Bottom-up market sizing starts with a list of real accounts

An investor points at the $4B on your market slide and asks where it came from. If the answer is an analyst report on a category one level above yours, the conversation moves on without you. If the answer is a count of companies and what each would pay you, it doesn’t.

That count is bottom-up market sizing. It takes longer than pulling a number from a report, and it’s the version investors and sales leaders can both use, because the same list of accounts becomes your territories and your outbound list.

Bottom-up market sizing is accounts times price

Bottom-up market sizing estimates a market by counting the customers who could buy and multiplying by what one of them would pay a year. For a B2B company the formula is short.

• Flat pricing. Market size = number of accounts × annual price per account.
• Seat or unit pricing. Market size = number of accounts × typical seats or units per account × price per seat or unit. Matrix Partners’ TAM guide (2017) uses HR software at $20 per employee per month, multiplied by the employees at companies of 100 to 5,000 people.
• Usage or take-rate pricing. Market size = number of accounts × volume per account × your take, not the total volume.

Use the same price measure everywhere, usually ACV or average revenue per account. Run the count with no limits and you get TAM. Apply the limits that are true today (regions, sizes, integrations) and you get SAM.

Top-down vs bottom-up market sizing

Top-down market sizing starts from a published industry total and narrows it with percentages. Bottom-up starts from one customer and multiplies out. Both have a place, but they answer different questions.

	Top-down	Bottom-up
Starts from	Analyst or government market total	A count of accounts and a price
Formula	Industry total × share that’s your category × share you can serve	Accounts × annual price, by segment
Speed	An hour	A few days
Best for	Sanity-checking scale	Pitch decks, sales planning, territories
Tends to run high when	The report bundles services, hardware or buyers you don’t sell to	Pilot pricing is assumed for every account
Tends to run low when	The category is too new to appear in reports	Upsell, expansion and later products are left out
Can you test it?	Hard; the inputs aren’t yours	Yes, against your CRM and your list

Investors lean the same way. Pear VC’s market sizing guide (2021) tells founders to use bottom-up and keep top-down to sanity-check the magnitude, because bottom-up assumptions can be tested against the rest of the pitch: the customer definition, the revenue model and the go-to-market plan. Matrix’s guide is blunter: a number lifted from an analyst report suggests the founder searched for a number big enough.

A third method, value theory, sizes a market by the value your product creates rather than what buyers spend today. It’s for products that create a new category, and we cover it in our post on TAM expansion.

A bottom-up market sizing example for a B2B startup

Illustrative numbers: a startup sells dock scheduling software to US third-party logistics (3PL) warehouse operators. Here’s the bottom-up approach to market sizing, step by step.

1. Define the account in filters you can search

“Logistics companies” can’t be counted. “US 3PL warehouse operators with 50 to 1,000 employees” can. Write the definition in the same firmographics you’d use to build a list: industry, headcount, region, and the systems they run if you depend on an integration.

2. Count the accounts, by size band

Price usually rises with company size, so count each band separately instead of using one average for everyone.

Size band	Accounts	Annual price	Market
50 to 199 employees	2,800	$12,000	$33.6M
200 to 499 employees	1,000	$30,000	$30.0M
500 to 1,000 employees	400	$60,000	$24.0M
Total	4,200		$87.6M

That $87.6M is the TAM for this definition. Notice that the smallest band holds two-thirds of the accounts but well under half the market. A single average price would have hidden that.

3. Apply today’s limits to get SAM

Say the product only integrates with the warehouse management systems used by about half of these operators. Rerun the table on those accounts and the result is SAM. Shipping the next integration would grow SAM more than any campaign.

4. Get to SOM from capacity

The last step is what your team can win in the period. Don’t pick a percentage of SAM. Build it from the accounts your reps can work, your conversion and your win rate, as in our guide on how to calculate serviceable obtainable market.

5. Check it against a top-down number

Now run the same market top-down. Say a report puts US warehouse software at $2B and you assume dock scheduling is a tenth of it. That’s $200M, more than twice the bottom-up count. The gap is the finding: the report includes large shippers that run their own warehouses, plus modules this startup doesn’t sell. Show both numbers and explain the difference, and the investor doesn’t have to.

Where the account count comes from

The count is the part that takes time. Pitch Doctor, written by a former VC, makes a useful distinction: investors distrust market size estimates from reports but are fine with a report’s count of the buyers. These are the sources that hold up when someone checks them.

• US Census business counts. County Business Patterns counts establishments by industry code and size class, and Statistics of U.S. Businesses counts firms by enterprise size.
• Regulator and registry lists. In regulated industries a license list is often the most complete count there is. In trucking, FMCSA’s Pocket Guide to Large Truck and Bus Statistics (2024 edition) counts 787,189 active interstate and intrastate hazmat motor carriers as of December 2023, 519,420 of them for-hire.
• Company databases. Firmographic data providers and LinkedIn Sales Navigator filter by industry, headcount and region.
• Enrichment tools. Clay combines several providers and dedupes them, which matters when the sources disagree.
• Your own CRM. Closed-won accounts tell you which filters actually predict a buyer.

A lookalike TAM build from your best customers is a fast way to test whether the count is realistic. If the lookalike list comes back much smaller than your estimate, the definition is too loose.

Mistakes that inflate a bottom-up market size

• One price for every account. If price grows with size, an average hides most of the market, as the table above shows.
• Counting accounts you can’t serve. Regions you don’t sell in and systems you don’t integrate with belong outside SAM.
• Total volume instead of your share. A payments or marketplace business sizes its fees, not the money that moves through it.
• Revenue that isn’t annual. If customers buy every two years, count half of them per year.
• Establishments counted as companies. Census establishment counts include every location, so a 3PL with 12 warehouses shows up 12 times. Use firm counts when you sell one contract per company.
• Never refreshing it. New products, prices and integrations all change the count.

Your market size should become your account list

A bottom-up market size is an account list with a price on it. Built properly, it tells you how to split accounts into territories, how many reps the list can support before they’re calling the same companies, and which segment to work first.

That’s why the count is worth doing carefully. Building and maintaining that account data for partners is part of our datasets work, and the ICP model build shows how we test the filters against closed-won before a rep touches the list.

When the count says the market is too small

Sometimes the bottom-up number comes back smaller than the plan needs. That’s useful to learn before the raise, not after. The options are a wider account definition you can actually serve, a higher price, or a second segment, which is the question our post on TAM expansion works through.

Start with the definition. Write your target account in filters a list builder could search, count it, and put that number next to the one on your slide.