I once watched a founder tell investors their TAM was $47 billion. The investor asked one question: “How many companies is that?”
Silence. Because the $47B came from a market report, not from counting anything. And a market size you can’t decompose into actual companies isn’t analysis. It’s decoration.
Let’s do the version you can defend.
📌 TL;DR: A defensible TAM is countable companies times annual contract value, built bottom-up from real market data. Top-down slices of analyst reports impress nobody anymore. And SAM and SOM are what keep the big number honest.
What is TAM analysis?
TAM analysis is the process of calculating your total addressable market, the full revenue opportunity if every potential buyer bought. It answers the sizing question every strategy conversation eventually hits: how big is this market, really?
And the total addressable market never travels alone. Two smaller siblings keep it honest. SAM (serviceable addressable market) is the slice your model and geography can actually serve. SOM (serviceable obtainable market) is the slice you can realistically win near-term, given competition and capacity.
| Level | Question it answers | Example |
|---|---|---|
| TAM | How big is the whole market potential? | All companies worldwide that could use the product |
| SAM | How much can our model serve? | Those in our regions, size bands, and languages |
| SOM | What can we win soon? | The share our team and brand can realistically take |
Three numbers, three different jobs. Investors read TAM for ambition, SAM for focus, and SOM for whether you’re honest.
And it’s not just for pitch decks. The same math sets sales targets, sizes territories, and tells you when a segment is saturated. Market potential is an operating number, not a fundraising prop.
Which method should you trust: top-down, bottom-up, or value theory?
Bottom-up, almost every time in B2B. But know all three, because you’ll meet all three:
Top-down starts from an analyst’s market figure and slices it: “the market is $50B, we can address 10%, so TAM is $5B.” Fast, and fragile. Every assumption is someone else’s, and the slicing percentages are usually invented on the spot. This is how $47B decorations get made.
Bottom-up starts from counting: how many companies fit your profile, times what each is worth annually. Slower, and every input is checkable. That checkability is the entire point.
Value theory estimates what buyers WOULD pay based on the value your product creates. It’s the right tool when you’re inventing a category and no pricing history exists anywhere. For most B2B products, it’s a supporting argument, not the main math. The finance-education treatments at Wall Street Prep and CFI walk through all three in more depth.
When bottom-up and top-down disagree (and they will), trust the one built from countable pieces.
Does top-down have any use at all? Yes, one. It’s a sanity check. If your careful count lands within shouting distance of the analyst figure, you can present both and look twice as credible. That’s the only job it does well.
How do you calculate TAM step by step?
You calculate TAM in five steps: define who counts, count them, price them, derive SAM and SOM, then sanity-check. The first two steps decide the quality of everything after.
- Define who counts. This is your ideal customer profile written as countable criteria: industries, size bands, regions. Vague profile, vague TAM.
- Count the companies. Use firmographic databases, industry registries, and provider counts to tally companies matching each criterion. This is a data job, not a guess. The same company data sources that power enrichment power sizing. Cross-check two sources; where they disagree, understand why before averaging.
- Price each segment. Multiply company counts by a realistic annual value per segment. Use your actual ACV (annual contract value) by size band if you have history, comparable pricing if you don’t. Enterprise and SMB deserve separate math. For consumer or usage-priced products, swap ACV for ARPU (average revenue per user); the structure of the formula doesn’t change.
- Derive SAM and SOM. SAM: apply your real constraints (regions served, sizes your product fits). SOM: apply competitive reality, your win rates, your capacity. Write the assumptions next to the numbers.
- Sanity-check against the top-down. If your bottom-up TAM is 40x the analyst figure, or a fortieth of it, one of you has a definition problem. Find it.
→ Bottom-up TAM = (number of companies that fit) × (average annual contract value)
And where does the count come from on day one? Often from your own CRM. If your records carry enriched firmographic fields, the account enrichment coverage report is the honest first draft of your market size, plus a gap number showing how much of the market you haven’t even logged yet.
A worked example
Say you sell compliance software to mid-size European manufacturers. A hypothetical bottom-up build looks like this:
- Manufacturers, 100 to 1,000 employees, in your six target countries: counted from firmographic data, say the tally comes to 18,000 companies
- Your ACV in that band: €12,000
- TAM: 18,000 × €12,000 = €216M
- SAM: you only support three languages today, so 11,000 companies, €132M
- SOM: given two entrenched competitors and your sales capacity, a defensible 3-year share of SAM, whatever YOUR win-rate evidence supports
Every number in that chain can be challenged, checked, and updated. That’s the whole point. An investor’s “how many companies?” now has an answer with a source. Visible.vc’s TAM modeling template shows the same build in spreadsheet form if you want a starting file.
💡 Geo tip: want to slice the TAM by territory? Location quotient shows where your industry actually concentrates, region by region. It turns a flat market number into a territory map, and it usually surprises you.
That’s the trick I lean on for regional planning: run the count, then check it against location quotient before drawing territory lines.
Turn the TAM into a list you can work
Here’s the part most TAM guides skip: the best thing about a bottom-up TAM is that it’s not just a number. It’s a list.
Because you counted actual companies, you can name them. The 18,000 manufacturers from the example aren’t an abstraction; they’re rows. Filter them by your SAM constraints and you’re holding your target-account universe. Score them by fit and you’ve got tiers. Hand them to sales and the market-sizing exercise just became pipeline.
A top-down TAM can’t do any of that. You can’t email 10% of an analyst estimate.
So when you finish the count, don’t file it. Export it. The same spreadsheet that convinced the board becomes the list your team works Monday morning. That double use is why I’ll always argue for the slower method: the decoration number dies in the pitch meeting, but the counted number keeps working after it.
The TAM mistakes that undermine credibility
Five patterns show up in almost every shaky market-size slide:
- The decoration number. A giant top-down figure with no company count underneath. Everyone senior has seen through it before.
- Counting the industry, not the profile. “All manufacturers” isn’t your market if your product only fits ones above 100 staff.
- One blended ACV. Averaging enterprise and SMB pricing into one number distorts both.
- Static TAM. Markets move. Re-run the count yearly; with current firmographic data it’s an afternoon.
- Confusing TAM with a target. TAM is the ocean. Your plan lives in SOM. Presenting TAM as a goal reads as either naive or evasive.
Notice what all five have in common? Each one is a shortcut. The decoration number skips counting, the blended ACV skips segmenting, the static TAM skips maintenance. Market sizing rewards the boring work, and punishes every shortcut in front of exactly the audience you least want watching.
🧠Remember: nobody funds you to win the TAM. The TAM proves the ocean exists; SOM proves you know which bay you're fishing in first. Present all three levels and let each do its own job.
The board deck that taught me to count
My own conversion moment came in Hamburg, building a board deck. The old slide said our market was “over €2 billion,” sourced from a report nobody had read past the summary.
So I rebuilt one segment bottom-up. DACH software companies, 50 to 500 employees, counted from two firmographic databases. It took me about three afternoons, mostly reconciling why the two sources disagreed by a few thousand companies.
The new number was smaller. Noticeably smaller. And the board liked it MORE, because when someone asked “how many companies?”, I had a count, a source, and the assumptions written on the slide. The smaller honest number beat the bigger decorated one. It always does.
How I know this (and what to check yourself)
The method here matches the standard finance-education treatments (Wikipedia’s TAM entry is a fine neutral overview, and Antler’s TAM/SAM/SOM guide covers the startup framing), plus my own years of building sizing models from firmographic counts.
Two honest limits. Company counts differ between databases, so always cross-check two sources and note the gap. And any ACV you borrow from comparables is an assumption; label it as one. For the fuller market picture beyond sizing, pair your TAM with industry benchmarks.
Frequently Asked Questions
What is TAM vs SAM vs SOM?
TAM is the total revenue opportunity if everyone who could buy did; SAM is the portion your business model can serve; SOM is the portion you can realistically win near-term. Each filters the previous: TAM, then constraints, then competition and capacity.
How do you calculate TAM, SAM, and SOM?
Count the companies that fit your profile, multiply by annual contract value for TAM. Apply your real serving constraints (regions, sizes, languages) for SAM. Then apply win rates and capacity for SOM. Bottom-up, with sources named at each step.
What is considered a good TAM?
Big enough to support the outcome you’re aiming for, and honest enough to survive diligence. Venture-scale outcomes usually want a TAM in the billions; a healthy bootstrapped business can thrive on a fraction of that. A good TAM is mostly a credible one.
Is market size the same as TAM?
Not quite. TAM is one specific measure of market size: the total revenue potential if every possible buyer bought. “Market size” gets used loosely for current spending, forecasts, or TAM depending on who’s talking, so always ask which definition is on the slide.
Is TAM calculated annually?
TAM is usually expressed as annual revenue potential, and yes, it’s worth recalculating every year. Markets shift, your ICP sharpens, and databases update. With current firmographic data, the yearly re-count is an afternoon, not a project.
What is TAM in data analytics?
Outside market sizing, TAM appears as an acronym for other things (like technical account manager). In a strategy or analytics context, though, TAM almost always means total addressable market, the metric this guide covers.
It’s time to count your market for real
Take one segment of your ICP. One industry, one size band, one region. And actually count the companies this week. Not estimate. Count, from data.
That single number changes the quality of every strategy conversation you’ll have this quarter.
You’ve got this. Tell me in the comments: does your current TAM slide have a company count behind it, or a market report? No judgment. Okay, a little judgment.
