TAM, SAM and SOM explained: definitions, methods and a worked example

What do TAM, SAM and SOM actually measure?

Each layer answers a different question, and each is narrower than the one above it. The total addressable market describes the whole category in a defined geography and year. The serviceable available market removes the customers your product, price position or delivery area cannot reach. The serviceable obtainable market is a forecast: the slice of SAM you expect to win once competitors, budget and sales capacity are taken into account.

All three can be expressed in value (euros a year) or in volume (units, customers, orders). Pick one and keep it consistent across the layers, because a TAM in euros and a SOM in customers cannot be compared on the same slide.

Top-down or bottom-up: which route should you use?

The top-down route starts from an official total, usually population or the number of firms, and narrows it with percentages until it reaches the target group. It is fast and auditable, but it is only as good as the match between statistical categories and your product.

The bottom-up route starts from what can be counted directly: customers reached by your channels, conversion rates from test campaigns, order values from your own sales data. It sits closer to real behaviour but depends on assumptions about how far early results scale.

A defensible estimate uses both. When the two routes land far apart, the gap tells you which assumption is weakest. A full market sizing study documents that reconciliation step by step, so a sceptical investor can follow it.

Worked example: a refill-cleaning subscription in the Netherlands

Example (hypothetical): a start-up sells plant-based cleaning concentrates online, delivered as refills on a subscription. It wants to size the Dutch market before a funding round. The market is defined as household cleaning products bought online by consumers in the Netherlands, per year.

Two inputs come from official statistics. The rest are labelled assumptions that a survey would later measure.

Step 2 contains a deliberate shortcut. Eurostat's online-shopping rate refers to people aged 16–74, while the World Bank figure covers every age. Multiplying the two overstates the base, which is why the result is marked as an upper bound. In a live model, replace the total with the 16–74 population from Eurostat's demographic tables before anyone quotes the figure.

The example also mixes years: population and online shopping are 2025 values, while the price levels used in the next section are 2024, the latest Eurostat publishes. Mismatches like this are normal; what matters is that they are written down next to the number.

How does the same model look in Bulgaria?

A common shortcut is to scale a TAM by population. Bulgaria had 6,433,302 inhabitants in 2025 according to the World Bank; the Netherlands has about 2.8 times as many. Dividing the Dutch TAM by 2.8 would be wrong twice over.

First, online shopping is far less widespread. Eurostat reports that 51.2% of individuals in Bulgaria bought online in 2025, the lowest share in the EU, against 94.2% in the Netherlands and 73.6% for the EU as a whole.

Second, spending a year does not travel between countries at face value. Eurostat's 2024 price level index for household consumption is 56.9 in Bulgaria and 121.0 in the Netherlands (EU-27 = 100). Real consumption per head also differs: in 2024, actual individual consumption came to 19,300 purchasing power standards (PPS) in Bulgaria against 31,400 in the Netherlands. Combining price level and volume puts nominal household consumption per head in Bulgaria at roughly 29% of the Dutch level.

With 3.29 million online shoppers and the same 30% category assumption, the start-up gets about 0.99 million potential buyers in Bulgaria. The spend assumption then decides the outcome:

The price-adjusted Bulgarian TAM is about one-eleventh of the Dutch one, although the Dutch population is only 2.8 times larger. The honest presentation is a range, roughly €14m to €23m, with a note that Assumption A was copied from the Netherlands and has not yet been measured locally. Our overview of Bulgaria's consumer market in numbers covers the other indicators a foreign entrant should check, and a market entry study tests whether the opportunity survives contact with distribution and competition.

Checking the SOM from the bottom up

A SOM built only by taking a percentage of SAM is a wish. The bottom-up check asks what the company's own resources can deliver.

In the Dutch example, suppose the media plan reaches 1.2 million online shoppers a year through paid social and two partner web shops, and test campaigns converted 3% of reached shoppers into subscribers. That gives 36,000 subscribers, or €1.8 million a year at €50 each. The top-down SOM was €2.6 million, which equals about 51,100 subscribers.

The two routes differ by roughly 30%. That is close enough to present as a range of €1.8–2.6 million, and the gap points at a concrete question for the founders: does the five per cent share assume more reach than the budget buys? Either the media plan grows, or the SOM comes down.

Five mistakes that make investors distrust a TAM slide

• Borrowing a headline figure. A global report's "cleaning products market" includes industrial cleaners and offline retail. Unless the category definition matches yours, it is not your TAM.
• Mixing bases. Shares calculated for people aged 16–74, or for enterprises with ten or more employees, must be applied to that same base, not to the whole population or every registered firm.
• Scaling by population. As the Bulgarian case shows, online adoption and price levels can shrink a market far faster than headcount does.
• One number instead of a range. A single point implies a precision the inputs do not have. Show low, central and high values and say which assumption drives the spread.
• A SOM with no capacity behind it. If the sales team, budget or warehouse cannot serve the forecast customers, the share is not obtainable, whatever the percentage says.

Where do the inputs come from?

Population, online behaviour, price levels and business counts are published free of charge by Eurostat, the World Bank and national statistics offices; our guide to free market data sources for Europe lists where to find each one, and the latest figures are charted in our consumer data section.

Category penetration and spend per buyer, Assumptions A and B above, rarely appear in official tables. They are measured with a short online survey of the target group. Our sample size guide explains how many respondents give a usable margin of error for those two inputs.

To try your own numbers, the TAM/SAM/SOM calculator runs both routes side by side. If the estimate will go in front of a board or an investor, request a proposal for an independent sizing model with sources and sensitivity analysis.