Market Sizing That Doesn’t Suck: Ditch TAM/SAM/SOM for Something Useful

Tired of the TAM/SAM/SOM theater? Explore two practical, persona-first alternatives to market sizing that go beyond inflated top-down estimates.

Let’s be honest: TAM/SAM/SOM is mostly theater.

If you don’t know, TAM, SAM, and SOM are market size metrics commonly used to assess a business's potential: Total Addressable Market (TAM) is the entire market, Serviceable Addressable Market (SAM) is the portion a company can realistically target, and Serviceable Obtainable Market (SOM) is the portion a company can realistically capture.

It’s that set of concentric circles that shows up in every pitch deck. It makes your market opportunity look big, clean, and exciting. It signals that you’ve done your homework. But it’s also mostly made-up numbers based on some combination of analyst reports, wishful thinking, and the need to make your startup look like a future unicorn.

“If we could only get 1% of China” or “sales reps around the world would want this.”

Sounds great in theory. 

I’ve never had much confidence in TAM/SAM/SOM. And the more I work with product managers—especially those in B2B software—the more I see how disconnected that framework is from how we actually build and sell products.

Instead of starting with some imaginary market definition, I prefer to flip the process: start with the persona, then figure out how many people actually fit that profile and have the problem you solve. From there, you can model a market that’s grounded in reality.

After all, products solve problems for personas. So, how many personas have the problem? And are willing and able to pay for a solution?

Here are two practical, persona-first market sizing techniques I recommend instead of the TAM/SAM/SOM theater show.

Persona-based Sizing

This method starts with a very simple question:

“How many people or companies look like my ideal customer, and what could I reasonably sell to them?”

Start by defining a persona who fits the ideal customer for your business. Not in fluffy aspirational terms, but with actual firmographics or demographics: company size, industry, geography, team structure, job titles, tech stack, maturity level—whatever dimensions make sense for your market.

Once you’ve got that persona defined, start counting. This is where tools like LinkedIn Sales Navigator, ZoomInfo, Apollo, Crunchbase, or even job boards become your best friends. You’re looking for a way to observe how many people or companies actually match your persona.

Say you’re building a product for customer success managers in SaaS companies with over $10M in annual recurring revenue (ARR). Great. Go find out:

• How many SaaS companies in that size range exist?

• How many have dedicated customer success teams?

• How many customer success people do they employ per company?

It’s detective work. It’s messy. But it’s grounded.

Then you multiply:

• Number of ideal customer accounts × average number of users per account × annual revenue per user.

That gives you your realistic, reachable market. Not a theoretical total market if everyone on Earth suddenly adopted your product—but a number you could plausibly build a business on.

If you want to get fancy, model this as a pyramid:

• Tier 1: Closest fit (core ideal customers)

• Tier 2: Adjacent segments that might adopt later

• Tier 3: Broader ecosystem

It gives you a roadmap for expansion, not just a one-and-done number.

Problem-based Sizing

This one’s especially useful when your product addresses a recurring pain—something that costs time, money, or reputation every time it happens.

Rather than starting with personas, this approach starts with the problem itself.

Here’s the question:

“How often does this problem show up, and how much is it worth fixing each time?”

Let’s say your product helps reduce system downtime for DevOps teams. You’d model:

• Number of target companies (based on your ideal customers)

• Average number of critical incidents per year

• Estimated cost per incident (e.g., lost revenue, support costs, service level agreement penalties)

Let’s throw in some fake math:

• 10,000 target companies

• 12 major incidents/year/company

• $5,000 cost per incident

That’s a $600M problem space. Now, if your product could realistically help reduce 25% of those costs, that’s a $150M opportunity. That’s a useful market sizing conversation—not “the global DevOps market is $3.2B,” which is a stat that sounds good but helps no one plan anything.

You can also layer on buying triggers. Maybe not all 10,000 companies feel enough pain to act. But 2,000 do. Now your model isn’t just about market potential—it’s about urgency. That makes it actionable.

Bonus: Add Deal Size and Sales Model

Once you’ve sized the reachable market using personas or incident frequency, multiply it by your expected average contract value or annual revenue per user. This will give you a top-down revenue opportunity that matches your bottom-up reality.

You’ll also want to be realistic about your sales model. If you’ve got a high-touch sales process, don’t kid yourself about scaling to 10,000 accounts with five salespeople. Build a model that reflects how many customers you could realistically reach and close, not how big the theoretical market is.

Addendum: Use Category Proxy Analysis as a Sanity Check

If you’re using persona- or problem-based sizing (and you should be), it’s still smart to cross-check your numbers with a third approach: Category Proxy Analysis. Sounds fancy but it’s not.

Here’s how it works:

• Find an existing product or category that’s adjacent to yours.

• Look at how big they are—how many customers, what their average deal size is, how they’ve grown over time.

• Use their traction as a reality check on your assumptions.

If you’re claiming a $200M market but the best-known player in the space is doing $15M in annual recurring revenue after five years, you might want to take a closer look at your model. Either you’ve found a magic go-to-market engine no one else has—or you’re making assumptions that don’t hold up.

It’s not about copying their market size. It’s about asking:

“Does my model hold up when compared to real-world analogs?”

When all three approaches—persona-based sizing, problem-based sizing, and category proxy analysis—roughly align, you’ve got something you can actually plan against.

Final Thought: Make It Useful, Not Impressive

Market sizing isn’t about impressing investors. It’s about informing decisions:

• Should we invest in this product?

• How big could it reasonably get?

• Where should we focus first?

TAM/SAM/SOM might make for pretty slides, but real market sizing is done in spreadsheets and back-of-the-envelope math with messy inputs and testable assumptions.

Start with the persona. Map the real pain. Run the numbers. Then ask yourself the one question no TAM/SAM/SOM diagram ever answers:

Is this a business worth building?

If the answer’s yes, that’s all the market sizing you really need.

---

Do you want more practical methods about managing products? Learn more about our Fundamentals of Managing Products program.