The short answer
TAM, SAM and SOM are three nested measures of a market. TAM (total addressable market) is the annual revenue if every potential buyer in the category bought. SAM (serviceable addressable market) is the part your product and sales channels can actually reach. SOM (serviceable obtainable market) is the revenue you can realistically win against competitors with your current capacity, usually stated for year 3. Build them bottom-up: number of buyers, times units per buyer, times price.
What TAM, SAM and SOM measure
A market size answers one question for an investor or a board: is there enough revenue available for this plan to be worth doing? Each of the three figures answers a narrower version of it.
| Measure | The question it answers | What limits it |
|---|---|---|
| TAM | How big is the whole category, if every buyer bought? | Only the definition of the category |
| SAM | How much of that can our product and channels serve? | Segments, geographies and buyer types you can reach |
| SOM | How much can we realistically win by year 3? | Competitors, sales capacity and win rates |
The three figures matter for different decisions. TAM tells an investor whether the category is worth backing. SAM tells a board whether the current business has room to grow. SOM is the only one of the three that belongs in a revenue plan, because it is the only one limited by what the company can actually sell.
Top-down against bottom-up
A top-down estimate starts from a published industry total and takes a percentage of it: “the compliance software market is $X billion, and we can win 1%”. It is fast, but the percentage is a guess and the industry total rarely matches your product’s definition.
A bottom-up estimate builds the market from its units: the number of buyers who fit, times the units each buys (seats, sites, users), times the average contract value. Every input can be checked, and changing one shows how the answer moves. Bottom-up is the method investors and boards trust, and the one to use whenever the inputs exist.
As an illustration of the arithmetic: 8,000 firms that fit, with an average of 50 seats each at $1,500 a seat a year, gives 8,000 × 50 × $1,500 = $600 million a year.
How to size a market in six steps
1. Define the buyer precisely
Name the type of organisation, its size range, its geography and the job your product does for it. “Mid-sized professional-services firms in North America that bill by the hour” can be counted; “businesses that need productivity” cannot.
2. Count the buyers
Use a source you can name: industry associations, government business registers, company databases. Record the source next to the number.
3. Set units and price from your own data
Take the average seats per customer and the average contract value from your current customers, not from a competitor’s marketing. Multiply to get TAM.
4. Cut TAM to SAM
Remove the segments your product or channels cannot serve today: the wrong size, the wrong region, requirements you do not meet. State each exclusion and its size.
5. Cut SAM to SOM
Start from sales capacity: how many deals your team can close in a year, at your current win rate, against the competitors present. SOM is what that capacity produces by year 3.
6. Find where the profit sits
A profit pool is the share of the industry’s profit, not revenue, earned in each segment. The largest segment by revenue is often not the most profitable one, and the plan should aim at the profit.
Worked example: Northwind
The scenario
Growth down from 40% to 12%, and a board asking why
Northwind is a B2B software company with $40 million of annual recurring revenue (ARR), the subscription income that arrives every year. Its growth has fallen from 40% to 12% in a year, and net revenue retention, what last year’s customers spend this year, has slid from 118% to 96%. Before approving a plan, the board wants to know whether the market is the problem.
Sized bottom-up from buyer counts, seats and average contract value, the total category is worth $1.2 billion a year. The part Northwind’s product and channels can serve is $630 million, and the new revenue it can realistically win by year 3 with its current sales capacity is $19 million.
The ratios answer the board’s question. Northwind’s $40 million of ARR is about 6% of its $630 million SAM, so the market is not what stopped growth. The constraint is internal: existing customers stopped expanding, because the company was acquiring price-sensitive buyers who took minimum seats and never upgraded.
The profit-pool view pointed the plan in a direction. The most profitable segment was firms that bill by the hour, which buy on time saved. That became Northwind’s ideal customer profile (ICP), the description of the customer it should spend its sales effort on.
Why AI market sizes are often wrong
Ask an AI tool “what is the TAM for my product?” and it will usually return a large, confident number. The problems are in where the number comes from.
| Where it goes wrong | What usually comes back | What a board needs |
|---|---|---|
| Method | A top-down industry total, source unstated | Buyers × units × price, each input sourced |
| Figures | Market reports it cannot verify, sometimes invented | Your own customer data, with estimates marked |
| SOM | A round percentage of TAM | Built from sales capacity and win rate |
| Conclusion | “The market is large and growing” | Whether the market is the constraint, and which segment to win |
Try it: a shortened Prompt 04
This is a shortened version of Prompt 04 from the Strategy & Consulting pack, set up to run on its own. Paste it into ChatGPT, Claude or Gemini and fill in the brackets.
You are a former strategy consultant who has sized markets for private equity due diligence. A market size is only as good as its inputs, so you build it bottom-up from buyers, units and price, name the source of every input, and never quote an industry figure you cannot trace. THE BUSINESS: - What you sell and to whom: [product, buyer type] - Your current revenue and number of customers: [figures] - Average units per customer and average contract value: [e.g. seats per customer, $ per seat per year] THE MARKET: - Who could buy, and how many of them there are: [buyer definition, count and its source] - Segments or regions your product or channels cannot serve today: [list] - Your sales capacity and win rate: [deals a year, win rate, main competitors] Produce: 1. TAM. Buyers x units x price, with the arithmetic shown and each input's source. 2. SAM. TAM minus each segment you cannot serve, with the size of every exclusion. 3. SOM BY YEAR 3. Built from sales capacity and win rate, not a percentage of TAM. 4. THE CONSTRAINT. Whether market size limits growth, using current revenue as a share of SAM, and the one segment where the profit sits. Mark every figure you estimated as an estimate. If a count or price is missing, list it as a gap; do not fill it from memory.
The full Prompt 04 has eight parts rather than four, including growth vectors ranked by size and speed, and it reads the Strategy Brief built by prompts 01 to 03, so the sizing tests the question the board is actually asking.
Checks before you show it
- Every input has a named source, or is marked as an estimate.
- TAM, SAM and SOM are each shown with their arithmetic.
- Every exclusion from TAM to SAM is listed with its size.
- SOM is built from sales capacity, not a percentage of TAM.
- The conclusion says whether the market is the constraint.
Questions
What is the difference between TAM, SAM and SOM?
TAM is the whole category if every buyer bought. SAM is the part your product and channels can serve. SOM is what you can realistically win by year 3, given competitors and your sales capacity.
Should I use top-down or bottom-up market sizing?
Bottom-up wherever you can count the buyers and know your price. Top-down is useful only as a cross-check, because its percentage share is an assumption and its industry total rarely matches your product.
What SOM do investors expect?
One they can trace to sales capacity and win rates. A SOM stated as a round share of TAM, such as 1%, tells an investor the number was not built.
Can ChatGPT calculate my market size?
It can do the arithmetic and structure the method well. It cannot supply reliable buyer counts or industry totals from memory, so give it your inputs and their sources, and ask it to mark anything it estimated.
Written by a former Gartner Managing Partner and investment banking SVP
The guides and the prompt templates on this site come from a career spent building these documents: board decks, forecasts, business cases and hiring decisions, as Managing Partner at Gartner, SVP in investment banking and Country Manager at international subsidiaries. The worked examples are published in full on each product page. Browse the Prompt Library.
Northwind is a fictional company and its figures are illustrative.
Output from any AI tool should be reviewed before use.