Market sizing
Market Sizing for Founders: TAM, SAM and SOM Without the Guesswork
Three letters, one honest number, and the two ways to get there.
If you searched this because you're prepping for a McKinsey case interview, this isn't your page — go find the interview-prep guides, they're good at that job. If you searched it because you're trying to figure out whether your idea is worth building, keep reading. Those are different questions, and most of what currently ranks for “TAM SAM SOM” answers the first one, not the second.
That's not a knock on the case-interview content — it's good at what it's for. It's just built for someone estimating the number of pianos tuned in Chicago, not someone deciding whether to spend the next six months building something. The method is similar. The stakes and the honesty required are not.
What TAM, SAM and SOM actually mean
- TAM (Total Addressable Market) — everyone who could theoretically use what you're building, worldwide, no constraints. It's a ceiling, not a forecast, and it's almost always the biggest, least useful number in the exercise.
- SAM (Serviceable Available Market) — the slice of TAM you could actually reach given your product, your geography, and the channels you'd actually use to sell. This is where most founders should spend their attention.
- SOM (Serviceable Obtainable Market) — the slice of SAM you could realistically capture in a defined window, given competition and your own capacity to sell and deliver.
The number that should actually drive a go/no-go decision is SOM. TAM is the number that looks good on a pitch deck slide and tells you almost nothing about whether you should spend the next six months building — a $50B TAM is meaningless if your realistic first-year SOM is $40,000 and your costs are $60,000.
The two ways to calculate each one — and why they disagree
Top-down
Top-down starts from a big public number — industry revenue, a market research report, a government statistic — and slices it down by assumption. “The lawn care industry is worth $80B, and if we capture even 0.1% of that, we're a $80M business.” It's fast, and it's the version investors are trained to distrust, because every assumption in the chain compounds. A 20% error at each of three steps isn't a 20% error in the final number — it can be a 70%+ error, and nobody flags it because the arithmetic looks clean.
Bottom-up
Bottom-up starts from a real, checkable count: how many people search for this, how many businesses exist in the category, what the average transaction is worth, multiplied out. It's slower to build and it's the version that survives a follow-up question, because every number in the chain has a source you can point to — a search volume figure, a public business count, a price you can verify.
The move most founders skip
A worked example
Say the idea is a subscription lawn-care service for a mid-size metro.
| Layer | Method | What it captures |
|---|---|---|
| TAM | Top-down | US residential lawn care is roughly a $60–100B annual category, depending on the estimate used. |
| SAM | Bottom-up | Households in your metro that own a home with a yard and don't mow it themselves — metro household count × homeownership rate × an estimated “pays someone to do this” rate, informed by category search volume in that city. |
| SOM | Bottom-up | Of that SAM, the share reachable in year one given crew capacity — a handful of zip codes, not the whole metro. |
Notice what happened: the number that actually tells you whether to build this is the smallest one in the table, and it's the one that almost never makes it onto the cover slide.
What “size the market” doesn't answer
A market can be enormous and still be the wrong one for you — because competition already owns it, because the margins are thin, or because the demand is seasonal in a way a single TAM figure hides completely. Market size answers is there room. It doesn't answer can I get into the room, which is a competition question (see our competitor analysis guide), and it doesn't answer will people actually pay, which is a question about intent, not size.
The honest version of this exercise
Most market-sizing content — including a lot of what currently ranks for these terms — treats the output as a single confident number. It shouldn't be. A defensible market size is a range, built from sources you can name, with the assumptions stated next to the number instead of hidden behind it. If a market-sizing exercise hands you one clean figure and no visible math, that's a sales document, not an analysis — and it's worth being suspicious of anyone, including yourself, who presents it that way.
That's the standard we hold our own reports to: every market figure comes with its source and its confidence level stated plainly, never presented as more precise than the underlying data actually supports. When a number can't be measured well, the report says so instead of guessing and hiding the guess.
Signal & Demand · $129 per idea
Get the bottom-up number pulled from real search and competitive data — for your specific idea, not a generic category.