I once spent six weeks building a pitch deck around a market I'd sized at $4.2 billion. Then an investor asked me one question: "Where did that number come from?" I said "Statista." He closed the deck and said, "Come back when you've counted your customers." That was three years ago, and I've never sized a market the same way since.
Evaluating market size for a startup idea isn't about finding a big number. It's about proving you understand who pays you, how many of them exist, and what they'll realistically spend. Get that wrong and you'll build for a market that doesn't exist, or pitch to investors who've seen a hundred founders make the same mistake.
Key Takeaways
- Bottom-up beats top-down for credibility, but you need both to sanity-check each other.
- Your realistic obtainable market matters more than the total addressable market—investors know this and so should you.
- Average revenue per customer is the number founders get wrong most often.
- A $500,000-revenue business is typically worth 2–4x revenue or 4–8x EBITDA, depending on the sector.
- Test demand with paid acquisition before you commit to a market thesis.
How to evaluate market size for a startup idea without fooling yourself
Most market sizing guides tell you to calculate TAM, SAM, and SOM. Fine. But they skip the part where founders inflate every one of those numbers by 10x and call it research.
I've done this. You probably have too. The trick is building a process that forces you to defend each assumption.
Start with the market boundary, not the number
Before any calculation, define who you actually serve. Not "small businesses." Not "healthcare." Those aren't markets, they're categories. A market has a boundary: a specific customer type, a specific problem, a specific buying trigger.
When I worked on a scheduling tool for dental clinics, my first draft market definition was "healthcare practices in North America." Useless. The real boundary was single-location dental practices with 2–5 chairs that already use digital booking software. That narrowed roughly 200,000 practices down to about 18,000. Small enough to actually count.
Ask yourself: if I had a list of every potential customer, what would be on it?
The bottom-up method: number of customers × revenue per customer
This is the approach that wins investor meetings, and honestly, it's the only one I trust anymore.
The formula is simple:
- Step 1: Estimate how many potential customers exist (from a public source, a directory, a census, or a scraper)
- Step 2: Estimate what percentage could realistically buy your product
- Step 3: Multiply by average revenue per customer per year
- Step 4: Sanity-check the result against a top-down figure
Here's a concrete example. Say you're building a project management tool for freelance video editors.
According to the U.S. Bureau of Labor Statistics, there are roughly 40,000 film and video editors in the country. Suppose 60% work freelance (24,000), and you estimate you can reach and convert 8% of those over five years (1,920 customers). If you charge $30 per month, that's $691,200 in annual recurring revenue at full penetration.
Is that a venture-scale business? Probably not. Is it a solid bootstrapped business? Absolutely.
That result—unsexy as it is—tells you more about your business than any $10 billion TAM figure ever will.
The top-down method: TAM, SAM, SOM
Top-down works the other way. Take an industry-wide revenue figure and slice it down to the portion you can realistically capture.
| Layer | What it means | Example: coffee subscription for offices |
|---|---|---|
| TAM | Total addressable market—every dollar spent on the category globally | Global office coffee services: ~$12 billion |
| SAM | Serviceable addressable market—the slice you can serve with your model | U.S. offices with 20–200 employees: ~$1.8 billion |
| SOM | Serviceable obtainable market—what you can realistically capture in 3–5 years | 1% of SAM: ~$18 million |
Two things kill most top-down estimates. First, the sources are usually paid and outdated. Second, founders confuse TAM with SAM and pitch the whole thing as if they're going to own it.
Rule of thumb: if your SOM is more than 3% of your SAM within five years, you're probably lying to yourself. Real companies take years to grab single-digit share in a fragmented market.
Cross-checking the two methods
The real value comes from running both and comparing. If your bottom-up estimate gives $50 million and your top-down gives $5 billion, something is broken. Usually it's the top-down number, because you included every possible customer instead of the ones who'd actually pay.
I once saw a founder pitch a $40 billion market for a niche B2B compliance tool. His bottom-up estimate was $15 million. Guess which one the investors believed.
Where to get real market data (without paying $5,000 for a report)
You don't need a Statista subscription to size a market. You need patience and a browser.
Free data sources that actually help
- Government statistics: The U.S. Bureau of Labor Statistics, Census Bureau, and Eurostat publish free industry data
- Trade associations: most industries have one, and they publish annual reports with member counts and average revenue
- Public company filings: SEC EDGAR is a goldmine if your space has any listed players
- Reddit and niche forums: not for the numbers themselves, but to find the right questions to ask
- LinkedIn Sales Navigator: filter for your exact customer profile and see how many results come back
The LinkedIn trick is underrated. I once built a rough customer count in an afternoon just by filtering job titles and industries. Took me two hours. Compare that to a $3,000 analyst report that told me the same thing.
Estimating average revenue per customer when you have no customers
This is the hardest part, and it's where most founders hand-wave.
Here's what I do: find 5–10 comparable products, look at their public pricing pages, and build a range. If you're selling a $40/month SaaS and your competitors charge between $25 and $80, use the median. If your pricing is fundamentally different (usage-based, enterprise contracts), find companies with similar models and look at their disclosed average contract values.
Another trick: look at job postings at competitor companies. Sales roles often list quota targets, which gives you a rough sense of deal sizes.
Be honest about this number. If you assume $500/month per customer and the market median is $50, you're not sizing a market—you're writing fiction.
How much is a business worth with $500,000 in sales?
A business with $500,000 in annual revenue is typically valued at 2–4x revenue, or 4–8x EBITDA, depending on the industry and growth trajectory. That puts most such businesses in the $1 million to $2 million range—but this varies enormously.
SaaS companies with strong retention can hit 5x revenue. Service businesses with low margins might only get 1x. Retail and restaurants often trade closer to 2–3x EBITDA, which for a $500K-revenue restaurant with 10% margins is barely $100,000.
Here's the distinction that trips people up: revenue is not market size. A business doing $500K in sales is not proof that the market is worth $500K. It's one data point about what one operator has captured. Your market sizing should tell you how many businesses like that could exist.
The mistakes that will sink your market sizing
I've made most of these. Learn from my scars.
Mistake #1: Confusing TAM with SAM
Yes, the global market for "productivity software" is enormous. No, that doesn't mean you can sell to it. Every founder does this at least once. Catch yourself before an investor does.
Mistake #2: Founder bias
If you're building the product, you naturally believe more people want it than actually do. I once spent three months convinced that 30% of marketing agencies would buy my tool. The real number, based on 40 customer interviews, was closer to 4%. That's a 7x overestimate. Painful to discover, more painful to ignore.
Mistake #3: Trusting one source
If your entire market thesis fits on a single Statista slide, you haven't done the work. Triangulate. Ask customers. Test with ads.
Testing real demand before you commit
Market sizing is theory. Testing is proof.
Run a small paid ad campaign targeting your supposed market. Spend $500 on Google or LinkedIn ads with a clear value proposition and a landing page. If you get a 0.5% click-through rate and zero signups, that's data. It doesn't mean the market is dead, but it means your assumptions deserve scrutiny.
Pre-sales are even better. If you can convince 10 people to pay you $100 before the product exists, you've learned more about market viability than any report will teach you.
I've done this three times. Twice it validated the thesis. Once it saved me from six more months of building something nobody wanted.
Final thought: the number matters less than the reasoning
No investor will fund you because your TAM is $50 billion. They'll fund you because your reasoning is tight, your customer definition is sharp, and your path to capturing a meaningful slice is believable.
The size of a market isn't a fact you look up. It's an argument you make—and the stronger the argument, the smaller the market can afford to be.
