Open almost any pitch deck and you will find the same slide: three numbers, three nested layers, a billion-dollar headline. Market sizing, the practice of estimating how much revenue a market could yield, is where ambition meets arithmetic. Done well, it tells you whether a bet is worth making and where to point a team first. Done badly, it produces a confident number that nobody believes and no one can act on.

The quick version

  • TAM is the whole prize (everyone who could ever buy), SAM is the slice you can actually serve, and SOM is the part you can realistically win in the next few years.
  • Two ways to count: top-down (start with a big industry figure and cut it down) and bottom-up (start with one customer's spend and multiply up). Build both; trust the bottom-up.
  • The number that matters most is the smallest one. SOM is what your plan, your hiring, and your next two quarters are made of.
  • Size the need, not the product category. The biggest sizing errors come from defining the market as "what exists today."

The idea in depth

The three-letter ladder is straightforward once you stop treating it as jargon. TAM (Total Addressable Market) is the revenue you would earn if every possible buyer bought from you and only you, the whole pie. SAM (Serviceable Available Market) narrows that to the buyers you can actually reach given your product, your geography, and your business model. SOM (Serviceable Obtainable Market) narrows it again to what you can plausibly capture in a defined window, usually three to five years, against real competitors with a real go-to-market plan.

Think of it as three concentric rings. The outer ring inspires; the inner ring commits. Investors and boards like the outer ring because it tells them whether the opportunity is big enough to matter. Operators live in the inner ring, because that is the number a sales plan and a hiring plan are built from.

flowchart TB
    TAM(["TAM, everyone who could ever buy
the whole revenue opportunity"])
    SAM(["SAM, the slice you can serve
your product, geography, model"])
    SOM(["SOM, what you can win in 3–5 years
against real competitors"])
    TAM --> SAM --> SOM
The market-sizing ladder: each ring narrows the one above it from ambition toward commitment. Leaders Loop

Two ways to count, and why they should disagree

There are two honest routes to a market number. Top-down starts with a published figure, an industry analyst's estimate of total spend, and applies a chain of percentages to carve out your share. It is quick, and it is easy to inflate: every assumption is a number you chose. Bottom-up starts from the unit economics: how many potential customers exist, what each one pays per year, and how often. You multiply up. It is slower, harder to fake, and far more useful, because every input is something you can defend with a name, a price, and a count.

Practitioner writing in venture capital is consistent on which to trust. Forum Ventures, summarising what investors actually want, puts it plainly: bottom-up should be your primary approach when presenting to investors, because it is built from inputs you control rather than estimates you borrowed (Forum Ventures, "Market Sizing for Startups"). The two methods are most valuable when you run both and they disagree: a wide gap is a signal that one of your assumptions is wrong, and finding out which is the entire point.

So the move is: never present one number. Present a top-down estimate and a bottom-up estimate side by side, name the assumption that drives each, and explain the gap between them. A board trusts the leader who shows the seams, not the one with a single suspiciously round figure.

The trap: sizing the product instead of the need

The most expensive sizing mistake is defining the market as the thing that already exists. The classic warning is Theodore Levitt's, in his 1960 Harvard Business Review essay "Marketing Myopia": the American railroads declined, he argued, because they assumed they were in the railroad business when they were in fact in the transportation business, product-oriented where they should have been customer-oriented. Size the railroad and you measure a shrinking market; size transportation and you see the whole opportunity, including the cars and planes about to eat your lunch.

The most-cited modern version of this fight is real. In June 2014, NYU valuation professor Aswath Damodaran valued Uber at roughly $5.9 billion, building from a $100 billion global taxi-and-limousine market and an optimistic 10% share (Damodaran, "A Disruptive Cab Ride to Riches," 9 June 2014). Benchmark's Bill Gurley, an Uber investor, replied with "How to Miss By a Mile" (11 July 2014). His objection was not arithmetic but framing: Damodaran had assumed "the future will look quite like the past" by anchoring on the historical taxi market. Ask instead how big the market is for cheaper, better transport you can summon from your phone, including trips people never used to take and cars people might stop owning, and Gurley's range ran from $450 billion to $1.3 trillion.

A market-size number is a story about the future with a dollar sign in front of it. Argue the story first.

The honest limitation: Gurley's essay is also a cautionary tale about the other direction. Sizing the need rather than the product can justify almost any number, because you are now estimating a market that does not yet exist. Damodaran's narrower frame was wrong about Uber, but a narrow frame is not wrong by default, and a trillion-dollar TAM has been the warm-up act for plenty of companies that never found a single profitable customer. The discipline is to be ambitious about the need and ruthless about the obtainable slice. Which is why the bottom of the ladder matters more than the top.

Why SOM is the number that earns its keep

TAM tells you if the room is big enough to bother entering. SOM tells you whether you can survive once you are in it. The evidence for taking the bottom of the ladder seriously is sobering: CB Insights' March 2026 analysis of 431 venture-backed companies that shut down found that running out of cash was the top cited reason at around 70%, but the report is explicit that this is usually the symptom, with poor product-market fit (cited for about 43% of failures) among the root causes, and that two-thirds of those PMF failures were early-stage companies that "never found a market" (CB Insights, "Why Startups Fail," 5 March 2026). A market you sized at a billion and could not actually reach is, operationally, no market at all.

So treat SOM as a plan, not a fraction. Don't reach it by multiplying SAM by an invented "we'll get 1%." Build it the way you would build a sales forecast, named target segments, reachable channels, a realistic win rate against the alternatives customers have today. If you cannot describe the first hundred customers, you have not sized a SOM; you have decorated a SAM.

A worked example

All figures below are illustrative, invented to show the method, not researched market data.

Suppose you are launching a scheduling tool for independent physiotherapy clinics in Australia. Here is the same opportunity counted both ways.

Top-down. A market report (say) puts total Australian spend on practice-management software for allied health at $400m a year. You reason that physio is roughly a fifth of allied health, so your slice of that category is about $80m, your TAM for this segment. You can serve clinics on cloud billing systems, which you estimate is 60% of them, giving a SAM of about $48m. You judge you can win a tenth of those in five years: a SOM of roughly $4.8m. Quick, clean, and entirely dependent on three percentages you chose.

Bottom-up. Now count from the clinic. Industry directories list (say) about 4,000 independent physio clinics in Australia, that is your reachable population. Your plan charges $250 per clinic per month, or $3,000 a year. Multiply: 4,000 × $3,000 = $12m. That is your bottom-up SAM, built from a count you can verify and a price you set. To get SOM, you don't take a percentage, you build a plan: your two-person sales team can onboard 15 clinics a month, so in three years you reach roughly 400 clinics, or about $1.2m in annual recurring revenue.

flowchart LR
    A(["Reachable clinics
4,000"]) --> B(["× annual price
$3,000"])
    B --> C(["Bottom-up SAM
$12m / year"])
    C --> D(["Onboarding capacity
~400 clinics in 3 yrs"])
    D --> E(["SOM
~$1.2m ARR"])
Bottom-up sizing builds from one customer's spend and your real capacity to reach them. Leaders Loop

Notice the gap. Top-down handed you a $48m SAM; bottom-up handed you $12m. That fourfold disagreement is the most useful output of the whole exercise, it forces the question: is the top-down category figure too broad, or the bottom-up clinic count too narrow? Chasing that single question down produces a number you can stand behind in a board meeting. The arithmetic was never the point. The argument was.

Frequently asked questions

Should I lead with TAM or SOM in a pitch?

Lead with TAM to establish the prize is big enough to be worth your time, then spend most of the slide on the bottom-up SOM, because that is what proves you understand how you will earn it. A huge TAM with a hand-waved SOM reads as ambition without a plan; experienced investors discount it on sight.

How big does a TAM need to be?

It depends on who you are sizing it for. Venture investors typically want a TAM large enough that even a modest share returns their fund, often $1 billion or more. A bootstrapped or family business has no such requirement; a $20m market you can dominate profitably beats a billion-dollar TAM you lose money chasing. Size to your funding model, not a generic threshold.

Where do I get the numbers without paying for an analyst report?

For bottom-up, you often don't need one: industry directories, professional-body membership counts, government statistics, public filings, and your own pricing get you most of the way. Triangulate two or three independent sources for any load-bearing figure, a sized market is only as trustworthy as its citations.

Isn't "we'll capture 1% of a huge market" a reasonable SOM?

No, it is the single biggest tell that a market hasn't really been sized. One percent is a number chosen because it sounds humble, not because anything in the plan produces it. A defensible SOM is built from named segments, reachable channels, and a win rate against the alternatives customers use today. If you can't show the mechanism, the percentage is decoration.

How often should we re-size the market?

Re-size whenever a core assumption changes: a new segment, a price change, entry into a new geography, or a competitor reshaping what customers expect. Treat the model as a living spreadsheet, not a slide you build once for fundraising. The act of updating it is often where you notice the market has moved before your revenue has.

Related in the Toolkit

Where to go next