पाठशाला Pathshala · विचार Vichār, The idea · Lesson 04 · Start

Market sizing an investor will believe

A market is a count of people who could pay, multiplied by what one pays a year. The rest is a claim about how many you can reach and win, and investors read that claim, not the big number.

Pathshala, The Founder Library · 10 October 2026 · 9 min read

Every investor has sat through the slide: a consultancy’s number in the tens of thousands of crores, a one per cent share pencilled under it, and a founder who could not say where either figure came from. The slide is not wrong because the market is small. It is wrong because nothing on it can be checked, and an investor who cannot check a number discounts it to zero.

This lesson gives you the other way of doing it. It takes about an afternoon, it produces a number that is usually smaller and always more persuasive, and it does something more useful than filling a slide: it tells you what you would have to believe for the company to be large, and which of those beliefs to go and test first.

Why the big number does not work

Top-down sizing starts with a total and takes a slice. Indian healthcare is so many lakh crore; digital health is some fraction; we will take a per cent of that. Each step is a guess about a guess, and the final figure depends most on the step nobody can defend, the share. Michael Seibel, writing for Y Combinator on how to pitch a company, puts the problem plainly: the top-down pitfall is not narrowing the customer enough. A number that includes everyone includes nobody in particular, and a product is bought by someone in particular.

Bottom-up sizing starts with the buyer. How many of them are there, counted from a source you can name? How many of those can you actually reach with the product you have, through the channel you have, in the languages and cities you can serve? Of those, how many can you win in a sensible period against whatever they do today? And what does one of them pay in a year? Four factors, each a single defensible sentence. The market is their product.

TAM, SAM and SOM done honestly

The three acronyms are useful only if they are nested figures from the same arithmetic, not three numbers from three reports. Total addressable market is every buyer who could use the product multiplied by the annual price: the ceiling if you served everyone. Serviceable addressable market is the part of that you can reach with the business as it will exist in the planning window: your geography, your language, your channel, the segment your product actually fits. Serviceable obtainable market is the share of that you can win in about five years, which is the number the business plan has to be built on.

Three rules keep it honest. Annualise everything: a product bought once every two years counts half its buyers each year. Use the revenue you keep, not the money that passes through: a marketplace sizes on its take, a lender on its fees, not on gross transaction value. And resist the round number for share. A rule of thumb investors repeat is that a leader in a fragmented market holds five to fifteen per cent after a decade, and most companies hold far less than that at the point they list. Ten per cent in five years is a claim, not a default.

An investor does not fund the TAM. They fund the SOM and the believability of the route to it.

The sequence of defensible factors

Start with the count, because the count is the only factor you can source rather than argue. India publishes more of these than founders expect. The Ministry of MSME reports over 7.83 crore enterprises registered on Udyam as of February 2026, with state-wise tables, which is the starting point for anything sold to small business. The Health Facility Registry under the Ayushman Bharat Digital Mission had about 2.35 lakh verified facilities by early 2024, of which 69,633 were private. The Household Consumption Expenditure Survey tells you what an urban Indian spends a month and on what. The Census and the population projections tell you how many of them there are. When the count you need does not exist, say so on the slide and derive it from two that do, showing the step.

Then reach, which is a statement about your company rather than the market: the cities you will operate in, the languages the product speaks, the channel that gets it in front of the buyer. Then share, which is the factor you will be challenged on, so argue it from analogues: what share did the leader take in a comparable category, over what period, with what advantage. Then price, which should come from customers you have actually quoted, not from what you hope to charge once the product is finished. Set the four sliders below from your own numbers and read the three discs.

Worked example one: software for Indian clinics

A founder is building practice-management software for small private clinics: appointments, records, billing, at ₹3,000 a month, so ₹36,000 a year. Count. The Health Facility Registry had about 2.35 lakh verified facilities in February 2024, 69,633 of them private. Registration is voluntary, so the private figure is a floor rather than a census of India’s clinics; the founder should say so and treat the gap as an assumption to be tested, not a number to be inflated. Reach. Government facilities do not buy this product, so the serviceable market is the private share, about 30 per cent of the count. Share. The category is fragmented and switching is painful, so the founder claims 10 per cent of reachable clinics in five years and defends it against what the current leaders hold. Price. ₹36,000 a year, taken from signed pilots.

Set the tool to 2,35,000 buyers, 30 per cent reach, 10 per cent share and ₹36,000 a year. TAM is about ₹846 crore. SAM is about ₹254 crore. SOM is about ₹25 crore a year from roughly 7,000 paying clinics. That is a good business. It is not, on this arithmetic, a venture business, and the sizing says exactly why: the count is modest and the price is low. The founder now has two honest routes. Widen the count, by showing with a source that the number of private clinics is far larger than the registry, or deepen the price, by adding payments, pharmacy procurement or insurance claims so that a clinic pays ₹1.5 lakh a year rather than ₹36,000. At that price the same 7,000 clinics are a ₹105 crore SOM. The slide should show both numbers and name the one the next six months will test.

Worked example two: a D2C snack brand

A founder is launching a premium packaged snack sold online and through modern trade in large cities. The temptation is to quote the size of the Indian snacks market and take a per cent. The bottom-up version starts with what people actually spend. The Household Consumption Expenditure Survey 2023–24 puts average urban monthly spending at ₹6,996 a person, with 11.09 per cent of it on beverages, refreshments and processed food: about ₹776 a month, or ₹9,310 a year. Count. The Census found 37.7 crore urban Indians in 2011 and the National Commission on Population projects roughly 59 crore by 2036; about 50 crore today is an interpolation, and the slide should label it as one with both anchors shown. A premium snack is not bought across the whole distribution, so the buyer is defined as the top fifth of urban spenders: about 10 crore people, an assumption with a government anchor. Reach. Online delivery plus modern trade in the cities the brand will stock reaches perhaps 40 per cent of those people in the planning window. Share. Two per cent of a category where the leaders are national FMCG companies, defended with what recent D2C food brands reached in five years. Price. Not the ₹9,310 category wallet, which includes eating out and beverages, but the ₹150 a month a buyer might plausibly spend on this kind of snack: ₹1,800 a year, an assumption to be replaced by repeat-purchase data within a quarter of launch.

Set the tool to 1,00,00,000 buyers, 40 per cent reach, 2 per cent share and ₹1,800 a year. TAM is ₹18,000 crore. SAM is ₹7,200 crore. SOM is ₹144 crore a year from about 8 lakh repeat buyers. Every factor is either sourced or labelled, and an investor reading the slide can disagree with one of them specifically, which is a conversation, rather than with the whole slide generally, which is a rejection.

Markets that grow because the product exists

The bottom-up method has a known failure: it sizes the market as it is, and the best companies change what the market is. In 2014 a respected valuation professor sized Uber against the global taxi market at about $100 billion and gave it a tenth. Bill Gurley, an Uber board member, answered with How to Miss by a Mile: a product that is faster, safer and cheaper than a taxi is used by people who never took taxis, for trips that were never taxi trips, and the real comparison is the cost of owning a car. The count was right for the old product and wrong for the new one.

The lesson is not that founders may inflate the count. It is that if your argument is that the product expands the market, say so as a separate, explicit step, with the mechanism: lower price brings in these buyers, a new use case brings in those, and here is the early evidence. Show the market as it is, then the market as the product makes it, and let the investor see the join. Hiding the expansion inside the TAM is how a defensible slide becomes an indefensible one.

How to put it in the deck

One slide. A table with four rows, one per factor: the number, where it came from, and how confident you are. Below it the three figures in rupees and the SOM also as a count of customers, because a founder who says eight lakh repeat buyers or seven thousand clinics has made a promise that can be checked against next year’s dashboard, and a founder who says ₹144 crore has not. Then one line naming the assumption you are least sure of and how you will test it in the next ninety days. Seibel’s advice on pitching applies to the whole slide: make it sound simpler than you think it should, use a concrete example with specific numbers, and keep the big number off the slide if you cannot defend the route to it.

Expect the questions in this order. Where did the count come from. Why that reach. Who else holds that share, and how long did it take them. What did the last customer actually pay. If the slide answers all four before they are asked, the market conversation ends in five minutes and the meeting moves to the part that decides it, which is you.

Before the deck goes out, run the sizing again from scratch in the tool above without looking at the old numbers, and see whether you land within twenty per cent of yourself; if not, one of the factors is a hope. Check that every count has a URL you could paste into the room. Check that price is a number a customer has said yes to. Check that the SOM is stated as customers as well as rupees. Then write the market expansion argument, if you have one, as a separate paragraph in the notes, and decide whether it belongs on the slide or in your mouth. Repeat the exercise every six months, because the one thing the tool cannot do is notice that the count has changed.

The figures in the worked examples are illustrations built on public data and clearly labelled assumptions, not forecasts for any company. The sources are below; the two government documents are worth an hour each.