How to run a market sizing and TAM validation study
Most TAM numbers are built from analyst reports and hope. This guide shows you how to validate market size with primary research you can actually defend to investors.
To run a market sizing and TAM validation study, you build two independent estimates of your market, one top-down and one bottom-up, then validate the assumptions inside them with primary research from verified target buyers. The number that survives both methods and holds up against real demand, adoption, and willingness-to-pay data is the one you can defend to investors.
Most TAM slides fail the moment someone asks how you got there. A single analyst report, a “1% of a huge market” hand-wave, and no evidence that anyone actually wants the product. This guide walks through the method to fix that: define your market layers, run both estimation approaches, collect the primary data that makes the numbers real, and triangulate into a range you trust.
What a TAM validation study actually proves
A market sizing study answers “how big is the opportunity.” A TAM validation study goes further. It tests whether the assumptions inside that number, adoption, budget, willingness to pay, are true for real buyers. Sizing without validation is arithmetic. Validation turns it into evidence.
You are trying to answer three questions:
- How large is the addressable market in revenue terms?
- What share of it can we realistically win?
- Do actual buyers show enough demand and willingness to pay to make those numbers credible?
The first two are estimation. The third is primary research, and it is the part most teams skip. It is also the part that separates a number you can defend from a number you invented.
Step 1: Define TAM, SAM, and SOM
Before you estimate anything, define the three market layers. Conflating them is the most common sizing mistake.
| Layer | What it measures | Question it answers |
|---|---|---|
| TAM (Total Addressable Market) | Total annual revenue if every possible buyer in the category used your product | How big is the whole category? |
| SAM (Serviceable Addressable Market) | The portion of TAM your product, segment, and geography can actually serve | What can we reach today? |
| SOM (Serviceable Obtainable Market) | The realistic share you can capture in a set period given competition and capacity | What can we win next? |
TAM is the ceiling. SAM narrows it to who you can serve given your current product, pricing, language support, and regions. SOM is what you can realistically win in the next one to three years given your sales motion and competition.
Write each layer as a sentence with explicit boundaries. For example: “TAM is all mid-market and enterprise B2B SaaS companies globally that run customer research; SAM is those in North America and Europe with a dedicated research or product ops function; SOM is the ones actively evaluating a research platform in the next year.” Clear boundaries make the estimates that follow far easier to defend.
If you are pre-product or pre-revenue, defining these layers is doubly important, because your assumptions are all you have. Our guide on how to do customer research with no users yet covers how to gather signal before you have a customer base.
Step 2: Build a top-down estimate
Top-down sizing starts from a published market figure and narrows it with filters. It is fast, useful for a sanity check, and inherently limited because it inherits someone else’s assumptions.
The logic is:
Top-down TAM = published market size x relevant segment filters
Start with a credible category figure from a recognized source. Government statistics agencies and industry census data are stronger anchors than a single vendor’s press release. For example, the U.S. Census Bureau’s Statistics of U.S. Businesses gives defensible counts of firms by size and industry, which you can use to bound how many potential accounts exist.
Then apply filters that match your SAM definition: geography, company size, industry vertical, and the share of that group with the underlying problem you solve. Each filter should map to a boundary you set in Step 1.
The weakness of top-down is obvious once you write it out. Every filter is a percentage you assumed. That is exactly why you do not stop here.
Step 3: Build a bottom-up estimate
Bottom-up sizing builds the market from your own unit economics. It is slower but far more defensible because every input is something you can measure or validate directly with buyers.
For a B2B product, the core formula is:
Bottom-up TAM = number of target accounts x adoption rate x annual contract value (ACV)
- Target accounts: How many companies fit your ICP? Count them from firmographic databases, industry directories, or census data, filtered to your segment.
- Adoption rate: What share of those accounts would realistically buy a product like yours? This is the input most teams guess. It is also the one primary research validates best.
- ACV: What would each account pay per year? This comes from your pricing and, critically, from willingness-to-pay research, not from a spreadsheet cell you like.
For a bottom-up SOM, layer in reach and conversion:
SOM = target accounts you can reach x expected win rate x ACV
The power of bottom-up sizing is that it forces you to name your assumptions. The risk is that a wrong adoption rate or ACV, multiplied across thousands of accounts, produces a wildly wrong number. That is why the adoption and ACV inputs must be validated with real buyers rather than assumed. If your ICP itself is shaky, tighten it first with ICP validation and a customer segmentation study before you multiply anything.
Step 4: Compare the two methods
Now put both numbers side by side. This comparison is the heart of a defensible TAM.
| Dimension | Top-down | Bottom-up |
|---|---|---|
| Starting point | Published market figure | Your own unit economics |
| Speed | Fast | Slower |
| Main risk | Inherits analyst assumptions | Wrong adoption rate or ACV scales up |
| Defensibility | Low on its own | High when inputs are validated |
| Best use | Sanity check and outer bound | Primary estimate you present |
If the two methods land in the same range, you have a strong signal. If they diverge sharply, you have found a flawed assumption, and finding it is the point. A 10x gap between top-down and bottom-up usually means either your adoption rate is fantasy or your segment filters are wrong. Reconcile before you present anything.
Step 5: Validate the assumptions with primary research
This is where a sizing exercise becomes a validation study. The two inputs that most need real evidence are adoption intent and willingness to pay. Both come from surveying and interviewing the actual buyer.
Who to survey
Survey the person who owns the budget and makes the decision in your target segment. A general consumer panel cannot validate a B2B adoption rate, because the respondents are not the buyers. You need verified firmographics: the right role, company size, and industry. When the people behind your adoption and ACV numbers are provably your ICP, the numbers become credible.
This is one reason teams use a verified B2B panel. CleverX is a two-sided research platform with an 8M+ panel of professionals verified by work email and LinkedIn, across 150+ countries, so the firmographics behind your bottom-up inputs hold up when an investor probes them. For guidance on reaching the right people, see how to recruit B2B research participants and, for senior buyers, how to recruit enterprise buyers for research.
Sample sizes
You do not need thousands of responses for directional validation. A practical rule:
- 30 to 50 verified buyers per segment is often enough to move adoption and pricing from guesswork to a defensible range.
- Larger samples (100+) tighten your margin of error if you need to defend precise percentages to a board or an investor.
- 5 to 10 interviews per segment before the survey to write questions in the buyer’s language and surface assumptions you missed.
The survey and interview questions that validate demand
Validation rests on three question themes: demand, adoption, and willingness to pay. Interview first to understand the language, then field a survey to quantify.
Demand and problem severity
- How do you handle [problem] today?
- How much time or money does that cost you per month?
- How painful is this on a scale where you would actively look for a solution?
Adoption intent
- If a product solved [problem] in [way], how likely are you to evaluate it in the next 6 months?
- Who else would need to approve that decision?
- What would have to be true for you to switch from your current approach?
Willingness to pay
- What do you spend today on tools or people to handle this?
- Which budget line would this purchase come from?
- Structured pricing questions, such as a Van Westendorp battery or a simple minimum-acceptable and too-expensive pair.
Never rely on a single stated price. Combine direct questions with a structured method and cross-check against current spend. Our pricing sensitivity survey and Van Westendorp guide and the deeper collect willingness-to-pay data from B2B buyers walk through the mechanics, and B2B SaaS pricing research methods covers the wider toolkit.
Well-run interviews are their own skill. If your team is new to them, 50 user interview questions that uncover real insights and 5 common user interview mistakes that ruin your research will keep your demand signal clean.
Step 6: Triangulate into a defensible range
Triangulation means arriving at your estimate from more than one independent direction and showing that they agree. A defensible TAM validation combines:
- Top-down estimate as the outer bound and sanity check.
- Bottom-up estimate as the primary number you present.
- Primary research validating the adoption rate and ACV inside the bottom-up model.
Present a range, not a single point. State your assumptions explicitly, show which primary data supports each key input, and cite your sources. Investors trust a smaller, well-derived number over a giant number with no derivation. The credibility comes from the method, not the size.
A quick defensibility checklist:
- Did two independent methods land in the same range?
- Is every major assumption written down and sourced?
- Are the adoption and pricing inputs backed by verified buyers, not guesses?
- Can you explain how you got from TAM to SAM to SOM in one sentence each?
If you want the strategic context around sizing, basics of market research and the complete walkthrough to product research methods place this study in a wider program, and pre-launch demand testing with real buyers is a natural next step once your sizing holds up.
Common mistakes to avoid
- The “1% of a huge market” trap. Starting from a giant number and claiming a small slice proves nothing about demand. Build bottom-up instead.
- Surveying the wrong people. Consumer panels cannot validate B2B adoption. Verify roles and firmographics.
- One method only. A single estimate with no cross-check is a guess with a spreadsheet.
- Stated price as gospel. Buyers overstate willingness to pay. Triangulate against current spend.
- Skipping interviews. Without qualitative work first, your survey asks the wrong questions in the wrong language.
Where CleverX fits
Bottom-up sizing is only as strong as the buyers behind the numbers. CleverX lets you field both the qualitative interviews and the validation survey against verified professionals, typically delivered in about 2 to 5 days on pay-as-you-go credits, so you can move from assumption to evidence inside a single planning cycle. Because participants are verified by work email and LinkedIn, the adoption and willingness-to-pay inputs in your model carry firmographic proof when someone challenges them. When you are ready to gather that evidence, start recruiting verified participants on CleverX.
If you would rather compare platforms before you commit, our roundup of the best platform for market sizing and TAM validation covers the tooling options in detail.
Frequently asked questions
What is the difference between TAM, SAM, and SOM?
TAM is the total annual revenue available if every possible buyer used your category. SAM is the slice you can reach with your current product, segment, and geography. SOM is the realistic share you can capture in a defined period given competition, sales capacity, and budget.
Should I use top-down or bottom-up market sizing?
Use both and compare them. Top-down starts from a published market figure and narrows it down, which is fast but inherits analyst assumptions. Bottom-up builds from your real unit economics, which is slower but far more defensible. When both methods land in the same range, your estimate is credible.
How much primary research do I need to validate a TAM?
You do not need thousands of responses. For directional validation of adoption intent and willingness to pay within a segment, 30 to 50 verified target buyers per segment is often enough to move from guesswork to a defensible range. Larger samples reduce your margin of error if you need to defend precise numbers.
Who should I survey for a TAM validation study?
Survey the actual buyer and budget owner in your target segment, not a general consumer panel. Verify their role, company size, and industry so the firmographics behind your adoption and pricing assumptions hold up. Interviewing a few buyers before you field the survey helps you write questions in their language.
How do I estimate willingness to pay in a sizing study?
Combine direct pricing questions with structured methods like Van Westendorp or a simple gag-and-max approach, and cross-check against what buyers currently spend on alternatives. Never rely on a single stated price. Anchor willingness to pay to the budget line the purchase would come from.
How do I make my TAM defensible to investors?
Triangulate at least two independent methods, show your assumptions explicitly, and back the key inputs like adoption rate and ACV with primary evidence from verified buyers. Investors trust a smaller number with a clear derivation more than a huge number with no sources.