glossary · strategy
TAM, SAM, SOM
Three concentric market tiers (Total Addressable, Serviceable Addressable, Serviceable Obtainable) that a PM uses to justify roadmap investment, frame market entry, and bound realistic capture.
TAM, SAM, and SOM are three concentric market tiers: Total Addressable Market (every dollar spent on this problem globally), Serviceable Addressable Market (the subset your product can actually reach and serve today), and Serviceable Obtainable Market (the share you can realistically capture given your sales motion, CAC, and distribution). PMs use them to justify roadmap investment and market entry decisions, not just to raise money. When an interviewer asks you to size a market, they are checking whether you understand the difference between a market existing and a market being capturable.
The canonical real-world example: Airbnb’s 2009 pitch deck sized TAM at 1.9 billion total worldwide trips, SAM at 532 million budget or online trips, and SOM at 10.6 million trips in year one. Each tier was bounded by a different, named constraint. That precision is what interviewers want to see.
Why PMs use these differently than VCs
A VC uses TAM to argue that a big enough prize exists to justify the bet. A PM uses TAM as a ceiling and SOM as the planning unit. The question a PM is answering is: “Is this problem large enough to warrant the team, roadmap space, and investment, and can we capture enough of it to sustain the business?” That requires a defensible SOM, not just a plausible TAM. Conflating the two is the most common interview failure.
Top-down vs bottom-up: use both, lead with bottom-up
Top-down starts with a published market total and carves down. Example: 50 million knowledge workers, $164 in annual software spend per worker, gives a TAM around $8.2 billion for B2B project management software (cross-validated against Gartner’s collaboration software figure). Useful as a ceiling check. Its weakness is that industry reports now lag AI-driven market shifts by 12 to 18 months.
Bottom-up builds from unit economics: count customers you can actually sign, multiply by ACV or ARPU. A Year 1 SOM worked to this level of specificity: 500 customers at $2,000 ACV equals $1 million revenue. Bottom-up is roughly twice as credible with interviewers in 2026 than top-down alone, because Carta’s 2025 benchmarks found that pitch decks with credible bottom-up SOMs closed Series A rounds 40% faster than those relying on top-down claims.
The self-check: when bottom-up and top-down align within about 15%, your assumptions are likely solid. State that explicitly in the interview: “my bottom-up gives $X, my top-down ceiling is $Y, and they’re in the same order of magnitude, which gives me confidence in the range.”
The 2026 reframe: “serviceable” no longer means technically feasible
Before 2025, SAM was partly constrained by what you could build. Technical feasibility filtered who you could serve. In 2026, feasibility is close to free: you can ship an AI-powered version of almost any workflow in weeks. So the meaningful constraint on SAM is no longer “can we build it for this segment?” It is “will this segment actually change how they work, trust the product, and pay for it?” SAM is now bounded by distribution reach, regulatory clearance, and behavioral adoption, not engineering capacity.
This matters in interviews. Candidates who define SAM by technical reach look behind the curve. Candidates who define SAM by “who will actually adopt and pay given current market conditions” show 2026 PM thinking. The same reframe applies to SOM: the limiting factor is not your team’s ability to build, it is your sales motion, trust signals, and CAC relative to LTV.
One practical consequence: TAM recalculation that used to be an annual exercise now needs to happen closer to quarterly. AI-driven market shifts can expand or compress SAM within a single product cycle.
What interviewers penalize
The 1% fallacy: claiming a $10 billion TAM and saying “we just need 1% to hit $100 million.” Interviewers have heard this hundreds of times. It signals market-existence awareness without market-capture thinking. 60% of pitch decks cite inflated TAM figures; interviewers now explicitly penalize this move.
Collapsing SAM and TAM: saying “our TAM is $8 billion and our SAM is $6 billion because we focus on English-speaking markets” is a rounding error, not a real filter. A meaningful SAM constraint changes the number by an order of magnitude, not 25%. The failure mode is picking a filter that sounds specific but removes almost nothing.
Indefensible SOM: getting from SAM to SOM by saying “we’ll capture X% because our product is better” is not a SOM, it is a wish. Interviewers want named levers: sales team headcount and ramp time, CAC relative to LTV, a named distribution partnership, regulatory clearance others lack, or a trust signal (named reference customer, security certification) that lowers buyer hesitation in a market where buyers evaluate three to five alternatives before committing.
Citing the total AI market as a TAM: the AI software market was $244 billion in 2025. That number is useless for any specific product. Naming it as your TAM is the 2026 version of the 1% fallacy.
Over-filtering SAM into irrelevance: drilling SAM so narrow that SOM becomes trivially small, with no justification for why the broader market is unreachable. If your SAM filter removes 99% of TAM with one sentence, the interviewer will ask what makes the rest genuinely out of scope.
Strong vs weak interview answer
weak
"Our TAM is the $10B project management software market. Our SAM is $6B because we're focused on English-speaking markets. If we capture just 1% we hit $100M." Three problems: SAM barely narrows TAM (not a real constraint), SOM is derived by assertion rather than unit economics, and no assumption is named or flagged. The interviewer stops taking the answer seriously at "1%."
strong
"I'll lead bottom-up for SOM credibility and cross-check top-down. TAM: 50M knowledge workers who use project software at roughly $164 average annual spend gives about $8.2B. That aligns with Gartner's collaboration software figure, so I'm confident in the ceiling. SAM: I'm filtering to mid-market teams of 10 to 200 employees in three English-speaking markets where we have distribution coverage and no regulatory blockers, and where the segment has demonstrated willingness to replace their current workflow. That cuts to roughly $1.2B. I'm not filtering by technical feasibility, since we can build for any of these segments; I'm filtering by who will actually adopt and pay. SOM year one: a two-person outbound team reaching 5,000 accounts, converting at 10%, at $2,400 ACV, equals about $1.2M. The assumption that drives the most variance is conversion rate, so I'd run a 20-account pilot before committing to the model. My top-down ceiling and bottom-up SOM are consistent in direction. That gives me confidence the range is right, not that the exact number is."
Connecting sizing to the rest of PM work
In a strategy deck or roadmap conversation, TAM justifies that the problem is worth solving at all. SAM sets the scope of market entry: which customers, which geographies, which channels. SOM grounds the first planning cycle in something defensible. When the three tiers are consistent with each other and each is bounded by a named, real constraint, the sizing exercise earns trust. When any tier is bounded by a percentage applied to the tier above it, it loses trust.
For the broader 2026 context on why feasibility is no longer the constraint on SAM, see feasibility is free and proving viability. For how SOM connects to launch sequencing, see go-to-market.