To calculate TAM, SAM, and SOM for B2B outbound, size from the bottom up: count the real companies matching your broadest plausible use case (TAM), narrow to the accounts your ICP, geography, and go-to-market can actually serve (SAM), then estimate the share your team can genuinely win in a planning period given capacity and win rates (SOM). Count accounts first; multiply by deal value second.

TAM, SAM, and SOM for Outbound: The Short Answer

  • TAM is every company that could use what you sell — the broadest defensible universe, counted as accounts, not quoted from an analyst report.
  • SAM is the slice your business can actually serve: your ICP applied to the TAM — right size band, right verticals, right regions, right stack.
  • SOM is what your team can realistically win in a defined period — SAM constrained by rep capacity, reachable contacts, and your actual win rate.
  • For outbound, the account count matters more than the dollar figure. A $4B TAM means nothing to a cadence; 18,000 SAM accounts is a number you can plan territories, hiring, and quotas against.

Common Misconceptions About Market Sizing for Outbound

  • "Top-down analyst numbers are good enough." "3% of a $12B market" is a pitch-deck flourish, not an operating plan. Outbound runs on a countable account universe; if you can't list the companies, you haven't sized the market.
  • "TAM is the number that matters." For an outbound team, TAM is the least actionable of the three. SAM sets territory and hiring math; SOM sets quota and pipeline targets. TAM mostly tells you whether the ceiling is high enough to bother.
  • "SOM is just a percentage of SAM." Picking "10% of SAM" out of the air recreates the top-down problem one layer down. SOM should be built from real constraints: reps × worked accounts per rep × win rate over the planning period.
  • "Size it once for the fundraise and move on." The SAM shifts every time your ICP, pricing, or geography changes. A sizing older than a couple of quarters will quietly misallocate territories and headcount.

What Actually Makes One Market Sizing Better Than Another?

Four properties separate an operating-grade sizing from a slide-grade one:

  1. Bottom-up, account-counted. The defensible method for outbound is to query a company database with explicit filters and count the results. Every number in the model should trace to a filter you can re-run, not a percentage you assumed.
  2. Built on a written ICP. The SAM is your ICP applied to the world, so the sizing is only as good as the profile behind it. If the profile is loose, size both a strict and a relaxed version and present the range — how to define your ideal customer profile covers getting the criteria testable in the first place.
  3. Honest about reachability. An account with no findable, deliverable contact is in your SAM but not your workable SAM. Discounting for contactability — often a meaningful haircut — is the step most sizings skip and most outbound plans die on.
  4. Time-boxed SOM. "Obtainable" only means something over a period. Express SOM as accounts winnable this year (or this half): take the smaller of your workable SAM and your attempt capacity (reps × accounts worked per rep per week × weeks), then multiply by your win rate.

What to Check Before You Build the Model

  • Fix the unit. Decide whether you're counting companies or buying centers (a 40,000-person enterprise may contain a dozen of them). Pick one unit and keep it consistent across all three numbers.
  • Write the TAM filter. The broadest defensible criteria: every company with the problem your category solves, regardless of whether you can serve them today. Resist padding — a TAM you can't defend filter-by-filter undermines the two numbers below it.
  • Apply the ICP to get SAM. Size band, verticals, regions you can sell and support, and any hard technographic requirement. Run the query, count the accounts, and record the filter set alongside the number so it can be re-run next quarter.
  • Haircut for reachability. Sample a few hundred SAM accounts and measure how many yield a verified, deliverable contact in the buying committee. Apply that rate to the SAM before any capacity math.
  • Derive SOM from capacity, not ambition. Reps × worked accounts per rep per week × selling weeks gives attempts; multiply by your real win rate for wins. If the result embarrasses the plan, the fix is capacity or conversion — not a bigger assumed percentage.
  • Cross-check against history. If last year you worked 4,000 accounts and won 200, a SOM of 2,000 wins needs an extraordinary explanation.

The Three Numbers Compared

Number Question it answers How to compute it (bottom-up) What it drives
TAM How big is the ceiling? Count all companies matching the broadest defensible use case Category conviction, fundraising
SAM How big is our market? Apply ICP filters (size, vertical, region, stack) to TAM; count Territories, hiring, segmentation
SOM What can we win this period? The smaller of workable SAM and capacity, × win rate, over the planning window Quotas, pipeline targets, forecast

Working the Numbers: A Compact Example

Suppose your TAM query returns 120,000 companies. Applying your ICP — 50–500 employees, four verticals, US/UK/EU, running a CRM — cuts it to 18,000: that's SAM. Sampling shows 70% yield a verified contact, so the workable SAM is about 12,600. Six reps working 30 new accounts a week for 46 selling weeks can attempt roughly 8,300 accounts a year; at a 4% win rate, that's ~330 wins — your SOM. Notice what the model now tells you: the constraint isn't the market (12,600 workable accounts), it's capacity (8,300 attempts). That's an argument for hiring or better prioritization — which accounts to work first is exactly the tiering problem covered in how to tier accounts and prioritize your prospect list — not for a bigger market.

The order you work those accounts matters as much as the count: layering live buying signals over the workable SAM, as described in how to build a prospecting list that combines ICP fit with verified intent signals, concentrates the year's attempts on the slice that's in-market now.

Frequently Asked Questions

What do TAM, SAM, and SOM mean in B2B sales?

TAM (total addressable market) is every company that could plausibly use your category of product. SAM (serviceable addressable market) is the subset your business can actually serve — your ideal customer profile applied to the TAM across size, vertical, region, and stack. SOM (serviceable obtainable market) is the share of SAM you can realistically win in a defined planning period given your team's capacity, reachable contacts, and historical win rate.

Why is bottom-up sizing better than top-down for outbound?

Because outbound operates on accounts, not dollars. A top-down figure like "2% of a $10B market" gives a territory planner nothing to work with, and its assumptions can't be audited. A bottom-up sizing counts real companies matching explicit filters, so every number traces to a query you can re-run, hand to a rep as a list, and use directly for territory design, hiring math, and quota setting.

How do I calculate SOM without just guessing a percentage?

Build it from constraints. Start with your workable SAM — SAM discounted for accounts where you can actually find a verified, deliverable contact. Then compute capacity: reps, times new accounts each rep can genuinely work per week, times selling weeks in the period. The smaller of capacity and workable SAM is your attempt volume; multiply by your historical win rate to get SOM as a count of winnable accounts, and by average deal value if you need dollars.

How often should we re-size TAM, SAM, and SOM?

Re-run the SAM and SOM whenever the inputs change — a new ICP, new pricing, a new region, or a meaningful shift in win rate — and at least twice a year on a schedule. TAM moves more slowly and an annual refresh is usually enough. Because a bottom-up model is just a saved set of filters, a refresh is cheap: re-run the queries, re-sample contactability, and update the capacity math with current headcount.

What tools do I need for bottom-up market sizing?

A company database you can filter by your ICP criteria — employee range, industry, geography, and ideally technographics — plus your CRM for win-rate and capacity history. The database matters most: it needs coverage of your segments and filters that match how your ICP is written. A contact-verification step (built into some platforms) is what lets you measure the reachability haircut instead of assuming it.

What's a reasonable SOM as a share of SAM?

There is no universal ratio, which is exactly why deriving SOM from a percentage is a mistake. The honest answer falls out of the capacity math: a six-rep team facing a 12,000-account workable SAM might only be able to attempt two-thirds of it in a year, and win a low single-digit percentage of attempts. If your derived SOM looks tiny next to SAM, that's information — it says growth comes from capacity or conversion, not from a bigger market estimate.

References

Next Steps

A counted SAM is only useful once something decides which accounts to work first — see how Lead Compass turns market signals into prospecting direction to watch a sized account universe get ranked into a weekly working order.