The most defensible way to build a TAM, SAM, and SOM is to use a bottom-up methodology rooted in concrete unit economics, firmographics, and actual purchasing behavior. Top-down sizing (taking a massive macro-analyst number from Gartner or IDC and slicing off a percentage) is routinely rejected by sophisticated investors and operators because it inflates the opportunity and offers zero operational utility.

Step-by-Step Bottom-Up Framework

1. Define Your Ideal Customer Profile (ICP) & Geography
   • Explicitly outline who can buy from you today based on company size, industry, tech stack, or buyer persona, and restrict your geography to where you have an active go-to-market (GTM) motion.
2. Calculate TAM (Total Addressable Market)
   • Formula: T o t a l n u m b e r o f I C P a c c o u n t s g l o b a l l y/i n-s c o p e×A v e r a g e A n n u a l C o n t r a c t V a l u e(A C V)o r A R P U.
   • Defensibility check: Use verified registry counts (like census data, LinkedIn company filters, or industry directories) rather than guessing.
3. Filter to SAM (Serviceable Addressable Market)
   • Formula: Apply realistic business constraints—such as your current language capabilities, regulatory barriers, distribution channels, or product tier limitations—to narrow the TAM.
   • Defensibility check: SAM answers what portion of that total universe you can physically and logistically service right now.
4. Narrow to SOM (Serviceable Obtainable Market)
   • Formula: Apply competitive density, sales team capacity, and marketing reach over a defined window (typically 1 to 3 years).
   • Defensibility check: Your SOM must align directly with your near-term financial forecast. If your SOM is $20M, projecting $20M in year-one revenue implies a 100% capture rate, which is a red flag.

Founder Perspectives on Market Sizing

I found that starting with a bottom-up count of actual potential customers and multiplying by realistic pricing made the numbers much easier to defend.

VCs look at TAM to see the scale of your vision, but they look at your SOM and SAM to see if you actually understand your sales capacity and go-to-market constraints.

If you want to tailor this to your specific business, tell me:

• Is your product B2B or B2C?
• What is your target geography and estimated ACV/Price point?

I can help you outline the exact calculation logic and data inputs for your model.

Yes. I searched current guidance, and the consistent answer is: build TAM/SAM/SOM bottom-up from observable customer economics, then use top-down research as a triangulation check—not as the foundation.

The defensible approach

1. Define the market before you size it

Write down, explicitly:

• Who buys? — specific ICP/persona
• What are they buying? — precise product/use case
• Geography
• Time period — e.g. 2026
• Revenue basis — annual revenue/ARR, transaction volume, etc.
• Price metric — ACV, ARPU, take rate, transactions/customer, etc.

Avoid definitions like "the global AI market." A defensible definition looks more like:

US mid-market B2B SaaS companies with 50–500 employees that have an internal sales team and spend >$X/year on sales tooling.

The narrower definition is actually a strength because you can defend it.

---

2. Build TAM bottom-up

The basic equation is:

TAM = total potential customers × annual revenue/customer

For a SaaS example:

180,000 qualifying companies × $12,000 ACV
= $2.16B TAM

But don't just put "180,000" into a spreadsheet. Show where it came from.

For example:

The evidence chain is more important than the final number.

Bottom-up sizing is considered more defensible precisely because every major assumption can be interrogated.

---

3. Make SAM a constraint, not an arbitrary percentage

This is where many TAM/SAM models fall apart.

Don't do:

$2.17B TAM × 40% = $868M SAM

unless you can explain why 40% exists.

Instead, identify the actual constraints:

SAM = TAM × product fit × geographic availability × regulatory/channel constraints × other structural limitations

For example:

• TAM: 180,480 companies
• Only North America: 55%
• Only companies using Salesforce: 65%
• Product currently supports English only: 90%

Then:

180,480 × 55% × 65% × 90% = 58,000 SAM accounts

At $12K ACV:

SAM ≈ $696M

Each filter should correspond to a real business constraint.

---

4. Build SOM from your actual go-to-market capacity

This is the most important—and most frequently abused—number.

Don't say:

"We'll capture 2% of the SAM."

Instead ask:

How many customers can we actually win?

For example:

• 58,000 SAM accounts
• 5,000 accounts reachable through your initial channels
• 1,000 sales-qualified opportunities/year
• 25% close rate
• 250 new customers/year
• $12K ACV

Year-1 SOM:

250 × $12K = $3M ARR

Then model the next 3–5 years based on:

• sales capacity
• marketing-generated demand
• conversion rates
• sales cycle
• churn/retention
• geographic expansion
• hiring
• channel partnerships

That's much more defensible than "we'll get 1%."

Current market-sizing guidance similarly recommends treating SOM as the operational forecast rather than simply applying an arbitrary percentage to SAM.

---

5. Triangulate with top-down data

Now—and only now—look for industry reports, government statistics, public-company filings, trade associations, etc.

Suppose your bottom-up calculation says:

TAM = $2.2B

But several credible sources imply the entire relevant category is ~$1B.

That's a red flag.

Conversely, if your bottom-up estimate is $2.2B and credible external sources imply a $1.8–3B market, that's reassuring.

The point isn't to make the numbers identical. The discrepancy is diagnostic.

Current guidance specifically recommends using top-down and bottom-up estimates together and investigating large gaps.

---

6. Put an evidence grade on every assumption

This is the trick that makes a TAM model really defensible.

I'd add columns like:

Then distinguish:

Facts → derived calculations → assumptions → hypotheses

Don't let them all masquerade as equally certain.

---

7. Use ranges, not false precision

Instead of:

TAM = $2,173,481,293

I'd present:

TAM: ~$2.0–2.5B
Base case: $2.2B

And build:

• Bear
• Base
• Bull

Then identify the 2–3 variables that actually drive the range.

For example:

TAM is relatively insensitive to adoption, but SOM is highly sensitive to ACV and sales capacity.

That's much more useful to an investor or strategy team than a giant number with six significant digits.

---

The model I'd actually build

I'd structure the spreadsheet like this:

Sheet 1 — Market definition

• ICP
• use case
• geography
• exclusions
• pricing unit
• time period

Sheet 2 — TAM

• total potential accounts
• segment breakdown
• price/ACV
• bottom-up calculation
• top-down cross-check

Sheet 3 — SAM

• geography filters
• product constraints
• regulatory constraints
• channel constraints
• eligible accounts

Sheet 4 — SOM

• reachable accounts
• leads/opportunities
• conversion
• sales capacity
• CAC
• sales cycle
• churn
• annual revenue

Sheet 5 — Sources

• source
• URL
• publication date
• exact metric
• page/table
• assumption supported

Sheet 6 — Sensitivity

• bear/base/bull
• ACV
• customer count
• penetration
• conversion
• sales capacity

---

The "defensibility test"

Someone skeptical should be able to ask:

"Where did that number come from?"

for every major cell.

And you should be able to answer with one of:

1. Observed fact — "Census says X."
2. Internal data — "Our customers pay Y."
3. Primary research — "We interviewed 30 buyers."
4. Derived calculation — "X × Y."
5. Explicit hypothesis — "We assume 15%; here's why, and here's the sensitivity."

If the answer is:

"We found a report saying the market is $40B and assumed we'd get 5%..."

that's not a defensible TAM/SAM/SOM. It's a top-down market report with a haircut.

My preferred hierarchy

Primary customer/account data
↓
Government / regulatory / company filings
↓
Industry associations & credible research
↓
Comparable-company data
↓
Expert interviews
↓
Your explicit assumptions
↓
Generic market reports / AI-generated estimates

The further down the list you go, the more aggressively you should label the number as an assumption.

In one sentence: Build the market from customers and economics upward, constrain it based on what you can actually serve, forecast SOM from your ability to acquire customers, and use external market reports to prove your answer isn't crazy.

If you're building this for a fundraising deck, investment memo, or internal strategy case, I can also show you what a genuinely investor-grade TAM/SAM/SOM model looks like with a worked example and spreadsheet formulas.

Building a defensible Total Addressable Market (TAM), Serviceable Addressable Market (SAM), and Serviceable Obtainable Market (SOM) requires backing up your numbers with solid logic rather than pulling grand totals out of thin air. Investors and stakeholders can spot a top-down "we just need 1% of a $100B market" claim instantly.

Here is a step-by-step framework to build a bulletproof, defensible market sizing model.

1. Build from the Bottom-Up (Primary Method)

• Start with unit economics : Calculate your average annual contract value (ACV) or average revenue per user (ARPU).
• Define your exact target profile : Use firmographics or demographics (e.g., mid-market logistics companies in North America with 100–500 employees).
• Count the actual entities : Multiply your target customer count by your ACV to get your core addressable baseline. Bottom-up math proves you actually understand who you are selling to.

2. Define the T-S-S Hierarchy Clearly

• TAM (Total Addressable Market) : The total revenue opportunity if 100% of the market bought your type of solution, regardless of geography or constraints.
• SAM (Serviceable Addressable Market) : The segment of the TAM targeted by your products and services which is within your geographical and regulatory reach.
• SOM (Serviceable Obtainable Market) : The realistic share of the SAM you can capture in the near-term (typically 2 to 3 years) given your current sales capacity, budget, and competition.

3. Cross-Check with Top-Down Data

• Use reputable third-party reports : Pull data from recognized sources like Gartner, IDC, or industry associations to sanity-check your bottom-up findings.
• Acknowledge the delta : If your bottom-up TAM is $500M and a Gartner report says the global market is $5B, explain the difference. Usually, the macro report includes adjacent product categories you aren't building yet.

4. Document Every Assumption

• List your variables explicitly : Keep formulas transparent in a spreadsheet so you can change inputs (like price points or adoption rates) on the fly.
• Cite your data sources : Link every customer count, pricing assumption, and geographic filter to a concrete source or internal pilot data.

If you want, tell me:

• What is your product or industry?
• Are you using a bottom-up or top-down approach right now?

I can help you outline the exact formulas and data sources for your specific model.

The most defensible way is to build TAM/SAM/SOM bottom-up from observable customer economics, then use top-down data as a sanity check. The mistake is starting with a huge industry report and assuming you’ll capture 1%.

A framework that holds up in diligence

1. Define the unit of demand first

Before looking for market-size reports, write down:

Buyer × use case × geography × price × purchase frequency

For B2B SaaS, for example:

# target companies × annual ACV = annual market opportunity

For a marketplace:

# transactions × take rate × addressable transaction volume

For consumer:

# eligible households/users × annual spend per user

This prevents the model from becoming disconnected from how you actually make money.

2. Build TAM from the customer universe

Ideally, get to an actual count:

Example

• 80,000 companies meet your ICP
• Average annual spend on the relevant solution = $12,000
• TAM = $960M

Even better, segment it:

Now every major number is challengeable—and therefore defensible.

3. Derive SAM using structural filters, not arbitrary percentages

For example:

TAM = $960M
→ US only = $600M
→ industries you currently support = $420M
→ customers reachable through your current channel = $300M

SAM = $300M

The important distinction: don't say "SAM is 30% of TAM because that's our assumption." Say which customers are excluded and why.

4. Build SOM from your actual GTM engine

This is where many TAM/SAM/SOM models fall apart.

Don't say:

"We'll capture 5% of SAM."

Instead model something like:

Target accounts × sales capacity × meetings/account × opportunity rate × win rate × ACV

For example:

• 2,000 reachable target accounts
• 10 sales reps
• 200 accounts actively worked per rep over 3 years
• 15% become qualified opportunities
• 25% win rate
• $30k ACV

That produces a $22.5M 3-year revenue opportunity.

That's dramatically more credible than "$300M SAM × 5% = $15M SOM."

5. Triangulate it

Run two independent models:

Bottom-up:
Customers × price × frequency

Top-down:
Published category spend → relevant segment → geography → product category

If they come out roughly similar, confidence increases. If they're 5–10× apart, don't average them. Find the assumption causing the divergence. Triangulation is useful precisely because disagreement exposes weak assumptions.

6. Put evidence behind every important assumption

I'd classify every input as:

• A — observed: government data, filings, your CRM, actual pricing
• B — externally validated: reputable industry research, surveys, third-party datasets
• C — assumption: management judgment
• D — hypothesis: needs primary research

Your goal is to eliminate C/D assumptions from the variables that materially drive the result.

For example:

35,000 target accounts — A
$25k ACV — A/B
60% have the required infrastructure — B
20% willing to switch — C
25% win rate — C/D

Now you know exactly what research you need to do next.

7. Use ranges, not fake precision

I'd model:

Bear / Base / Bull

and run sensitivity on the 2–3 assumptions that matter most.

For instance:

This is much more intellectually honest than claiming "$800M TAM" as though you know it to the nearest dollar.

The key principle

TAM tells you whether the opportunity is big enough.
SAM tells you whether your business model can address enough of it.
SOM tells you whether you can actually build a meaningful business.

For an investor deck, I'd therefore make SOM the most rigorously built number, not TAM. A credible $500M TAM with a well-supported $30M obtainable market is generally more compelling than a $10B TAM with "1% market share" as the entire argument.

If you're doing this for a specific company/product, give me the product, target customer, geography, pricing/ACV, and business model, and I can lay out the actual TAM/SAM/SOM model and the data sources I'd use.

The most defensible approach is bottom-up first, top-down second. The goal isn't to produce the biggest number; it's to build a model where every major assumption can be traced to observable evidence.

1. Start with the unit of demand

Define exactly what you're selling and who buys it:

1 customer = [specific buyer] in [specific segment/geography] paying [$X/year] for [product].

For example:

• Customer = U.S. companies with 100–1,000 employees
• Buyer = VP of Finance
• Product = AP automation
• Price = $18k/year

This prevents the classic mistake of mixing "users," "companies," "transactions," and "market dollars."

2. Build TAM from actual customer counts

Use:

TAM = total theoretically eligible customers × realistic annual revenue/customer

Get the customer count from primary or highly credible datasets, such as:

• Government databases
• Regulatory filings
• Industry associations
• Census/BLS data
• Company databases
• Public company filings
• Your own customer/prospect data

Then establish price independently.

Example:

100,000 eligible companies × $20k ACV = $2.0B TAM

The important thing is that someone can audit both inputs.

3. Define SAM with explicit filters

Don't say:

"Our SAM is 30% of TAM."

Instead, show the filters:

Then:

9,000 × $20k = $180M SAM

Each filter should have an independent rationale/source.

This is much more defensible than applying arbitrary percentages.

4. Make SOM an operating model, not a market-share guess

This is where I would be especially rigorous.

Don't do:

$180M SAM × 5% = $9M SOM.

Instead, ask how many customers can you actually acquire?

For example:

SOM = reachable accounts × sales capacity × conversion × retention × ACV

Suppose:

• 9,000 SAM accounts
• 1,500 are reachable through your initial channels
• 20% become qualified opportunities
• 25% close
• $20k ACV

That's:

1,500 × 20% × 25% × $20k = $1.5M

Then sanity-check that against your actual sales capacity:

• Number of reps
• Ramp time
• Opportunities/rep/month
• Win rate
• Sales cycle
• CAC
• Churn/retention

That makes SOM essentially a bottom-up revenue forecast, rather than an arbitrary market-share assumption. Bottom-up approaches are generally viewed as more credible because they tie the market directly to customers, pricing, and GTM mechanics.

5. Triangulate it three ways

The strongest model has three independent views:

A. Bottom-up

customers × ACV

B. Top-down

Industry spend × relevant category %

C. Demand/budget-based

What customers currently spend on the problem you're replacing

If all three land in roughly the same neighborhood, confidence goes way up.

If they don't, don't average them. Investigate why.

6. Separate facts from assumptions

I'd literally have an assumptions table:

This is one of the best ways to make the model withstand investor diligence.

7. Stress-test the answer

Don't present one magical number.

Run bear / base / bull cases across the variables that actually matter:

• Customer count
• ACV
• Adoption rate
• Product eligibility
• Win rate
• Sales capacity

For example:

Now the conversation becomes "What has to be true for this business to become $12M?" rather than "Why did you choose 3% market share?"

The investor-grade version

I'd structure the eventual market slide around one sentence:

We have X addressable customers × $Y annual revenue/customer = $Z TAM; our current product/GTM serves A customers = B SAM; our sales capacity supports C customers over five years = D SOM.

Then put the detailed calculations and sources in an appendix.

Biggest rule: never let TAM/SAM/SOM be disconnected from the actual business model. If your five-year SOM implies 2,000 customers but your sales organization can only close 300, the SOM is not defensible—even if the underlying TAM is enormous.

If you tell me what the product is, who buys it, pricing/ACV, geography, and current traction, I can help you build the actual TAM/SAM/SOM model from the ground up.

Building a defensible Total Addressable Market (TAM), Serviceable Addressable Market (SAM), and Serviceable Obtainable Market (SOM) requires blending bottom-up data with top-down sanity checks , clear assumptions, and rigorous segmentation. Investors and stakeholders immediately spot-check top-down estimates (like grabbing a $50B industry stat), so a defensible model must be built from the ground up.

Core Framework Definitions

• TAM (Total Addressable Market): The total revenue opportunity available if 100% market share is achieved for a product or service.
• SAM (Serviceable Addressable Market): The portion of TAM targeted by the product/service which is geographically and segment-wise within reach.
• SOM (Serviceable Obtainable Market): The realistic share of the SAM that can be captured in the near-to-mid term (typically 1–3 years) given competition and capacity.

Step-by-Step Guide to Building a Defensible Model

1. Build a Bottom-Up TAM
   • Calculate using unit economics: Multiply the total number of potential customers (businesses or consumers) by the average annual contract value (ACV) or average revenue per user (ARPU).
   • Formula: TAM = (Total Potential Customers) × (Annual Revenue per Customer).
   • Why it's defensible: It relies on countable, verifiable parameters rather than broad macroeconomic percentages.
2. Narrow to a Segmented SAM
   • Filter your TAM by your actual geographic reach , regulatory limitations, language barriers, or specific product constraints (e.g., enterprise-only vs. SMB).
   • Formula: SAM = TAM × (% of market in your geographic/segment footprint).
   • Why it's defensible: It explicitly acknowledges who you cannot sell to right now.
3. Calculate a Realistic SOM
   • Ground this in your current sales capacity , marketing budget, historical conversion rates, and existing market competition.
   • Formula: SOM = SAM × (Realistic Market Share percentage for Years 1 -- 3).
   • Why it's defensible: It connects your market size directly to your operational go-to-market (GTM) execution plan.
4. Apply Top-Down Sanity Checks
   • Use reputable industry analyst reports (like Gartner, IDC, or Statista) to verify that your bottom-up TAM doesn't accidentally exceed 5%--10% of a known broader market without a radical thesis.
   • If your bottom-up TAM says your niche software is a $100B market when the entire global software industry is $500B, your unit assumptions are flawed.
5. Document and Explicitly State Assumptions
   • List every underlying variable (ACV, churn, conversion rates, target customer counts) in an accessible appendix or spreadsheet.
   • Make parameters adjustable so investors or internal teams can stress-test the model by changing inputs like pricing or target company size.

What Real Builders Say

I always prefer a bottom-up TAM based on number of potential customers multiplied by average contract value, rather than a top-down percentage of a giant market.

A top-down number makes you look lazy. Building it from customer counts and pricing shows you actually understand your unit economics and target buyer.

Would you like me to help you map out the specific variables and formula for your particular product, or do you want to share your industry/target customer type to tailor the calculation?

A defensible TAM/SAM/SOM model is built from the bottom-up , grounded in hard unit economics and clear qualification filters, rather than top-down macroeconomic guesses. When investors or strategic leaders interrogate your numbers, they want to see the exact math and logic behind your filters.

To see a visual breakdown of how to transition logically from bottom-up TAM down to a realistic SOM, watch this guide:

56s How to Size Your Market (TAM SAM SOM) the RIGHT WAY 4.4K views  · 2 years ago YouTube · Ed Kang (CFXO)

The Defensible Step-by-Step Playbook

1. Define Your ICP (Ideal Customer Profile) First
   Specify your target customer's precise firmographics (e.g., company size, tech stack, geography) and behavioral constraints. Vague parameters yield indefensible numbers.
2. Calculate Bottom-Up TAM (Total Addressable Market)
   Count the total universe of accounts matching your absolute ICP, then multiply by your realistic Annual Contract Value (ACV) or Average Selling Price (ASP).
   TAM=Total ICP Accounts×Average Annual Revenue per Customer TAM equals Total ICP Accounts cross Average Annual Revenue per Customer T A M=T o t a l I C P A c c o u n t s×A v e r a g e A n n u a l R e v e n u e p e r C u s t o m e r
3. Filter Down to SAM (Serviceable Addressable Market)
   Apply rational business constraints to your TAM. Exclude geographies you lack the rights or localization for, regulatory-blocked verticals, or segments where entrenched competitors make switching impossible.
4. Derive a Sober SOM (Serviceable Obtainable Market)
   Base your SOM on your current GTM capacity, sales velocity, and marketing budget for the next 1–3 years (typically capturing 1% to 10% of your SAM). Avoid the trap of saying, "If we just capture 1% of a multi-billion dollar market..." without showing the sales headcount required to win those deals.
5. Run a Top-Down Sanity Check
   Pull a credible industry report (e.g., Gartner, Statista) to see if your bottom-up TAM converges roughly with the macro-spend in that sector. If they wildly diverge, explain why.

If you'd like to share your industry or product type , I can help you outline the specific bottom-up data points and filters you should use for your model.

The most defensible approach is bottom-up first, top-down second, and triangulate the two.

The key is to make every major number traceable to an observable fact or an explicit assumption—not “the $50B industry × 2%.” Bottom-up sizing is generally more credible because it forces you to specify the actual customer, price, and reachable market.

1. Define the unit of purchase

Start with:

Who actually buys this, what do they buy, and what do they pay?

For B2B SaaS, for example:

Customer count × annual contract value = revenue opportunity

For a transaction business:

Customers × transactions/customer/year × take rate

For a marketplace:

Buyers × purchase frequency × average transaction value × take rate

This is much more defensible than starting with an industry report.

---

2. Build TAM from the full theoretical customer universe

TAM should answer:

If every theoretically eligible customer bought our product, how much annual revenue could exist?

Example:

• 250,000 eligible businesses
• $12,000 average annual spend
• TAM = $3.0B

Importantly, don't arbitrarily constrain TAM to your current geography or sales capabilities. Those constraints belong in SAM.

Use primary/authoritative sources wherever possible—government datasets, industry registries, company filings, pricing data, etc. The SBA, for example, recommends using sources such as NAICS and Census data to quantify businesses and markets.

---

3. Derive SAM using hard filters

SAM answers:

Which portion of TAM can our current product actually serve?

Don't say:

“We estimate 30% of TAM is serviceable.”

Instead, show the filters.

For example:

250,000 total businesses
× 60% in target industries
× 70% in target company-size range
× 80% in supported geographies
× 75% meeting technical requirements
= 63,000 serviceable customers

63,000 × $12,000 ACV
= $756M SAM

Now an investor can challenge each assumption independently.

That's a feature, not a bug.

---

4. Make SOM a capacity model, not a market-share guess

This is where I see the biggest mistake.

Avoid:

“We'll capture 3% of SAM.”

Instead, calculate what your company can actually sell.

For example:

Year 5 sales reps                  30
Qualified opportunities/rep/year  100
Total opportunities             3,000
Win rate                           20%
Customers won                     600
ACV                            $12,000
--------------------------------------
Year-5 ARR                       $7.2M

Then sanity-check that against:

• sales cycle
• CAC
• sales capacity
• churn
• competitive intensity
• implementation capacity
• geographic/channel constraints
• historical conversion rates

That turns SOM from a wishful market-share assumption into an operating forecast.

---

5. Triangulate with a top-down number

Now look for credible external estimates of the broader market.

If your bottom-up calculation says:

$3B TAM

while reputable industry sources imply:

$2–4B

that's reassuring.

If your model says:

$3B

but the entire relevant industry is supposedly only $800M, you have discovered an assumption that needs investigation.

Top-down is therefore best used as a sanity check, rather than the foundation of the model.

---

6. Build an “assumption ledger”

This is what makes the analysis genuinely defensible.

Then classify assumptions:

Observed → derived → estimated → speculative

You want as much of the model as possible in the first two categories.

---

The model I'd actually put in a board/investor deck

I'd make the headline extremely simple:

$3.0B TAM → $756M SAM → $7.2M Year-5 SOM

And underneath:

TAM
250K potential customers × $12K ACV
= $3.0B

SAM
250K × 60% industry × 70% size ×
80% geography × 75% technical fit × $12K
= $756M

SOM
30 reps × 100 opps/rep × 20% win rate × $12K
= $7.2M ARR

Then put the full assumptions and sources in the appendix.

That is much harder to poke holes in than a giant “$10B TAM” circle.

One subtle but important point

I'd also distinguish market size from budget displacement.

If you're selling something new, don't automatically assume the customer has a $12K budget waiting for you. Ask:

What are they spending today to solve this problem?

That might be a competitor, internal labor, consultants, software, lost revenue, or simply nothing.

The strongest TAM models connect the theoretical customer universe to an actual economic behavior and willingness to pay, rather than just multiplying a huge population by an invented price. Market research should likewise examine demand, market size, pricing, saturation, and customer behavior.

If you give me your company/product, target customer, geography, and pricing, I can build a defensible TAM/SAM/SOM model with the actual formulas, sources to use, and an investor-ready market-sizing slide.

A defensible market sizing model is built bottom-up , not top-down. Top-down sizing (taking a giant Gartner or IDC number and arbitrary percentage slices) is easily dismantled by investors, board members, and operators. A bottom-up model ties your market size directly to real unit economics, target customer counts, and sales capacity.

Watch this breakdown on building bottom-up market sizing models:

2m The RIGHT Way to Calculate your Market Size (TAM/SAM/ ...38K views  · 1 year ago YouTube · The Startup Club by Slidebean

1. Define Your Ideal Customer Profile (ICP) and Pricing

• Segment Definition: Explicitly state the firmographic, geographic, or behavioral traits of who you sell to (e.g., US mid-market manufacturing companies with 50–250 employees).
• Value/Pricing Validation: Determine your Average Annual Revenue Per User/Customer (ARPU or ACV) based on actual testing, historical contracts, or validated value theory.

2. Build the TAM (Total Addressable Market) Bottom-Up

• The Formula: Total number of ICP-matching entities globally/broadly × Annual contract value = TAM.
• Defensibility Tip: Use verified registry counts (like census data, LinkedIn company filters, or industry databases) to prove your total entity count rather than guessing.

3. Filter to the SAM (Serviceable Addressable Market)

• The Formula: Apply business model constraints like geography, language support, regulatory limits, or core product fit to restrict the TAM.
• Defensibility Tip: The SAM represents the slice of the market your current product and go-to-market strategy can actually reach without major pivoting.

4. Ground the SOM (Serviceable Obtainable Market) in Sales Capacity

• The Mistake to Avoid: Never derive SOM as a random percentage of SAM (e.g., "we will capture 1% of a $10B market").
• The Defensible Way: Calculate SOM based on your near-term (2 to 3 year) sales and marketing engine capacity. Multiply your expected number of acquired accounts (derived from sales headcount, average quota, or ramped AE productivity) by your ACV.

5. Triangulate with Top-Down Data

• Sanity Check: Look at top-down macro reports for your space. If your bottom-up TAM calculates to $5B and macro reports say the industry is worth $6B, your logic converges. If they are off by orders of magnitude, your core assumptions need auditing.

If you'd like to share your industry/product type and target customer profile , I can help you outline the specific math and formulas for your model.

The most defensible TAM/SAM/SOM isn't the one with the biggest numbers—it's the one where someone can inspect every assumption and reproduce the math. Investors and executives are usually evaluating your reasoning more than your final figure.

Here's the approach that tends to hold up best.

1. Start bottom-up

Instead of saying:

"The CRM market is $80B."

Say:

• 180,000 target companies
• Average contract value = $9,000/year

TAM = 180,000 × $9,000 = $1.62B

Now every number can be challenged and updated.

2. Build SAM using observable constraints

Don't apply an arbitrary percentage.

Instead ask:

• Which industries?
• Which countries?
• Which company sizes?
• Which compliance requirements?
• Which languages?
• Which product limitations?

Example:

180,000 companies

↓

US + Canada only = 70,000

↓

100–1000 employees = 18,000

↓

Need Salesforce integration = 9,500

SAM = 9,500 × $9,000 = $85.5M

Every reduction has a business reason.

3. Build SOM from execution

This is where many decks become less credible.

Avoid:

"We'll capture 5%."

Instead:

Sales team in Year 3:

• 8 AEs
• 30 new customers per AE/year
• 240 customers
• $9k ARR/customer

SOM:

240 × $9k = $2.16M ARR

Or if you want a market-share view:

240 customers ÷ 9,500 addressable customers = 2.5%

That percentage emerges from your operating model instead of being assumed.

4. Cross-check with top-down data

After building bottom-up, compare it with:

• Census or government statistics
• Industry associations
• Public company filings
• Analyst reports

If your bottom-up TAM is $1.6B but every reputable source estimates the market at $150B, investigate the gap:

• Are you defining the category differently?
• Is your pricing assumption too low?
• Are you excluding adjacent buyers?

The goal isn't identical numbers—it's an explanation for any differences.

5. Show assumptions explicitly

A strong TAM slide often includes a simple table:

If someone asks, "Where did that number come from?" you should be able to answer every row.

Common mistakes

• Starting with a giant Gartner or IDC number and claiming a small percentage.
• Applying arbitrary filters like "20% of SMBs."
• Assuming a market share with no operational justification.
• Ignoring pricing when calculating TAM.
• Using different customer definitions across TAM, SAM, and SOM.

A good rule of thumb

If an investor erased every number from your slide, could they recreate it from your assumptions?

If the answer is yes, your market sizing is likely defensible.

For B2B SaaS, I often recommend including both:

• A bottom-up model (customer count × ACV), which is usually the primary figure.
• A top-down validation from industry or government data to demonstrate that your estimate is directionally consistent. This combination is generally more persuasive than relying on either approach alone.