TAM, SAM and SOM are three nested estimates of how much revenue your market holds. TAM is the whole market, SAM is the slice your business model can actually reach, and SOM is the share you can realistically win in the near term. Each is a subset of the one before it, and the most defensible way to calculate them is to multiply a count of customers by what each one pays per year.
This guide gives you the definitions, the formulas, two ways to size a market (top-down and bottom-up), worked examples with real arithmetic, and a checklist for making your numbers credible to an investor, a board or your own team.
TL;DR
- TAM is total annual revenue if you won 100% of your market. SAM is the part you can serve with your current model, geography and channels. SOM is the part you can realistically capture soon.
- The basic formula for all three is customers × annual revenue per customer. Only the customer count changes.
- Use bottom-up for any decision that costs real money, and use top-down as a cross-check. If the two disagree badly, an assumption is wrong.
- Present a range, show your sources and make every number traceable to a page someone can open.
- Define the market narrowly first. The definition decides the number more than any formula does.
Table of Contents
- What TAM, SAM and SOM mean
- How TAM, SAM and SOM differ and relate
- The formulas
- Top-down vs. bottom-up market sizing
- Worked example: B2B software
- More examples across industries
- Step by step: size your own market
- Making the numbers credible
- Common mistakes
- Our perspective on market sizing with live data
- Run your market sizing with Prowl
- FAQ
- Sources
What TAM, SAM and SOM mean
According to Stripe’s market sizing guide, TAM (total addressable market) is the total annual revenue a product or service would generate if it captured 100% of its market. SAM (serviceable addressable market) is the portion of TAM a company can reach with its current business model, geography and distribution channels. SOM (serviceable obtainable market) is the share of SAM it can realistically capture in the near term, given competition and its own resources, such as sales capacity and marketing budget.
HubSpot’s explainer frames them the same way: TAM is the maximum potential demand, SAM is the size of the TAM you can reasonably target, and SOM is the size of the SAM you can potentially convert. It also notes that these figures feed a business plan, helping you set realistic revenue goals and choose which markets to enter.
The analyst firm Forrester recommends three core types of market sizing: TAM, SAM and SOM. In its demand-unit approach, target demand equals the SAM, because it reflects the number of potential demand units in the market.
A plain-language version: TAM answers “how big is the pond?”, SAM answers “how much of the pond can our boat reach?”, and SOM answers “how many fish can we land this year or next?”
How TAM, SAM and SOM differ and relate
Stripe points out that each figure is a subset of the one before, so TAM is always the largest number and SOM the smallest. TAM counts every potential buyer, SAM counts the buyers you can serve, and SOM counts the buyers you can expect to win.
| TAM | SAM | SOM | |
|---|---|---|---|
| Question it answers | How big is the whole opportunity? | How much can our model reach? | How much can we win soon? |
| What narrows it | Market definition | Geography, segment, product fit, channels | Competition, capacity, budget, sales cycle |
| Typical time frame | Long-run ceiling | Medium term | 12 to 36 months |
| Main audience | Strategy, investors | Product and marketing | Sales, finance, investors |
| Size | Largest | Middle | Smallest |
Two habits keep the three straight. First, TAM is an upper limit, not a forecast: Stripe stresses that it assumes every potential customer buys and nobody competes. Second, the market definition drives everything. Stripe’s example is a startup selling payroll software to restaurants: its TAM is what restaurants spend on payroll software, not what all businesses spend on software.
The formulas
The bottom-up formulas in Stripe’s guide are the simplest place to start:
- TAM = total number of potential customers × annual revenue per customer
- SAM = number of potential customers you can serve × annual revenue per customer
- SOM = number of customers you can realistically win × annual revenue per customer
The price per customer stays roughly constant. What shrinks is the customer count, as you apply filters: region, company size, industry, whether your product fits their workflow, and whether you can reach them with the channels you have.
TechCrunch describes three ways to calculate TAM: top-down, using publicly available market research reports; bottom-up, using actual and projected pricing plus current and total projected use; and value theory, which applies to entirely new products, services and categories where no market exists yet. Value theory estimates the value delivered and then how much of it can be reflected in pricing.
Top-down vs. bottom-up market sizing
Top-down: start big and filter
The World Bank describes the top-down approach, also called the chain ratio method, as defining a “universe” target market and applying filters that keep reducing it to a “net” market. Its example: 20,000,000 adults × 20% farmers × 70% with mobile access × 40% who qualify = 1,120,000 people of market potential.
Top-down is fast and handy for a first sense of scale. Its weakness is that every percentage is an estimate. Scaler notes that top-down figures often inflate market size because each narrowing percentage is one guess multiplied by someone else’s guess.
Bottom-up: build from unit economics
Precise Research Solutions describes bottom-up sizing as building the market upward from granular unit economics: market size = total qualified buyer accounts × annual contract value (or ARPU). It asks you to identify discrete buyer profiles, verify pricing through primary research and count real account densities in specific territories.
The World Bank’s description is similar: identify the customer segments you intend to reach, then estimate their size and growth, in some cases with a partner’s customer data to sharpen the picture.
Which one to use
Scaler’s verdict is practical: use bottom-up for any decision that costs real money, and top-down only as a rough cross-check. A bottom-up number tells you which assumption broke when reality diverges from it. A top-down number gives you nothing to debug.
| Top-down | Bottom-up | |
|---|---|---|
| Starting point | Large market or industry total | Count of buyers and a price |
| Speed | Fast | Slower |
| Main risk | Stacked guesses inflate the result | Missing segments or wrong price |
| Best for | Early sense of scale, sanity checks | Budgets, hiring, fundraising, forecasts |
| Debuggable | Rarely | Yes, assumption by assumption |
Worked example: B2B software
Precise Research Solutions gives a clean example for automated compliance audit software sold to medical device manufacturers:
- TAM: 35,000 firms × $40,000 = $1.4 billion
- SAM: 6,000 qualified North American accounts × $40,000 = $240 million
- SOM: 300 won accounts × $40,000 = $12 million, or 5% share of the focused territory
Notice what happened between each step. TAM to SAM was a geography and qualification filter (35,000 firms down to 6,000). SAM to SOM was a sales reality filter (6,000 accounts down to 300). The price never moved.
If you sell software or an analytics product, the same structure applies. A team deciding where to focus content analytics for SaaS growth, for instance, would define SAM as the companies it can reach with its current channels and estimate SOM from the accounts it can close in the next year.
More examples across industries
Laundromat software. HyperResearch’s example counts 20,114 locations for a TAM of $42.4 million. SAM narrows to 10,890 staffed stores at $2,988 each, or $32.5 million. SOM is 540 stores × $2,988 = $1.6 million.
A paid tax-filing app in India. Scaler’s worked example starts with roughly 8.5 crore individual income-tax returns filed per year, 40% of which are self-filed (about 3.4 crore) and 85% of those on a smartphone (about 2.9 crore). At an annual price of ₹499, SAM is about ₹1,447 crore a year. A realistic year-three SOM of 2.3 million users gives about ₹116 crore a year. It shows how a consumer product can use the same chain of filters.
A healthcare finance product. The XLSlides calculator preset multiplies 18,500 potential customers by $120,000 in annual revenue for a TAM of $2.22 billion. Filters of 100% geography, 28% segment and 42% product fit leave 11.8% of TAM, so SAM is $261,072,000. A 4% capture rate gives a SOM of $10,442,880.
Specialty foods. In TechCrunch’s example, the plant-based foods category is the SAM at $7 billion. The “olives and peppers” category within it, at $2.5 billion, is the SOM. The company estimates 0.5% of the U.S. market in year one, 1% in year two and 3% in year three.
Across all four, the pattern holds: define the buyer, count them, multiply by price, then filter by what you can reach and win.
Step by step: size your own market
- Define the product and the buyer in one sentence. Be narrow. “Payroll software for restaurants” beats “business software.”
- Count potential customers. Use government and public databases such as the US Census Bureau, the Bureau of Labor Statistics and NAICS industry data, industry associations, and public company filings, the sources Praxia Insights recommends.
- Set annual revenue per customer. Use your real pricing, or pricing of comparable products if you have not launched.
- Calculate TAM. Customers × annual revenue per customer.
- Apply filters for SAM. Geography, segment, product fit and channel reach. Write each filter as a percentage or a count with its source.
- Build SOM from capacity. Praxia advises building SOM from unit economics instead of taking a percentage of SAM: how many customers can you close per month at current capacity, at what average contract value, projected over 12, 24 and 36 months.
- Cross-check with the other method. If you built bottom-up, run a quick top-down and compare.
- Present a range. Praxia suggests stating that your TAM ranges from one figure to another depending on how broadly you define the category.
Making the numbers credible
A credible estimate is one that someone else can rebuild. Ideoreto says a good bottom-up estimate shows its sources, formulas, assumptions and uncertainty, explains why the customer count and price are plausible, and compares itself with a top-down number. If the two are wildly apart, investigate rather than ignore the gap.
Kromatic’s guidance is stricter: a defensible TAM rests on at least two independent paths, one top-down from analyst or government data and one bottom-up from named-segment counts times annual contract value, that agree to within roughly 3 to 5 times. Every numeric claim should trace to a verifiable URL. It also observes that a polished $50 billion TAM pulled from one report is usually less credible than a $200 million TAM with airtight bottom-up math and visible sensitivity to its assumptions.
Investors read these numbers closely. TechCrunch’s author advises not to inflate estimates, because investors will expect you to achieve them and missing them can damage your credibility and your ability to raise money. They care most about SOM.
Common mistakes
- Treating TAM as a forecast. It is a ceiling that assumes no competition and universal adoption.
- Using one broad report as the whole answer. Cross-check it with a count of real buyers.
- Defining the market too widely. All software spend is not your market.
- Deriving SOM as an arbitrary 1% of SAM. Tie it to sales capacity, conversion and contract value.
- Burying assumptions. Show each filter and where it came from.
- Ignoring the gap between methods. A big disagreement between top-down and bottom-up is a signal that an input is wrong, and finding it is where the learning is.
Our perspective on market sizing with live data
We built Prowl for the moment a market-sizing spreadsheet stops being a template and becomes a research job. The formulas are easy. The slow part is finding buyer counts, pricing, competitor positions and demand signals, then keeping every figure tied to where it came from.
Prowl connects any AI agent or workflow to 444 marketing intelligence tools through one API. That means an agent can pull competitive, SEO, ad performance, pricing and market trend data into one workflow and cross-reference it, instead of you wiring up each source by hand. Our approach is designed so that each number in a report can be traced back to its source, which is exactly what a bottom-up estimate needs to survive scrutiny.
We also think a market-size estimate should be repeatable. When pricing or competitor data moves, an agent-driven workflow can rerun the same inputs and show which assumption changed, rather than leaving you with a stale slide.
Prowl Team
Run your market sizing with Prowl
If you size markets regularly, for clients, product bets or fundraising, Prowl lets your agent gather the inputs and build the report in one pass. Plans are published on the Prowl pricing page: Exploit is $60 per month, Blackops is $120 per month and Syndicate is $240 per month. If you prefer to pay as you go, pre-paid credits come in $10, $50 and $200 packs.
Want to see what the tool set covers first? Explore the Prowl market data platform or browse real-world use cases. For adjacent workflows, our guides on market share analysis and market entry analysis show how sizing feeds the next decision.
FAQ
What is the difference between TAM, SAM and SOM?
TAM is the total revenue available if you captured the whole market. SAM is the part of it your business model, geography and channels can reach. SOM is the share of SAM you can realistically win in the near term. Each is a subset of the previous one.
How do you calculate TAM, SAM and SOM?
Multiply a customer count by annual revenue per customer. Use all potential customers for TAM, customers you can serve for SAM, and customers you can realistically win for SOM. Check the result against a top-down estimate from analyst or government data.
Is top-down or bottom-up market sizing better?
For decisions that cost real money, bottom-up is the stronger choice because each input can be tested and corrected. Top-down is quicker and works as a cross-check. If the two disagree badly, review your assumptions before presenting either figure.
Can you give a simple TAM SAM SOM example?
For compliance audit software aimed at medical device makers, one published example uses 35,000 firms at $40,000 for a $1.4 billion TAM. It narrows to 6,000 North American accounts for a $240 million SAM, then 300 won accounts for a $12 million SOM.
How big should SOM be compared with SAM?
There is no fixed percentage. Build SOM from how many deals your team can close each month, at what average contract value, over 12, 24 and 36 months. The example above lands at 5% of the focused territory, but your own capacity decides yours.
Sources
- TAM, SAM, and SOM: How to Size Your Market (Stripe)
- TAM, SAM & SOM: What Do They Mean & How Do You Calculate Them (HubSpot)
- Types of Market Sizing: Use, Terms and Definitions (Forrester)
- How to calculate your startup’s TAM, SAM and SOM (TechCrunch)
- World Bank document on top-down and bottom-up market sizing
- Market Sizing: How to Calculate TAM, SAM and SOM (Precise Research Solutions)
- How Market Sizing Prevents Expensive Business Mistakes (Scaler)
- TAM SAM SOM Calculator for Market Sizing (XLSlides)
- How to Use AI to Build a TAM/SAM Analysis (Praxia Insights)
- Bottom-Up Market Sizing: Build a Realistic Estimate (Ideoreto)
- Secondary Market Research for Startups (Kromatic)