM as evidence of how rigorously you think.

What is bottom-up market sizing?

It is a method that builds market size from individual units: customers multiplied by realistic annual spend, with each input sourced or tested. It is usually more transparent and defensible than top-down estimates.

Can a startup have a small TAM and still be successful?

Yes. Many profitable companies serve focused markets with strong margins, loyal customers and low acquisition costs. A small TAM can work for bootstrapped businesses, niche software or companies that expand into adjacent markets over time. It is a harder fit for venture funding that expects very large outcomes.

Final Takeaway

TAM tells you how large the theoretical opportunity could be. SAM tells you which part is actually relevant. SOM forces you to confront what the business can realistically capture.

Credible market sizing is not about producing the biggest possible number. It is about building a chain of assumptions that a smart skeptic can follow, challenge and still find reasonable. Start bottom-up, cross-check top-down, label every input, connect SOM to your sales and financial plans and validate with real customers. The founders who do this are not the ones with the largest slide, but the ones who understand their market well enough to win it.