To calculate TAM (total addressable market), multiply the total number of potential customers for your product by the average revenue you could realistically earn from each one. There are three accepted ways to arrive at that number: top-down (starting from a broad industry figure and narrowing it), bottom-up (counting real potential customers and multiplying by your price), and value theory (estimating what buyers would pay for a genuinely new kind of product). For most B2B companies, bottom-up is the most defensible of the three, because it is built from a real, countable customer base instead of a percentage carved out of an industry report.
This guide is for founders and revenue teams building a TAM for a fundraising deck, a board update, or a go-to-market plan. It walks through all three methods with worked examples, explains how TAM relates to SAM and SOM, and covers the mistakes that make investors and buyers stop trusting the number.
What is total addressable market (TAM)?
TAM is the total revenue you could generate if every potential customer for your product bought it, at the price you charge, today. It answers one question: how big is the opportunity, not how much of it you will actually capture.
The basic formula is:
TAM = Total number of potential customers x Average revenue per customer
For a B2B product, "average revenue per customer" is usually your average contract value (ACV) or average annual recurring revenue per account. TAM is a ceiling, not a forecast. It ignores competition, sales capacity, and how much of the market you can realistically reach, which is what SAM and SOM are for (covered below).
What are the three ways to calculate TAM?
There are three standard methods, and most credible TAM estimates use at least two of them to cross-check each other:
- Top-down: start with a broad industry number from a market research firm or public data source, then narrow it down with segmentation assumptions until it matches your product.
- Bottom-up: start from your actual customer definition, count how many real companies (or individuals) fit it, and multiply by your price.
- Value theory: for a genuinely new category with no existing market data, estimate what target buyers would pay for the value your product creates, then multiply by an estimated buyer count.
How do you calculate TAM top-down?
- Find a broad market size figure from an industry report, analyst estimate, or public statistics (government census data, trade association numbers).
- Identify what share of that broad market is actually relevant to your product, using filters like geography, company size, or industry vertical.
- Apply those filters as percentages to narrow the broad number down to your segment.
- Multiply the resulting customer count by your average revenue per customer.
Example: an industry report says there are 2 million small businesses in your country. If your product only fits retail businesses with 10-50 employees, and that segment is roughly 6% of all small businesses, your addressable segment is about 120,000 businesses. At an average annual contract value of 2,000, that is a 240 million TAM.
Top-down is fast and easy to defend with a citation, but it inherits every assumption baked into the original report, and the "percentage of the market that's relevant to us" step is often a guess dressed up as a calculation.
How do you calculate TAM bottom-up?
- Write down your ideal customer profile (ICP) in concrete terms: industry, company size, geography, tech stack, or any other attribute that predicts a good fit.
- Count the actual number of companies that match that definition, using a company database, government business registry, or a tool that can search against your ICP description directly.
- Multiply that count by your average contract value.
- Sanity-check the result against your current pipeline: if your real-world win rate and deal size are very different from your assumption, adjust.
Example: your ICP is "B2B software companies with 50-500 employees that use a specific category of infrastructure tooling." You find that roughly 8,400 companies match that description. At an average contract value of 15,000, your bottom-up TAM is 126 million.
Bottom-up takes more work than top-down, but it is harder to dismiss in a due diligence conversation, because every number traces back to a real, countable group of companies instead of an industry-wide percentage. The step that usually breaks down in practice is step 2: getting an accurate, current count of companies matching a specific ICP, rather than a rough guess. Turning your best customers into that ICP definition is one starting point if you already have paying customers to learn from.
How do you calculate TAM with value theory?
Value theory is for products that create a new category, where there is no existing market size to start from. Instead of counting an existing market, you estimate the value your product creates for a buyer (time saved, revenue generated, cost avoided), estimate what share of that value a buyer would pay for, and multiply by an estimated number of buyers who would experience that value. It is the least data-grounded of the three methods and is best used as a directional estimate, cross-checked against a bottom-up count of the buyers you believe exist.
How is TAM different from SAM and SOM?
TAM, SAM, and SOM answer three different questions, and conflating them is one of the most common mistakes in a TAM calculation.
| TAM | SAM | SOM | |
|---|---|---|---|
| Question answered | How big is the total opportunity? | How much of it can this product actually serve? | How much can this company realistically capture? |
| Narrowed by | Nothing yet, it's the whole market | Product fit, geography, regulatory limits | Sales capacity, competition, go-to-market reach |
| Typical use | Investor pitch, market context | Go-to-market planning | Revenue forecasting, quota setting |
SAM (serviceable addressable market) is the slice of TAM your product can actually serve, once you exclude segments you don't support (wrong geography, wrong company size, a regulatory requirement you don't meet). SOM (serviceable obtainable market) is the slice of SAM you can realistically win in a given period, given your sales team size, brand, and competition.
What mistakes make a TAM calculation unconvincing?
- Using a percentage with no source. "We estimate we can capture 10% of this market" without explaining why 10% and not 3% or 30% is the single fastest way to lose credibility with an investor or a buyer.
- Never cross-checking top-down against bottom-up. If a top-down estimate and a bottom-up count of your actual ICP are off by an order of magnitude, one of them is wrong, and it is worth finding out which before you present either.
- Treating TAM as a number you calculate once. A TAM built for a fundraising deck two years ago reflects a market that has since changed: new companies formed, others were acquired or shut down, and your own ICP may have narrowed or expanded as you learned who actually buys.
- Mixing markets. Rolling a total software market number into a TAM for a narrow vertical product inflates the number without making it more true.
- Skipping the source. Citing "industry estimates" without naming where a top-down figure came from invites the exact question it should have pre-empted.
How does Retriever help here?
The step that most TAM calculations fake is the bottom-up company count: knowing your ICP is "B2B software companies, 50-500 employees, using X" is one thing, knowing how many real companies actually match that description right now is another. Retriever is a semantic ICP search engine: you describe your ideal customer in plain English and get back a ranked list of the real companies that match, each with a match score and the evidence behind it, instead of a percentage assumption.
Because Retriever also tracks buying signals (hiring, funding, tech-stack, leadership, and social activity) on top of the match, the same search that grounds your bottom-up TAM can double as a live view of which companies in that market are showing a reason to buy this quarter, which is a different question than TAM answers but one revenue teams usually need next. For a broader look at how AI search tools compare on this kind of ICP matching, see our guide to the best AI tools for finding companies that match your ICP.
See how Retriever works or book a demo if you want to run your own ICP against real, current company data before you finalize a TAM number.
Frequently asked questions
What is a good TAM for a startup? There is no universal minimum, but most investors want to see a large enough opportunity to support a venture-scale outcome, which depends heavily on stage, sector, and business model. A well-supported, smaller TAM with a credible bottom-up calculation is generally viewed better than an inflated one that does not survive scrutiny.
Should I use top-down or bottom-up to calculate TAM? Use both if you can. Top-down is faster and easier to source for a first pass; bottom-up is more defensible because it is grounded in an actual count of matching customers. When the two disagree significantly, that gap is worth investigating before you present either number.
How often should I recalculate TAM? Recalculate whenever your ICP changes meaningfully, or at least once a year, since the underlying market moves: companies are founded, acquired, or shut down, and your own understanding of who buys tends to get sharper over time.
Is TAM the same as market size? TAM is one way of expressing market size, specifically the revenue ceiling if you captured every potential customer. "Market size" is sometimes used loosely to mean TAM, SAM, or SOM depending on context, so it is worth clarifying which one is meant.
Do I need TAM if I'm not fundraising? TAM is useful beyond fundraising for go-to-market planning, quota setting (alongside SOM), and prioritizing which segments to target first, so it is worth calculating even without an investor audience.