How To Do Market Sizing: 10 Steps Explained
If you're a student who's just been asked, in an internship or a case interview, to size a market and then explain how you got there, you've probably felt the specific panic of staring at a blank page with no idea where the first number even comes from. Market sizing gets treated like a math problem, but the real skill is building a process you can defend, not landing on the exact right figure.
This piece walks through the ten steps that take you from a vague question, like how many people might actually buy this, to a number you can walk into a meeting or an interview and stand behind. Each step builds on the last, so the order matters more here than it does in most business skills you'll pick up on the job.
Market Sizing is a great way to understand businesses, and the skills used there are also useful for business internships. If you’re curious as to what you can learn, check out this blog here!
This piece walks through the ten steps of a full market sizing exercise: defining the market, choosing top-down or bottom-up, sizing the TAM, narrowing to the SAM, estimating the SOM, listing assumptions, sanity-checking, cross-checking with a second method, showing your math, and stating a confidence level. A defensible process matters more than landing on one exact "correct" figure. Whenever you have time, calculate your number two different ways and see if they land in the same range. It's useful for any student prepping for a case interview or asked to validate a business idea during an internship.
Why is market sizing crucial even outside of consulting interviews?
Market sizing shows up more often than most students expect, not just in consulting case interviews but inside real companies deciding whether an idea is worth building at all. A founder deciding whether to launch a new product, or a team deciding whether to expand into a new region, is really asking a market sizing question, even if nobody calls it that out loud.
At a program like Ladder Internships, where students are placed on real, project-based work with actual startups and nonprofits, this question comes up directly: a partner company wants to know whether an idea is worth the resources it would take to build, and an intern is often the one asked to run the first pass at the number. Getting comfortable with the process matters more than getting the exact figure right on your first attempt. More on that below.
What does a full market sizing process actually involve?
Before the ten steps, it helps to know there are really only two broad approaches, and most real exercises end up blending them. Top-down starts with a big, known number and narrows it down, while bottom-up starts small, with a real unit like one customer or one location, and builds up from there. Case interviews tend to reward whichever approach you can defend clearly, not one specific method.
The ten steps below work for either approach, and the later steps, on pressure-testing and presenting your number, matter just as much as the math itself. A number nobody can defend is barely more useful than no number at all.
Step 1: Define the market you're actually sizing
Before you calculate anything, write down, in one specific sentence, exactly what you're sizing and for whom. "The market for coffee" isn't a market sizing question, "the market for delivery coffee ordered by office workers in a specific city" is. The single most common mistake in market sizing is skipping this step and calculating the wrong market precisely.
Get specific about the product, the customer, and the geography before you touch a single number. Once you've defined it clearly, everything else in this process gets easier, and every one of the next nine steps depends on this definition staying fixed.
Step 2: Decide between top-down and bottom-up
Top-down starts with a large, already-known number, like total industry spend, and narrows it down through a series of percentages until you reach your specific market. It's fast and works well when solid industry data already exists.
Bottom-up starts from a single real unit, one customer, one store, one transaction, and multiplies up. It's slower to build but usually more defensible, because every number in a bottom-up calculation traces back to something you can actually justify, rather than a percentage you picked because it sounded reasonable.
Step 3: Size the total addressable market (TAM)
Your total addressable market, or TAM, is the full size of the opportunity if you captured every possible customer, with no competition and no constraints. It's the biggest, roughest number in the whole exercise, and it's supposed to be.
Look for the broadest number that's still genuinely relevant: total spending in the category, total number of potential customers, or total unit volume, whichever a reasonable source is likely to publish. Don't worry yet about how much of it you could realistically capture; that comes next.
Step 4: Narrow to your serviceable addressable market (SAM)
Your serviceable addressable market, or SAM, is the slice of the TAM your specific product or business could actually serve, given real constraints like geography, price point, or product fit. If your TAM is every coffee drinker in the country, your SAM might be office workers in cities where your delivery model actually works.
This is where a lot of the real judgment in market sizing happens. Be honest about which constraints are real, and don't inflate your SAM just to make the final number look bigger.
Step 5: Estimate your serviceable obtainable market (SOM)
Your serviceable obtainable market, or SOM, is the realistic share you could actually capture, usually within a defined timeframe like a first year or two, given competition, your resources, and how fast a business like this could realistically grow. This is the number that actually matters for a business decision.
Ground this in something concrete: a comparable company's early market share, a realistic sales capacity, or a conversion rate you can defend. A SOM pulled from thin air is the fastest way to lose credibility in an interview or a real pitch.
Step 6: Write down every assumption you're making
As you work through TAM, SAM, and SOM, every percentage and every number you didn't measure directly is an assumption, and it needs to be written down, not buried in your head. List the population figure you used, the percentage you applied, and where each one came from.
This list becomes the most useful part of your work later, both because you'll need to defend each one and because it's usually where you catch your own mistakes first.
Step 7: Pressure-test your assumptions with a sanity check
Once you have a number, run it against something you already know to be true and see if it holds up. If your calculation implies more coffee orders per day than there are office workers in the city, something upstream is wrong.
Sanity checks are fast, and they catch the errors that actually matter: an extra zero, a percentage applied twice, a population figure for the wrong region. Do this before you present anything, not after someone else catches it.
Step 8: Cross-check with a second method
Whenever you have time, calculate the same number a second way and see how close the two land. If your top-down and bottom-up estimates land within the same rough range, that's a strong signal your number is reasonable.
If they land far apart, that's useful information too. It usually means one of your assumptions is off, and finding which one is often more valuable than either number on its own.
Step 9: Show your math, not just the final number
The final number matters far less than the process that got you there, especially in an interview or a real project. Walk through your assumptions in order, state where each one came from, and be upfront about which parts are estimates rather than facts.
This is also the step where a real placement teaches you something a mock case interview can't. At a program like Ladder Internships, students present a final deliverable to an actual manager at the end of a project, and a partner company evaluating your market sizing wants to see your reasoning as much as your conclusion, since that's what they'll actually rely on when making a real decision.
Step 10: Round appropriately and state your confidence level
A market sizing exercise built on assumptions doesn't deserve a precise-looking answer. Round your final number and say so out loud: "roughly $40 million" reads as more credible than "$41.3 million," because it signals you understand the limits of your own estimate.
Pair the number with a short statement of confidence: which assumptions you're most and least sure about, and what additional data would tighten the estimate. That honesty is usually what separates a strong answer from a merely correct-sounding one.
How do you use a market sizing number to actually make a decision?
A market sizing exercise is only useful if it changes a decision. In an interview, that means connecting your number back to the original question, is this market worth entering, is this feature worth building, not just stopping once you've calculated a figure. The number should answer something, not just exist on its own.
On a real project, like the kind of placement Ladder Internships builds around pairing a student with an actual startup manager and a dedicated coach, that number usually feeds directly into a recommendation: whether an idea is worth pursuing, and at what scale. Treat the market size as the first input into a decision, not the final deliverable itself.
Common questions about market sizing
1. Do I need real research data, or can I use reasonable assumptions?
Reasonable, clearly stated assumptions are normal and expected, especially in an interview setting where you don't have access to real data. What matters is that each assumption is defensible and clearly labeled as an estimate, not presented as a fact.
2. What if my top-down and bottom-up numbers don't match?
That's common, and it's more useful than either number matching perfectly by coincidence. Look at which assumption is driving the gap, and use that to refine whichever method feels shakier.
3. How precise does the final number actually need to be?
Far less precise than most students assume. A well-reasoned range, or a rounded figure with a clearly stated confidence level, is usually more credible than a number with several decimal places behind it.