Market sizing

At its simplest, market sizing answers one question: if you sold this, how much could people or businesses buy in a year? You will rarely get an exact answer. You are building an estimate, from data where it exists and from assumptions you can explain where it does not.

Before a single number goes on the page, pin down three things.

What you are counting. Customers, units or money. "1.7 crore businesses", "17 crore school bags a year" and "₹12,000 crore a year" are all market sizes, but they answer different questions.

Over what period. Almost always one year, so that markets can be compared with each other.

Where the edges are. Which customers, which product and which geography. "Accounting software for GST-registered businesses in India" is a market you can size. "Software for businesses" is not, because nobody could build a number for it, let alone check one.

Skip this step and no amount of arithmetic will save you. Most weak market sizes are badly defined, not badly calculated.

It matters because a surprising number of decisions quietly depend on it.

A clever product in a small market is still a small business. So when a team is choosing between two ideas, sizing both is often the fastest way to see which one deserves the next six months. The same thinking decides which customer segment or city to go after first, and whether a revenue target is a sensible slice of the market or wishful thinking.

Investors ask about it too, although experienced ones care less about how big your number is than about how you got there. And in founder's office, consulting, product and strategy interviews, market sizing questions are one of the most common kinds of guesstimate, because they show how you reason when there is no data to lean on.

Most market sizing works in three layers, each one a smaller slice of the one before.

SAM is sometimes called the Serviceable Addressable market. The meaning is the same.

A simple way to remember the three: TAM is the ambition, SAM is the strategy and SOM is the plan. That is also why investors and interviewers tend to push hardest on SAM and SOM. Those are the layers where your assumptions about customers, competitors and your own ability to sell are out in the open.

There are three ways in. Good market sizing usually uses one to build the estimate and another to check it.

Value-based sizing sounds abstract, so here is a quick example. Say a clinic misses 10 appointments a month, at ₹500 each. That is ₹5,000 a month walking out the door. If a reminder tool brings back half of those patients, it creates ₹2,500 of value a month, and the clinic might happily pay a fraction of that. Multiply that price by the number of clinics with the same problem, and you have a market size, even though nobody talks about an "appointment reminder market" yet.

Whatever the market, the same seven steps hold.

Let's put that into practice with a B2B market.

We will size paid accounting and bookkeeping software for GST-registered businesses in India, measured in yearly subscription revenue.

Why GST-registered businesses, and not every business? This is where definitions start to matter. About 1.70 crore taxpayers are registered for GST (GSTN, August 2026). Yet about 8.84 crore enterprises are registered on the Udyam portals (Ministry of MSME, June 2026).

That is roughly five times as many, because the Udyam count includes a very large number of informal micro enterprises, such as very small shops and home-run businesses, most of which have no GST registration and little need for accounting software.

Starting from the GST base keeps us to businesses with genuine compliance needs. In an interview or a pitch, explaining this choice is worth as much as the number itself.

If every one of those businesses paid an assumed ₹6,000 a year, the market would be 1.70 crore × ₹6,000, or about ₹10,000 crore a year. That is the ceiling. Nobody gets all of it.

In reality, most GST registrations are tiny proprietorships that hand their books to an accountant and never open accounting software themselves.

Let's assume about 40% manage their own books, which gives 68 lakh businesses, and that half of those would pay for dedicated software. That leaves 34 lakh businesses, worth 34 lakh × ₹6,000, or about ₹2,000 crore a year.

A new product will not win all of that. It is up against established players and free tools.

Assume it can realistically capture about 3% of the SAM within three to five years. That is roughly 1 lakh paying businesses, and about ₹60 crore a year in recurring revenue.

Two quick checks, from two different angles.

The data. GSTN reports that about 8.56 lakh businesses generated e-invoices in June 2026. Those are larger firms that almost certainly use software already. Our SAM of 34 lakh is about four times that, which makes sense if plenty of smaller firms also pay for simpler tools.

The sales effort. Winning 1 lakh customers over three to five years means adding roughly 20,000 to 35,000 a year. For a product sold mostly online with a small sales team, that is ambitious but believable.

If you were the founder, the number to plan hiring and spending around is the ₹60 crore, not the ₹10,000 crore.

The gap between TAM and SAM is worth a second look too. The businesses that rely on accountants are not lost customers. They might be reachable through a different product, one built for the accountants who each manage the books of many clients.

Now a consumer market, where good public data does much of the work.

We will count new school bags bought each year for students in Classes 1 to 12 in India, in units and rupees, and leave out pre-primary children.

Here we are lucky, because the base number is published. UDISE+ 2024-25 records about 23.3 crore students in Classes 1 to 12, and roughly half of them study in government schools.

Government and private school families buy very differently, so that is the one split worth making.

• Government schools: 11.8 crore ÷ 2 = 5.9 crore bags × ₹300 = about ₹1,770 crore.
• Private and aided schools: 11.5 crore bags × ₹900 = about ₹10,350 crore.
• Total: about 17 crore bags and ₹12,000 crore a year.

Spread across 23.3 crore students, ₹12,000 crore works out to about ₹520 per student a year. For context, MoSPI's household survey puts average monthly spending per person at about ₹4,122 in rural India and ₹6,996 in urban India (2023-24). Spending well under one month's budget on a school bag each year feels about right.

The interesting part is where the money sits. Private schools have slightly fewer students than government schools, yet they account for most of the value. A new school-bag brand would probably start there. A company supplying state governments would be in a different business altogether, built around very low prices and large orders.

You do not have to invent every number. For most Indian markets, a solid starting figure already exists in official data. These are the sources worth bookmarking.

A few words of caution before you use them.

Definitions differ more than you would expect. GSTN and Udyam both count businesses, and one figure is about five times the other. Always check what a number actually counts.

Registrations are not the same as active users. A business registered once may have shut down since, and plenty of people hold more than one SIM.

The population base is older than it looks. India's last full count was in 2011. The houselisting phase of Census 2027 ran from April to September 2026, and population enumeration is scheduled for February 2027. Until those results are out, every current population figure is a projection.

One part of founder's office and strategy work surprises people: you will spend as much time checking other people's market sizes as building your own. A handful of questions catches most weak numbers.

• What exactly is the market? If a figure has no clear customer, product and geography, there is nothing to check.
• Was any of it built bottom-up? A TAM lifted straight from an industry report, with no count of real customers behind it, tells you very little.
• Is it leaning on "we only need 1%"? Claiming a small slice of a huge market skips the hard part, which is explaining how you win any customers at all.
• Is it revenue, or the value of everything sold? Marketplaces often quote the total value of goods sold through them, which is far larger than the revenue they keep.
• What is driving the growth? A high annual growth rate stretched over many years can turn a modest market into a huge one on paper. Ask what would actually make it grow.
• Does it end in a decision? A good market size leads somewhere. One that only ends in a big number is usually there to impress.

In an interview, market sizing becomes a guesstimate. You estimate the size of a market out loud, usually in five to ten minutes, with no data in front of you.

Everything above still applies, with a few adjustments. You build your own base numbers from rough anchors, such as India's population or household count, instead of looking them up. Speed and structure matter more than precision, so round your numbers, give each assumption a reason and check your answer from a second angle. And you must finish with a so what: what the number means for the business, such as which segment to go after first.

For the full interview method, read Guesstimates: how to solve them. To practise, try questions 2 and 10 in Guesstimate questions (with answers). Market sizing often turns up inside case interviews too.

Most mistakes are easy to avoid once you know to look for them.