How to Estimate Total Addressable Market (TAM) Step by Step

To estimate total addressable market, multiply the number of customers who could buy your product by the annual revenue you would earn from each one (bottom-up), then sanity-check that number against published industry spending (top-down). If the two figures land within roughly 2x of each other, you have a number you can defend. If they don't, the gap tells you which assumption is wrong.

You do not need a research firm for this. You need a spreadsheet, a few hours of public data digging, and the discipline to write down every assumption so someone else can challenge it.

What TAM actually measures

TAM is the total annual revenue available for a product category if every possible customer bought from you. It is a ceiling, not a forecast. Nobody wins 100% of a market, and nobody should plan around it.

Three terms get mixed up constantly, so fix them early:

• TAM (total addressable market): everyone who could use your type of product, at your price, worldwide or in your defined region.
• SAM (serviceable addressable market): the part of TAM you can reach with your current product, geography, language and sales channels.
• SOM (serviceable obtainable market): the share of SAM you can realistically win in the next three to five years given your team, budget and competition.

Investors care most about TAM and SOM. You should care most about SAM and SOM, because those drive hiring, spend and quota. TAM tells you whether the game is worth playing. SOM tells you what to do on Monday.

Top-down vs bottom-up at a glance

My opinion: treat bottom-up as the real answer and top-down as the audit. Founders who lead with a top-down number ("the market is $50B and we only need 1%") are telling investors they haven't done the work. That 1% argument is the most recognizable weak spot in any pitch.

Step-by-step: how to estimate total addressable market

Step 1: Define the customer and the job

Write one sentence: "We sell [product] to [specific customer type] so they can [job]." Be narrow. "Small businesses" is not a customer. "US independent veterinary clinics with two or more vets" is.

If you struggle to define the buyer sharply, work through customer segmentation methods first. A fuzzy customer definition is the root cause of most bad TAM numbers.

Step 2: Count the units

Find how many of those customers exist. Good free sources include government census and business registries, trade association member counts, industry directories, LinkedIn company searches filtered by size and category, and app store or marketplace listings. Pick a count, write down the source and the date.

Then filter. Not every company in the category fits your product. Remove those who are too small to pay, too large to buy from you, owned by a chain with centralized purchasing, or in regions you can't serve.

Step 3: Set a realistic annual revenue per customer

Use your actual or planned price, not an aspirational one. If you don't have a price yet, use the three closest competitors' published pricing and pick the middle. Annualize it. Include only recurring or repeatable revenue, not one-time setup fees you will not collect again.

If different customer types pay different amounts, build separate rows rather than averaging. A blended average hides the segment that actually matters.

Step 4: Multiply and document

Units times annual revenue per unit gives you bottom-up TAM. Put every assumption in its own labeled cell with a source and a confidence rating (high, medium, low). Anything marked low is what you test next.

Step 5: Run the top-down check

Find a credible figure for total spending in your category: an industry report summary, a public company's filings describing its segment, or a trade association's survey. Then narrow it using clear, defensible filters (region, product type, customer size). Each filter needs a reason, not a guess.

Step 6: Reconcile the gap

Compare the two numbers. A difference under 2x is normal and healthy. A difference of 5x or more means one method has a flaw, usually a category mismatch in the top-down figure or a missing segment in the bottom-up one.

Step 7: Narrow to SAM and SOM

Cut TAM down to what you can serve (SAM), then to what you can win (SOM) using your sales capacity. This last step is where the number becomes a plan.

Worked example: scheduling software for vet clinics

Say you are building appointment scheduling and reminder software for independent veterinary clinics in the US. Here are the assumptions, all illustrative.

Bottom-up

• US veterinary practices: roughly 30,000 (from trade association and directory data)
• Independent, not corporate-owned: 70%, so 21,000
• With two or more vets, the segment with real scheduling complexity and budget: 60% of those, so 12,600
• Price: $300 per month, or $3,600 per year

Bottom-up TAM: 12,600 x $3,600 = about $45M per year.

Top-down

Suppose a trade report suggests US veterinary software spending is roughly $1.2B per year.

• Share that goes to scheduling, reminders and client communication: 15%, so $180M
• Independent clinics only: 70%, so about $126M

Top-down estimate: about $126M.

Reconciling

The top-down figure is nearly 3x the bottom-up one. That is inside the range worth investigating, not panicking over. Two things explain it. The top-down number includes single-vet clinics, which the bottom-up model excluded on purpose. It also includes spend on bundled practice management suites that cost far more than a standalone $300 tool.

After adjusting for both, you decide the honest TAM is about $45M to $60M, with upside if you add features that justify a higher price. That is a useful, defensible answer.

SAM and SOM

You have two sales reps and sell mainly through trade shows and referrals in the Midwest and Southeast, covering about 4,000 qualifying clinics. That is your SAM: 4,000 x $3,600 = about $14.4M.

At a realistic 6% win rate over three years, you land roughly 240 clinics. That is a SOM of about $860K in annual recurring revenue.

For a bootstrapped company, that is a real business: it funds a small team. For a venture-backed plan expecting $50M in revenue, this market is too small unless you expand into adjacent products. That conclusion is the point of the exercise. You learned it in a day, not after a year of building.

Common mistake: the "1% of a huge market" shortcut

The most common error is starting from a giant industry number and claiming a tiny share. "The global pet industry is worth hundreds of billions, so even 0.1% is a huge business." It sounds modest but proves nothing, because the number includes food, toys and services you will never sell.

How to avoid it:

• Start from customers you can name and count, not from a market report.
• Use top-down only as a ceiling, and never as the main argument.
• For every percentage filter you apply, write one sentence explaining why that percentage is right.
• If you can't explain a filter to a skeptical customer, remove it.

A second mistake worth naming: counting everyone who has the problem instead of everyone who pays to solve it. Plenty of people dislike scheduling chaos. Far fewer will pay $300 a month. Anchor your count to people or companies who already spend money on something similar.

Stress-testing your number

Once you have a figure, try to break it before someone else does.

• Call 10 to 15 target customers. Ask what they pay now for the closest alternative. If most say "nothing" or "a spreadsheet," your price assumption is weak.
• Flip the price. Recalculate at half and double your planned price. If the market only works at one price, your TAM is fragile.
• Check the competition. Count how many customers your top competitors claim. If three of them together report more clinics than your TAM says exist, your unit count is too low. Our competitive analysis guide for small businesses shows how to pull those numbers quickly.
• Check your own bias. Founders anchor on the biggest number they've seen. The write-up on cognitive biases in business decisions covers anchoring and optimism bias, both of which inflate TAM.

For a broader review of whether the plan holds together, the stress testing guide is a good next step.

Turning TAM into decisions

A market size has no value unless it changes what you do. Use it this way:

• If SAM is under about $10M: you are in a niche. That can be great for a profitable small business, but plan for a ceiling and think about adjacent products early.
• If TAM is large but SOM is small: your bottleneck is distribution, not market size. Spend your effort on channels, which is covered in how to build a go-to-market strategy from scratch.
• If bottom-up and top-down disagree wildly: stop and fix the model before presenting it to anyone.

Revisit the estimate every quarter during the first year. Your price, win rate and segment definitions will all move once real customers are involved, and a stale TAM misleads you as much as a wrong one.

Key takeaways

• Build TAM bottom-up first: countable customers times realistic annual revenue per customer. Use top-down only as a ceiling and sanity check.
• Define the customer narrowly before you count anything. A vague buyer produces a vague, inflated number.
• A gap of more than about 2x between your two methods means a flawed assumption, so investigate before you present the figure.
• Always carry TAM down to SAM and SOM. Your plan should be built on what you can win, not on the total market.
• Never rely on the "1% of a huge market" argument. Write down the source and reasoning for every filter and percentage.
• Test price and demand with 10 to 15 real customer conversations, and update the model quarterly.