How to Estimate Market Size for a New Business

A founder finds a report saying their industry will be worth $20 billion within five years. The number goes into a pitch deck, usually next to a chart pointing sharply upward.

It may be a legitimate statistic. It may also include countries the business cannot serve, enterprise customers the product was not designed for, and several product categories that have little to do with the actual offer.

The figure describes an industry. It does not yet describe the opportunity available to that particular business.

A useful market-size estimate gets closer to practical questions: How many suitable customers exist? What could they reasonably spend? Which portion can the business serve now? How many of those customers could it realistically reach and win?

This piece explains how to estimate market size using TAM, SAM, and SOM, compare top-down and bottom-up methods, find useful data, and build a calculation whose assumptions can be understood and challenged.

What Market Size Is Supposed to Tell You

Market sizing is an estimate of the commercial opportunity available within a defined group of customers.

Businesses use it to compare segments, assess whether an opportunity is large enough to pursue, plan sales capacity, discuss growth with investors, and decide where further research is needed. UK government investor-readiness guidance for businesses in the space sector, for example, recommends presenting TAM, SAM, and SOM with clear evidence and assumptions rather than relying on unsupported headline figures.

That does not make market size a forecast of what the business will earn.

A market can be large while demand for one specific offer remains weak. Customers may already have satisfactory alternatives. They may be difficult to reach, slow to buy, or unwilling to pay enough to support the business model.

Market size is one part of an opportunity assessment. It should be considered alongside evidence that the problem matters, customer willingness to change, competitive alternatives, achievable pricing, acquisition costs, and the team's ability to serve the market.

The estimate becomes valuable when it exposes these assumptions instead of hiding them behind one impressive number.

TAM, SAM, and SOM Explained

TAM, SAM, and SOM describe three progressively narrower views of a market.

What is TAM?

Total Addressable Market, or TAM, is the total annual revenue opportunity for an offer if every relevant potential customer bought it.

For a software product, a simple starting calculation might be:

Number of potential customers × annual revenue per customer = TAM

If 50,000 organizations could use a product and each would pay an average of $1,000 per year, the estimated TAM would be $50 million in annual revenue.

TAM is theoretical. It helps establish the scale of the broad opportunity, but it ignores many immediate constraints. A company will not win every customer, serve every geography, or satisfy every use case included in that number.

It is still useful when the market boundary is clearly defined.

"The global software industry" would be far too broad for a small workflow product. "Independent accounting firms with 5 to 50 employees that manage recurring client document requests" is much closer to a market that can be examined.

What is SAM?

Serviceable Available Market, or SAM, is the portion of the TAM that the current offer and business model can actually serve.

The reduction from TAM to SAM may reflect:

supported countries and languages
customer size
industry focus
regulation
product capabilities
available integrations
delivery capacity
pricing model

Imagine a product that could eventually serve 50,000 businesses worldwide. The current version is available only in English, supports three accounting platforms, and is designed for firms with fewer than 50 employees.

After those limits are applied, perhaps 8,000 businesses remain. Those 8,000 potential customers form a more relevant SAM.

The filters should come from real constraints. Reducing TAM by an arbitrary percentage does not explain which customers are included or why.

What is SOM?

Serviceable Obtainable Market, or SOM, is the part of the SAM that the business could realistically capture within a stated period.

This is where market sizing becomes closely connected to execution.

SOM should consider the channels available to reach customers, sales capacity, expected conversion rates, competitive pressure, onboarding limits, retention, and the period covered by the estimate.

A common shortcut is to claim that the business will capture 1% of a large market. The percentage appears conservative, but it says very little.

One percent could still require thousands of customers, a large sales team, extensive capital, or a distribution channel the business does not have.

A stronger SOM estimate begins with how customers might actually be acquired.

Top-Down and Bottom-Up Market Sizing

There are two common ways to approach a market-size calculation. Using both can provide a useful check.

Top-down market sizing

A top-down estimate begins with a large market figure and narrows it using relevant filters.

Suppose an industry report estimates that European companies spend €5 billion per year on a broad category of business software. You might then estimate:

the share spent by small businesses
the share connected to your product category
the share within the countries you serve
the share represented by your target customer type

The result provides a broad estimate of the available market.

Top-down calculations are useful when credible industry data exists and you need an initial sense of scale. They can also help compare several markets quickly.

Their weakness is the distance between the source category and the offer being evaluated. A report about "business productivity software" may include project management, communications, document storage, analytics, and other products.

Applying several percentages to that total can create a neat calculation built on a poor match. Every filter introduces another assumption, and those assumptions should remain visible.

Bottom-up market sizing

A bottom-up estimate begins with the customers the business could serve.

The basic calculation is:

Number of suitable customers × realistic annual revenue per customer

For example:

4,000 suitable agencies × $1,200 annual revenue = $4.8 million annual market

This approach is often more practical for an early business because each part can be investigated.

You can ask:

Are there really 4,000 suitable agencies?
Does the count include inactive companies?
How many match the required size?
Is $1,200 consistent with the value created and current alternatives?
Would every suitable customer need one account or several?
Does the estimate include customers outside the supported region?

A bottom-up calculation may produce a smaller number than an industry report. That is not a weakness when the logic is more relevant to the actual business.

A Worked Market-Sizing Example

Consider a fictional SaaS product that helps small agencies transfer project information from sales to delivery.

The product collects signed scope, client expectations, deadlines, dependencies, and known delivery risks in a structured handover. It costs $100 per month, or $1,200 per year.

The company initially serves English-speaking agencies with 5 to 30 employees.

All numbers below are fictional and included only to demonstrate the process.

Step 1: Define the broad customer group

The founder estimates that there are 40,000 agencies across the wider set of countries the product could eventually support.

At $1,200 in annual revenue per customer:

40,000 agencies × $1,200 = $48 million TAM

The $48 million figure describes the broad long-term opportunity under the current pricing assumption.

It does not mean all 40,000 agencies experience the problem, want dedicated software, or would select this product.

Step 2: Apply the current service constraints

The company can currently sell and support customers in four countries. It also focuses on agencies with 5 to 30 employees.

Very small firms often handle handovers informally. Larger firms may require deeper integrations, permissions, and procurement processes than the current product supports.

After applying geography and company-size filters, the founder identifies approximately 9,000 agencies.

Further research suggests that around two-thirds regularly separate sales and delivery responsibilities. The others use the same person or small team across the client engagement.

9,000 agencies × 67% = approximately 6,000 suitable agencies

At the current annual price:

6,000 agencies × $1,200 = $7.2 million SAM

The 67% assumption needs evidence. The founder might test it through customer interviews, agency directories, job roles, process descriptions, and a sample of company websites.

A range may be more honest at first. If the proportion could reasonably sit between 55% and 70%, the estimated SAM would range from about $5.9 million to $7.6 million.

That range tells the reader more than a falsely precise answer.

Step 3: Estimate realistic acquisition capacity

The founder plans to acquire customers through targeted outbound, agency communities, and partnerships with consultants who improve agency operations.

Based on early tests, the business believes it could create around 300 qualified sales opportunities per year once the process is established. A 20% conversion rate would result in 60 new customers annually.

The estimate also assumes annual customer retention of 85%.

After three years, the business might have roughly 150 active customers, allowing for customers acquired in different years and some churn.

150 active customers × $1,200 = $180,000 annual SOM after three years

The SOM is far smaller than the $7.2 million SAM. It is also connected to a real acquisition model.

The founder can now examine the assumptions that matter:

Can the business consistently create 300 qualified opportunities?
Is a 20% conversion rate supported by early sales?
Will customers remain for more than one year?
Can the team onboard and support 150 accounts?
Could partnerships increase reach without raising acquisition costs too far?

As evidence improves, the SOM can be revised.

Where to Find Reliable Market Data

The quality of a market estimate depends partly on the data used to build it. No single source usually provides the finished answer.

The U.S. Small Business Administration's market research guidance recommends examining factors such as demand, market size, customer location, market saturation, and pricing.

Those are useful research questions. A founder still has to translate the available information into a market definition that matches the offer.

National statistics and business registers

Official sources can provide counts by industry, geography, employment size, revenue, or legal status.

In the United States, the Census Bureau's County Business Patterns provides annual data on establishments with paid employees, including industry, location, payroll, and employment size.

This can help estimate the number of organizations in a defined category, although it does not include every self-employed or non-employer business.

For European markets, Eurostat's Structural Business Statistics provides data on business activity and performance across countries, industries, and enterprise-size classes.

Industry classification codes are useful, but they rarely match a product's target market perfectly. A category such as "advertising agencies" may include businesses with very different services, team structures, and buying needs.

Use the official count as a starting point, then apply evidence-based filters.

Industry associations and regulators

Industry bodies may publish membership figures, annual reports, workforce data, licensing records, or surveys.

These sources can be particularly useful in regulated or specialized markets. A register of licensed clinics, schools, financial advisers, or construction firms may provide a more relevant base than a broad industry estimate.

Check what the data excludes. An association's membership list may cover only paying members rather than the entire market.

Company directories and marketplaces

Directories can help estimate how many businesses visibly operate within a niche. Marketplaces may also reveal service categories, customer activity, prices, review volume, and geographic concentration.

The data may contain duplicates, outdated profiles, and businesses that do not match the target segment. Sampling is usually more useful than treating the full directory count as fact.

For instance, reviewing 150 randomly selected agency profiles may show that only half have the size and service model relevant to the product. That proportion can then be applied cautiously to a broader count.

Your own commercial data

Once a business begins speaking with customers, internal information becomes increasingly important.

Sales calls reveal which companies experience the problem. Conversion rates show which segments are willing to act. Customer records provide evidence about pricing, retention, and account value.

An external database may estimate how many companies exist. Your own data helps determine how many behave like plausible customers.

How to Estimate a Market When Data Is Incomplete

Market data is often incomplete, especially for new categories or narrow customer groups.

A product may serve "operations consultants working with agencies," while official statistics classify consultants under a much broader category. No public database will provide the finished number.

In that situation, build the estimate in layers.

Start with the strongest available base figure. This could be the number of businesses in a related industry, the number of licensed professionals, or the number of organizations within a certain size band.

Then document each filter and its source.

For example:

Relevant businesses

Low estimate: 8,000
Base estimate: 10,000
High estimate: 12,000
Evidence: Official industry count

Businesses matching the target size

Low estimate: 35%
Base estimate: 45%
High estimate: 55%
Evidence: Sample of company records

Businesses using the relevant workflow

Low estimate: 50%
Base estimate: 65%
High estimate: 75%
Evidence: Interviews and role analysis

Annual revenue per customer

Low estimate: $900
Base estimate: $1,200
High estimate: $1,500
Evidence: Pricing tests and comparable alternatives

This creates a range rather than one rigid answer.

It also shows which assumption has the greatest effect. If the result changes dramatically depending on how many businesses use the target workflow, that question deserves more research.

A good model makes uncertainty easier to manage. It does not pretend the uncertainty has disappeared.

Common Market-Sizing Errors

Some errors produce visibly unrealistic numbers. Others are harder to notice because the spreadsheet still looks professional.

Using an entire industry as the TAM

A product serving one workflow inside the construction industry does not address all construction spending.

The relevant market should reflect the offer, customer, and buying unit. Broad industry revenue can provide context, but it should not automatically become the business's TAM.

Confusing users with paying customers

A collaborative tool may have ten users inside one company, while the company purchases one account.

Multiplying ten users by the full subscription price would overstate the opportunity unless pricing is genuinely per user.

The opposite can also occur. A platform might have several paid teams or locations within one organization. The calculation should reflect how purchasing actually works.

Applying unsupported percentages

Statements such as "we only need 1% of the market" avoid the difficult part of the analysis.

How many customers does 1% represent? Which channel reaches them? How many salespeople would be required? How long would acquisition take? What conversion rate is assumed?

A small percentage of a large market can still be operationally unrealistic.

Mixing incompatible sources

One source may count registered companies, another individual professionals, and a third total industry revenue. They may refer to different years, countries, definitions, and currencies.

Before combining them, check:

what is being counted
the period covered
whether the figure represents businesses, establishments, users, or transactions
which organizations are excluded
whether the data is nominal or adjusted
whether categories overlap

A calculation can be mathematically correct and conceptually wrong.

Treating precision as confidence

An estimate of 18,427 customers looks more authoritative than a range of 15,000 to 22,000.

Unless the underlying data supports that level of accuracy, the precise number is mostly formatting.

Round figures appropriately and explain the assumptions that create the range.

Market Size and Market Reach Are Different Questions

A market may contain thousands of suitable customers and still be difficult to enter.

The buyers may not search for the problem. Decision-makers may be hard to identify. Contracts may require lengthy procurement. The expected account value may be too low to support direct sales. Customers could be spread across several languages, regulations, and regions.

Suppose two segments each contain 5,000 suitable businesses.

In the first segment, the companies belong to a small number of associations, use similar software, and actively search for the problem. In the second, the companies are fragmented across local markets and rarely recognize the problem until someone explains it.

The market counts are equal. The route to customers is not.

This distinction matters most when estimating SOM. The number should be connected to a plausible acquisition model rather than treated as a fixed percentage of the broader market.

The VynaroAI guide to choosing the right customer acquisition channel explains how customer behavior, account value, trust, and available resources affect which part of a market can realistically be reached.

Turn the Estimate Into a Research Plan

The first market-size model will contain uncertain assumptions. That is expected.

Mark each assumption according to the quality of its evidence.

Strong evidence might include official statistics, customer records, signed contracts, or several consistent sources.

Moderate evidence could include a relevant industry survey, a carefully reviewed sample, or repeated customer interviews.

Weak evidence includes broad analogies, small samples, or assumptions borrowed from unrelated businesses.

Then ask which weak assumption has the greatest effect on the result.

For an agency software product, the total number of agencies may already be well documented. The real uncertainty could be the proportion with a separate sales-to-delivery handover. Ten well-selected interviews may improve the estimate more than another industry report.

For a consulting service, the customer count might be clear while the realistic annual spend remains uncertain. Testing several offers could provide the missing evidence.

Market sizing becomes more useful when it directs the next research step.

A Credible Estimate Is More Useful Than a Large One

TAM, SAM, and SOM can help a business understand the scale of an opportunity, but the three labels do not make an estimate credible by themselves.

Define the customer carefully. Use top-down data for context and bottom-up calculations for practical relevance. Show the filters between the broad market and the segment you can serve. Connect the obtainable market to real acquisition and delivery capacity.

Then keep updating the model as customer conversations, pricing tests, sales results, and retention provide better evidence.

A large number may look impressive on a slide. A transparent estimate is more useful when you need to decide what to build, which market to enter, and how much growth is realistically available.

Ready to examine your opportunity more systematically? Explore VynaroAI's structured business tools to work through your market, customer, positioning, and business assumptions.