How to Calculate TAM, SAM and SOM Without Inflation

Calculate TAM by multiplying the number of customers who could genuinely buy your defined product by realistic annual revenue per customer. Derive SAM by removing customers your current product, geography, pricing or operating model cannot serve, then set SOM at the smaller of the demand you can win and the capacity you can deliver within a stated period.

The defensible method is bottom-up: keep customer counts, prices, conversion assumptions and capacity limits as separate inputs, with a source and date attached to each one. Use top-down industry figures only as a reasonableness check, not as a substitute for customer-level evidence.

What TAM, SAM and SOM actually measure

TAM, SAM and SOM answer three different questions, so they should not be three arbitrary percentages of the same large industry forecast. All three must use the same product definition, revenue basis, currency, territory logic and period.

Total addressable market: the annual revenue available if every qualifying customer bought the defined product.
Serviceable addressable market: the part of TAM that the current product and business model can serve.
Serviceable obtainable market: the part of SAM the company could credibly win during a stated planning horizon.

The Amazon Ads market-sizing guide calculates TAM as the number of potential customers multiplied by average revenue per customer, then narrows the opportunity for SAM and SOM. The nesting is useful, but your filters should be auditable counts or rules rather than unexplained percentages.

Start with one precise market sentence: “Organizations of type X, in territory Y, with need Z, buying product A at annual revenue B.” If your TAM describes all business software while your product serves independent dental clinics, the definition is already too broad.

Build a calculation sheet that exposes every assumption

A market-sizing workbook separates three calculation methods and records the source date, unit and confidence of every input.

A useful workbook separates evidence from arithmetic. Create bottom-up, top-down and value-theory tabs plus a source register; the bottom-up tab should remain the primary model used in the business plan or pitch deck.

For every input, record the variable, value, unit, geography, segment, source URL or internal dataset, source date, retrieval date, calculation note, owner and confidence level. Do not paste a final market number into the model without retaining the counts and revenue assumptions that produced it.

Define one customer unit: account, household, venue, professional or transaction-producing buyer.
Set a common period, normally one year.
Separate segments when their prices or buying frequencies differ.
Calculate TAM from the unfiltered qualifying population.
Apply explicit serviceability rules to produce SAM.
Calculate a demand-based SOM and an operating-capacity ceiling.
Use the lower credible result as SOM.
Compare it with top-down and value-theory estimates, then investigate material differences.

This structure also makes revision straightforward. When pricing, geographic coverage or sales capacity changes, you can update one dated input instead of rebuilding an opaque headline figure.

Calculate TAM from customer counts and annual revenue

The core formula is TAM = qualifying potential customers × annual revenue per customer. For B2B businesses, the University of Regina worksheet states the equivalent formula as the number of accounts in the market multiplied by annual contract value.

Count only entities that share the problem the defined product solves. A company directory is not automatically a customer list: remove duplicates, inactive organizations and records outside the necessary size, industry or technical conditions before using the count.

Annual revenue per customer must match the business model. For subscriptions, use annual recurring revenue per account or monthly revenue multiplied by 12, with justified plan mixes separated. For transactional businesses, use annual purchase frequency multiplied by company revenue per purchase.

For marketplaces, keep gross merchandise value distinct from platform revenue. If the company earns a take rate, calculate revenue from transaction value multiplied by that rate and account for rebates or incentives that reduce the company’s revenue; do not present all spending conducted through the marketplace as its market revenue.

Do not insert an untested future premium price because it enlarges TAM. If observed pricing is unavailable, use a documented current price, comparable offers or customer research, and show low, base and high scenarios instead of concealing uncertainty in one average.

Turn TAM into SAM with serviceability filters

SAM includes customers you can serve with the product and operating model being evaluated, not everyone you might serve after several hypothetical expansions. Translate every boundary into a countable inclusion rule.

Geography: countries, states or travel radiuses where sales and delivery are available.
Product fit: required integrations, languages, use cases, organization sizes or customer capabilities.
Commercial fit: customers able to buy at the stated price and through the available contract structure.
Legal and operational fit: sectors or territories supported by the current compliance and fulfillment model.
Buying unit: accounts containing the relevant team, budget owner or decision-maker.

Calculate segments independently when their revenue differs: SAM is the sum of serviceable customers in each segment multiplied by that segment’s annual revenue. Avoid applying successive percentages to TAM unless each percentage comes from a dataset with the same territory, population and denominator.

Keep serviceability and reachability separate. An account may fit the offer even if you have not identified a contact, but a near-term outbound plan cannot assume access to decision-makers whom its chosen channels cannot reach. Preserve total serviceable accounts and currently reachable accounts as different rows.

Calculate SOM from demand and capacity

A demand forecast is capped by the startup’s sales and onboarding capacity to produce a defensible SOM.

SOM needs both a time horizon and an acquisition mechanism. “We can capture 2%” is not a calculation until the model shows how many prospects enter the funnel, how they convert, how long deals take and whether the company can onboard and support the resulting customers.

The Sopro market-sizing report distinguishes top-down, bottom-up and value-theory approaches and says SOM should reflect resources, reach, competition, sales capacity and market position. Convert those constraints into two numerical ceilings:

Demand-based obtainable revenue = reachable prospects × qualified-opportunity rate × close rate × annual revenue per new customer.
Capacity-based obtainable revenue = maximum customers that sales, onboarding, inventory and service teams can support × annual revenue per customer.

Your base SOM should not exceed the lower ceiling. Reconcile it with existing customers, expected losses and contract timing as well. If the model predicts 600 wins but implementation can onboard only 20 customers per month, a one-year forecast cannot treat all 600 as active for the entire year.

Do not model competition by dividing SAM equally among vendors. Ground expected wins in channel access, sales-cycle length, replacement opportunities and observed conversion data. If the startup has no operating history, label funnel assumptions as hypotheses and present a conservative range.

Apply the method to three business models

SaaS contracts, marketplace take rates and local-service capacity lead to different bottom-up market calculations.

SaaS example

As a hypothetical example, suppose a vertical SaaS company identifies 18,000 qualifying accounts at $2,400 in annual contract value. TAM is $43.2 million. If 7,500 accounts are in supported countries and use compatible workflows, SAM is $18 million.

Assume the startup can reach 3,000 accounts over the next 12 months, create qualified opportunities from 4% of them and close 20% of those opportunities. The result is 24 customers and $57,600 in annualized new-contract value. If onboarding can support 40 new accounts, demand is the binding constraint; reported first-year revenue would still need adjustment for contract start dates and losses.

Marketplace example

Consider a hypothetical marketplace with 60,000 relevant buyers, four purchases per buyer annually, an average transaction value of $75 and a 12% take rate. Annual gross transaction value is $18 million, while platform TAM revenue is $2.16 million. Presenting $18 million as company revenue would overstate the opportunity.

If the launch territory contains 15,000 qualifying buyers, SAM revenue is $540,000 using the same frequency, transaction value and take rate. SOM must then reflect how many active buyers and suppliers the company can acquire while sustaining completed transactions; registrations alone are not a revenue-producing unit.

Local-service example

For a hypothetical home-maintenance company, suppose 25,000 qualifying households would each generate $300 in annual company revenue. TAM is $7.5 million. If the current travel radius contains 6,000 of those households, SAM is $1.8 million.

Now test fulfillment. Five technicians completing three paid visits per day over 220 working days provide capacity for 3,300 visits. At $150 of company revenue per visit, the annual capacity ceiling is $495,000. Demand may support more, but SOM cannot exceed that ceiling without more technicians, higher productivity, additional working days or a different price.

Use top-down and value theory as cross-checks

A top-down check begins with a published category total and narrows it to the same geography, segment and revenue definition used by the bottom-up model. It can reveal a unit error or implausible customer count, but broad industry revenue may include products, channels and buyers outside the startup’s market definition.

Value theory estimates potential pricing from measurable customer benefit. As a hypothetical illustration, if a product creates $10,000 in verified annual benefit and research supports a price equal to 10% of that value, the implied annual price is $1,000. This can help assess a new category without reliable historical spending, but created value does not automatically become willingness to pay or vendor revenue.

Do not force the methods to match. Reconcile whether a gap comes from category boundaries, customer counts, pricing, purchase frequency or confusion between customer spending and the company’s recognized revenue. A large unexplained gap calls for more research, not selection of the largest result.

Remove the assumptions that inflate market size

Before publishing the model, challenge the inputs most likely to enlarge it. These checks catch common structural errors:

Do TAM, SAM and SOM use the same currency, annual period and revenue definition?
Are customers counted once, with buyers kept distinct from users, locations and transactions?
Does annual revenue reflect realistic plan mix, discounts, buying frequency and take rate?
Are SAM filters based on current serviceability rather than a future roadmap?
Does SOM name a time horizon and connect to a measurable acquisition funnel?
Can sales and delivery capacity support the customer count?
Are customer losses, contract start dates and recurring versus one-time revenue treated consistently?
Does every material input have a source date and an owner responsible for updating it?

Run sensitivity analysis on the weakest inputs. If changing an assumed conversion rate from 2% to 4% doubles SOM, show both outcomes and state what evidence is needed to narrow the range. Precise spreadsheet formatting does not reduce uncertainty in unsupported assumptions.

Prepare the figures for a business plan or pitch deck

Present one slide with the market definition, formulas, base-case figures and three to five decisive assumptions. Keep the complete calculation sheet available so a reviewer can trace each headline number to its inputs and evidence.

Label TAM as the theoretical opportunity under the defined offer, SAM as the portion served by the current model and SOM as a time-bound execution case. Connect SOM to reachable accounts, qualified opportunities, closed customers, onboarding or fulfillment capacity, contract timing and annualized revenue.

The practical next step is to populate the bottom-up tab with current customer records and pricing. Once the base case survives reachability and capacity checks, use the other tabs to challenge it, record unresolved uncertainty and assign dates for updating the inputs when the product, territory or go-to-market model changes.

How to Calculate TAM, SAM and SOM Without Inflating Your Startup’s Market