How to Size a B2B Software Market With TAM, SAM and SOM Without Paid Reports

To size a B2B software market without paid reports, count the companies that have the problem, multiply by what they would realistically pay per year, then narrow that total to the segment you can serve (SAM) and the share you can win in a few years (SOM). Build it bottom up, from public data.

Paid analyst reports have their place, but they are expensive, often a year old by the time you read them, and rarely cut the market the way your product does. A bottom up estimate built from public sources is cheaper, more transparent and easier to defend, because every number in it has a visible origin. This guide walks through the method step by step, with a worked example you can copy.

What TAM, SAM and SOM actually mean

The three terms describe nested circles. Getting the definitions right matters, because most sizing mistakes come from mixing them up.

• TAM, total addressable market. The annual revenue available if every company that has the problem bought a solution like yours. It is a ceiling, not a forecast.
• SAM, serviceable addressable market. The part of TAM your product can actually serve today, given its geography, language, integrations, company size focus and compliance coverage.
• SOM, serviceable obtainable market. The part of SAM you can realistically win in a defined period, usually three to five years, given competition, sales capacity and brand.

Investors look at all three, but they read them differently. TAM tells them whether the opportunity is big enough to matter. SAM tells them whether you understand your own product's limits. SOM tells them whether your plan is credible.

Top down versus bottom up

There are two ways to reach a number.

Use bottom up as the main estimate. If you can find a credible top down figure, use it as a sanity check. When the two land within the same order of magnitude, your estimate is probably reasonable. When they are ten times apart, one of your assumptions is wrong and it is worth finding out which.

Step one, define the buyer precisely

Bottom up sizing starts with a sentence like "companies in the United States with 50 to 1,000 employees that run payroll in house". Every word changes the count. Specify:

• Geography. Country or region.
• Company size. Employee band or revenue band, whichever drives purchase behavior in your category.
• Industry. All industries, or a vertical.
• Trigger. The condition that creates the need, such as "has a dedicated finance team" or "sells in more than one currency".

Step two, count the buyers from public data

Government statistics are the best free source for company counts. Most national statistics offices publish the number of enterprises by employee size class, and many publish it by industry as well. Useful public starting points include the US Census Bureau's County Business Patterns and Statistics of US Businesses, Eurostat's structural business statistics for EU countries, and the Office for National Statistics business population data in the UK.

From there, apply filters you can justify:

• If only companies above a size threshold need the product, take only those size classes.
• If the trigger applies to a fraction of companies, estimate that fraction from a public source, for example the share of companies in an industry that export, which trade statistics often publish.
• If you cannot find a source for a filter, state the assumption openly and show what happens to the result when you change it.

Step three, estimate realistic annual spend

Annual contract value is where most estimates go wrong, because founders use their own list price for every customer. A more defensible approach:

• Collect public pricing from competitors in each segment. Pricing pages are the most reliable public evidence of what buyers pay.
• Where vendors only quote on request, use third party estimates if you can find them and label them as estimates.
• Use a different annual value per size band. A 60 person company and a 900 person company do not pay the same.

If you are building a competitor table at the same time, our guide on building a competitor landscape shows how to capture pricing with a source and date for each row, which feeds straight into this step.

A worked example you can copy

Here is the structure for an imaginary expense approval tool aimed at mid-sized companies in one country. The numbers below are placeholders to show the arithmetic, not real market data. Replace each one with a sourced figure from your own research.

The value of this layout is that anyone reading it can challenge a single line without throwing out the whole estimate. If a reviewer thinks E is too high, you change one cell and everything else updates.

Step four, narrow TAM to SAM honestly

SAM should reflect what your product can serve today, not after the next three roadmap items. Common filters:

• Language and localization. If the product only works in English, non-English markets are outside SAM.
• Integrations. If you connect to two accounting systems, companies on other systems are outside SAM until you build more.
• Compliance. Regulated industries that require certifications you do not have are outside SAM.
• Company size. If your onboarding cannot handle enterprise procurement yet, enterprise is outside SAM.

Being strict here helps you. A smaller, credible SAM is more persuasive than a large one that falls apart under the first question.

Step five, set a SOM you can defend

SOM is a share of SAM over a stated period. There is no universal percentage, and quoting one without reasoning is the fastest way to lose a skeptical reader. Instead, build it from capacity:

• How many deals can your sales and marketing realistically close per month by year three?
• What win rate do you see today against the competitors in your landscape?
• How many strong competitors serve the same SAM, and are any of them losing ground?

Multiply deals per year by average annual value and compare the result with SAM. If your capacity based figure is a small single digit share of SAM, it is usually easier to defend than a large round percentage.

Sources that cost nothing

• National statistics offices for company counts by size and industry.
• Competitor pricing pages for spend per customer.
• Public company filings, where competitors are listed, for revenue and customer counts they disclose themselves.
• Job postings, which show which segments competitors are investing in.
• Industry associations, which often publish member counts and surveys.

Keep a list of every source with the date you accessed it. Public figures get revised, and a dated source list lets you update the model in minutes rather than redoing it.

Doing it faster

The slow parts of this method are collecting company counts, reading dozens of pricing pages and keeping the sources straight. MarketAgent's market sizing runs the bottom up method for any question you ask, shows the method for each number, marks estimates as estimates and links every input to its source. It sits inside the market map, so the size estimate and the competitor landscape come from the same research. You can see the plans on the pricing page, or try a one-time demo of our AI market research agent on the homepage.

Mistakes that undermine a market size

• Using the whole software category as TAM. "The global CRM market" is not the TAM of a niche CRM add-on.
• One price for every customer. Spend varies a lot by company size. Use bands.
• Unlabeled estimates. Mixing sourced facts and guesses without marking which is which makes the whole model look like a guess.
• Counting the same buyer twice. If a company buys at group level, do not count each subsidiary as a separate customer.
• No date. A market size without a year is impossible to compare or update.

Frequently asked questions

Is TAM, SAM, SOM still useful for early stage startups?

Yes, mostly as a way to show you understand who buys and why. Investors care less about the precise TAM figure and more about whether the logic behind it holds. A clear bottom up model is more convincing than a large top down number.

How long does a bottom up market sizing take?

A first version with sourced company counts and competitor pricing takes a focused afternoon for one market. Adding several regions or segments multiplies the work, which is why teams often start with the home market and extend from there.

Should I include free tools and spreadsheets in the market size?

Include the companies that currently use them in your buyer count, because they have the problem. Do not assign them zero spend if your plan is to convert them. Use the price they would pay for a paid solution in their size band.

What if there is no public data for my niche?

Use the closest broader statistic you can find and apply clearly stated filters. Show a low, base and high case. Readers accept uncertainty when it is visible. They do not accept hidden guesses.

How often should a market size be updated?

At least once a year, and whenever the inputs change materially, such as a major competitor changing its pricing. Weekly competitor monitoring tells you when one of your pricing inputs has moved.