The numbers Brief Episode map Timeline Takeaways Charts Quotes Clips Glossary Go next Transcript

Episode 71 · Consumer · 33 min

Six cities, and everything in them

NoBroker's arithmetic is deliberately unfashionable: six cities at 25% market share beat fifty cities at 1%, and 20 million customers beat a shallow hundred million. The claim underneath it is that a broker taking a month's rent from the owner and a month from the tenant (18% of an 11-month lease) was charging a bomb for an introduction, and that the money saved belongs in going deeper, not wider.

Real estate & proptech Founder journey Product strategy

Akhil Gupta

Co-founder & CTO, NoBroker · with Vishal Krishna

33:09

Said in this episode

▶ 3:08

18%

Broker's cut of an 11-month rental contract

One month's rent from the owner plus one from the tenant on an 11-month lease; the equivalent charge on a buy-sell transaction is 4-6%.

▶ 12:56

20 million

Customers served across six cities

Gupta's deliberate alternative to chasing 100 million shallow users. Fewer customers, but every housing need met on one platform.

▶ 14:26

₹1,000-3,000

Price of a specialist service on a free platform

The platform itself stays free after a first year to eighteen months of charging nothing at all; revenue comes only when a customer needs something extra.

▶ 14:46

2017

First year of city-level profitability

Profitability was reached city by city rather than company-wide, and the earnings were then deliberately reinvested into new cities and new business lines.

▶ 7:49

The entire product and engineering team in 2016-17

Product, engineering, design and testing put together, at the point when the multi-line platform was being built.

▶ 10:40

100+

Freshers hired off campus in seven years

From IITs and other campuses in every year but one; many are still with the company or have moved on to senior roles elsewhere.

The brief

The argument in sixty seconds

Gupta's claim is that real estate was the last big Indian market where an intermediary who added almost nothing still took a bomb: one month's rent from the owner and one from the tenant, 18% of an 11-month contract, or 4-6% of a sale, and that by late 2012 and 2013, with MakeMyTrip having done travel, BookMyShow entertainment and Flipkart clothing, its turn was overdue. What made the bet frightening was that no comparable platform existed anywhere in the world; what made it worth taking was that same fact. The second, less-quoted claim is about restraint. A platform scales infinitely and housing demand does not: you cannot discount somebody into moving house the way a food app discounts you into a second pizza. So NoBroker took six cities and roughly 20 million customers over fifty cities and a hundred million, on the arithmetic that six cities at 25% share beat fifty at 1%, and spent the difference stacking rental agreements, packers and movers, cleaning and painting, rent on credit card, loans, insurance and society management on top, an Amazon for real estate rather than a wider listings board. The platform stayed free; specialist services cost ₹1,000-3,000; cities turned profitable from 2017 and the profit went back into new cities and new lines rather than into the P&L. Behind all of it sits a chemical engineer who learned to code at Oracle, hires a CA if the aptitude is right, and once watched 50-60 brokers come to shut his office down, and 100 employees turn up the next morning anyway.

Worth your time if you are

Marketplace founders trying to disintermediate an entrenched broker

Growth leaders under pressure to add cities they cannot serve

Engineering leaders hiring for aptitude rather than pedigree

Renters who have handed two months' rent to a stranger

Episode map

Where the conversation travels

Every block is a chapter, coloured by what it's about. Click any of it to jump straight to that minute on YouTube.

Real estate & proptechFounder journeyProduct strategyHiring & talentUnit economicsLeadership & org

01Cold open: a valuation, not a windfall 0:00 Vishal opens the unicorn series and Gupta immediately deflates the tag. People assume the founders earned that much money when it is only the company's valuation, and the one concrete benefit is that recruiters no longer have to explain what NoBroker is. 02Roti, kapda, makaan, and the missing platform 1:17 The 2012-13 genesis: technology had eliminated the intermediary in travel, entertainment and clothing but not in the single biggest market, where a human who adds little to the transaction still takes a month's rent from each side of an 11-month lease or 4-6% of a sale, and a global search found no platform doing without him. 03Build the first version in what you know 4:55 The first product did one thing, connect owners and tenants, and was written in Java, Spring and NoSQL because a founder chasing product-market fit should not add moving parts he doesn't understand; microservices, heavy caching, search and multi-cloud came only afterwards. 04Referrals first, then responsibility 6:29 The early team came almost entirely from people who had worked with, studied with or lived around the founders (a family more than a hiring funnel), and the thing Gupta credits is handing them responsibility, back when product, engineering, design and testing together were 25 people in 2016-17. 05Each complaint became a business line 8:05 Customers who had found a house then asked who would do the rental agreement, the move, the cleaning and painting, the rent payment, the sale deed, the loan, the insurance and the society management, and each answer became a line that Gupta argues could be a unicorn on its own, which also kept engineers with new things to build. 06Degrees don't matter, aptitude does 9:39 A chemical engineer who learned to code inside Oracle's training programme now hires a CA and a B.Com bootcamp graduate to write code, invests three to six months in them, and has taken 100-plus freshers off IIT and other campuses in every year but one of the last seven. 07Why a scalable platform stayed in six cities 10:56 Scaling the technology and the operations was never the constraint. Housing demand is: nobody moves house for 20% off the way they order a second pizza for 50% off, so NoBroker chose to be penetrated deeply in six cities with ten new business lines and about 20 million customers rather than thin across fifty. 08Free platform, paid extras, profitable from 2017 13:26 After a free first year to eighteen months the model became charge only for specialist services at ₹1,000-3,000, cities turned profitable from 2017, and the founders then chose to reinvest that profit into new cities and lines, because the barometer is customers served, and six cities at 25% share beats fifty at 1%. 09Registries, migration and 600 million urban Indians 16:46 Market entry is sized off the government registry of property transactions rather than any organised competitor data, against a projection of 600 million Indians in urban centres by 2030 that six or eight cities cannot possibly absorb, which is why Jaipur, Indore and their peers, boosted by work-from-home, are the next set. 10Product company is just a fancy tag 18:58 Gupta refuses the product-versus-services label (a product nobody uses is not a product) and argues the only test is whether it makes a customer's life easier, automates something, saves money or adds convenience; the same logic makes partners central, since packers, cleaners and painters had no route to demand beyond Yellow Pages. 11Cloning yourself instead of coding everything 20:40 The founder who once had no time to talk to anyone because he was coding now describes his job as making more Akhils, Amits and Saurabhs, handing over ownership, and passing on the patterns behind the bad experiences so the team does not have to repeat them. 12The unicorn tag does not open floodgates 22:21 Money in the bank signals stability but does not by itself attract engineers (young people come for the problem), and hiring stays hard because too many startups with too much money give good candidates too many choices while most of India's engineering graduates are not employable. 13Two days in office, three at home 24:05 Watching 300-400 people back on one floor, Gupta argues hybrid is permanent: two hours of Bombay or Bangalore commuting is pure loss and remote processes now work, but brainstorming and human contact still pull people in a couple of days a week. 14The night fifty brokers came 25:25 Fifty to sixty brokers arrived to thrash the team, the office was locked within hours and the police would not let them work from it, and the next morning all 80 to 100 employees showed up at two improvised locations, which Gupta reads in hindsight as proof the disruption was real. 15Quitting with a baby on the way 28:20 He left his job while his wife was pregnant with their first child and his business family backed him without understanding what he was building; the wider argument is that startups compress three to five years of a big company's work into one, and that the founders of 2015 and of 2020 are visibly different people. 16Still water, running water, and two kids 30:15 Not a reader, Gupta says problems are the motivation. If January's problem is still June's problem your life is stagnant. He advises young engineers to bring solutions rather than complaints, and closes on the daughter who is exactly as old as NoBroker and the son whose childhood he finally got to watch.

Timeline

The arc, briefly

Eight years of removing the broker, and the decision to go deep in six cities instead of thin in fifty.

2007-08Saurabh registers the domain. nobroker.in was booked years before there was anything to put on it.

Late 2012The one intermediary technology had not touched. MakeMyTrip had done travel, BookMyShow entertainment, Flipkart and Myntra clothing. Real estate, the biggest market of the lot, still ran through a person who added little and charged a bomb: one month's rent from the owner and one from the tenant, 18% of an 11-month contract, or 4 to 6% on a sale. They looked across the globe and found no platform working without a broker, which was frightening and was also the reason to build one.

2014NoBroker starts. Vishal met Amit and Saurabh that year while they were pitching it. Gupta, a chemical engineer from IIT Bombay who learned to code inside Oracle's training programme, wrote the first version in Java and Spring on NoSQL because that was what he already knew and a founder chasing product-market fit should not add parts he does not understand. He quit with his first child on the way, and his daughter is exactly as old as the company.

2014-15Free, and then the night the brokers came. For the first year to eighteen months the platform made no money at all. In those early years fifty to sixty brokers arrived at the office to thrash the team, the office was locked within hours and the police would not let anyone work from it. All 80 to 100 employees turned up the next morning, split across two improvised locations.

2016-17Twenty-five people carry the whole product. Engineering, design and testing included. Hiring ran on referrals and aptitude rather than degrees, with a chartered accountant and a B.Com graduate out of a bootcamp writing code, and three to six months invested in someone before anyone expected output.

2017Cities start turning profitable, Bangalore first. The platform stays free and a specialist service costs ₹1,000 to 3,000. Gupta says they could harvest Bangalore and Mumbai, raise nothing and hire nobody, and have decided instead to put every rupee earned into new cities and new lines.

TodayTwenty million customers, six cities, ten lines. Each unanswered customer question became a business: rental agreements, packers and movers, cleaning and painting, rent paid by credit card, sale deeds, home loans, insurance, society management. Six cities at 25% share beats fifty cities at 1%, and you cannot discount somebody into moving house the way a food app sells a second pizza. The unicorn tag mostly means recruiters no longer have to explain who NoBroker is.

2030600 million Indians in urban centres. Six or eight cities cannot absorb that, so Jaipur, Indore and their peers are the next set, helped along by the work-from-home shift that sent people back to tier two and tier three towns and kept them productive there.

Takeaways

Ideas to carry out of this hour

Real estate was the last intermediary standing

By late 2012 technology had removed the middleman from travel, ticketing, clothing and food, but the single biggest market still ran on a human who introduced two parties and then left them to negotiate alone. The price of that introduction is the argument: a month's rent from the owner and a month from the tenant, 18% of an 11-month contract, or 4-6% on a buy-sell, at rents that are already 20-40% of monthly income. When the founders looked worldwide for a platform that worked without a broker, they found none. Scary because it suggested it might be impossible, exciting because it meant the opportunity was still open.

▶ 2:21 ▶ 3:08 ▶ 3:24

The unicorn tag buys recruiting, not much else

Gupta's first move in the interview is to separate valuation from money: people assume the founders have earned that much when it is the company that has been valued. What actually changed is smaller and more practical: you no longer have to explain who NoBroker is when you go out to hire. He reads the tag mainly as a testimony to seven or eight years of building a business model he says had not been attempted anywhere in the world.

▶ 0:32 ▶ 0:48 ▶ 1:03

Depth beats coverage: six cities at 25% beats fifty at 1%

Scaling the technology was never the hard part and neither was scaling operations, since NoBroker has no feet on the ground. The binding constraint is that housing demand cannot be manufactured: you cannot offer somebody 20% off to change house, the way a food app can sell a pizza to someone who has already eaten. So the barometer became customers served rather than cities entered, and the arithmetic is blunt: six cities at 25% share is 150, fifty cities at 1% is 50. Prove the market once, and replication into the next set is faster and cheaper.

▶ 11:41 ▶ 12:26 ▶ 16:13

Every unanswered customer question became a business line

Customers who found a house on a free platform immediately asked who would write the rental agreement, and that question, repeated, built the stack: packers and movers, cleaning and painting, rent paid by credit card, encumbrance checks and sale deeds, home loans, home insurance, and finally society management with its visitors and finances. Gupta's claim is that several of these are unicorn-scale businesses in their own right, and that being the Amazon for real estate to 20 million customers is worth more than being a listings board to a hundred million.

▶ 8:05 ▶ 8:21 ▶ 13:09

Stay free, charge for the specialist service

For the first year to eighteen months NoBroker made no money at all; the model that followed keeps the core platform free and charges ₹1,000-3,000 only when someone needs a specialised service on top. That was enough for city-level profitability from 2017, and Gupta is explicit that the company could simply harvest its established cities, raise nothing and hire nobody. The deliberate choice instead is to route every rupee earned into new cities and new lines, because the question is whether you want short-lived success or a business that outlasts you.

▶ 14:26 ▶ 14:46 ▶ 14:55

Build the first version in the stack you already know

NoBroker started on Java and Spring with NoSQL for an unglamorous reason: that is what Gupta had used at Oracle and in his previous startup. His rule is to minimise moving parts. Building the business, the website and the product is already an enormous thing, and you cannot do it in a language you are still learning. Speed to a testable product mattered more than architecture; microservices, caching layers, search work and multi-cloud all arrived later, once there was something worth re-engineering.

▶ 5:11 ▶ 5:27 ▶ 5:59

Hire for aptitude, then spend six months on the person

Gupta is a chemical engineer from IIT Bombay who knew nothing about coding until Oracle's training programme and its mentors taught him, and he has generalised that experience into hiring policy: the degree does not matter. The company has taken on a chartered accountant and a B.Com graduate out of a bootcamp to write code, and invests three to six months in them before expecting output. Across seven years it has recruited 100-plus freshers from IIT and other campuses, missing only one year.

▶ 9:39 ▶ 9:53 ▶ 10:40

The night fifty brokers came to shut the office

The darkest moment was not a funding round that fell through, though those happened too. Fifty to sixty brokers came to the office to thrash the team; within hours the office was locked and the police would not let anyone work from it, leaving a company of 80 to 100 people without premises overnight. Every one of them showed up the next day and split across two improvised locations. That is the moment Gupta reads as proof both that the disruption was real and that self-belief plus the people around you is what carries a founder through.

▶ 26:52 ▶ 27:06 ▶ 27:22

The numbers, drawn

What the episode measures

Every figure below was said on air, with timestamps included and caveats kept.

What the middleman charged

% of transaction value

As stated in conversation: one month's rent from the owner and one from the tenant works out to 18% of an 11-month contract; buy-sell brokerage was described as 4-6%.▶ 3:08

The arithmetic behind staying small

cities × market share

Gupta's own arithmetic, said on air as the answer to pressure to expand. It is an illustrative comparison rather than reported market-share data.▶ 16:13

Every unanswered question became a line

what customers asked next, after they found the house

Gupta's claim is that several of these could be unicorn-scale on their own, and that the stack is what makes 20 million customers worth more than a hundred million shallow ones.▶ 8:05

Free, then paid, then spent again

where the money goes once a city works

The platform stays free; the fee only appears when somebody wants a specialist service on top. Gupta says the company could harvest its established cities instead, raise nothing and hire nobody.▶ 14:26

Worth keeping

Lines that stay

A lot of people think that we have earned this much money — but this is the company valuation which we have got.

▶ 0:32

You should not have a lot of moving parts. Building the business, building the website, building the product — that itself is an extremely big thing. You cannot do it in something which you don't know.

▶ 5:27

I cannot tell you I will give you 20% off on your movement, please change the house. It's not a food business where I've got an amazing pizza and I'll give you 50% off even when you have had dinner.

▶ 11:56

Do you want to win 50 cities and take one percent market share in each city, or be in six cities and have 25 percent? You should not spread yourself thin — you should be thick into one market.

▶ 16:13

If you don't have any problem at this point of time, that means your life is stagnant. It's like there's still water and there's running water.

▶ 30:30

Clips that travel

Short on time? Start here

Marketplace founders hunting an unbroken intermediary

Why nobody had killed the broker

The 2012-13 origin story, the 18% brokerage arithmetic, and why finding no such platform anywhere in the world was both the scare and the signal.

1:17 → 4:55 · 4 min ▶ Watch clip

Engineering leaders rethinking their hiring filter

Degrees don't matter, aptitude does

A chemical engineer explains why he hires a CA and a B.Com bootcamp graduate to code, and what the three-to-six-month investment buys.

9:30 → 10:56 · 1 min ▶ Watch clip

Growth leaders under pressure to add geographies

Six cities deep, not fifty cities thin

The strongest stretch: demand you cannot discount into existence, the 25%-of-six arithmetic, free-then-paid monetisation, and profit reinvested rather than banked.

10:56 → 16:46 · 6 min ▶ Watch clip

Founders disrupting an incumbent with something to lose

The night fifty brokers came

The office locked by the police, 100 people without a desk, and everyone turning up anyway the next morning.

25:25 → 28:20 · 3 min ▶ Watch clip

Young engineers looking for a working philosophy

Still water, running water

Audit your problems every three or four months, bring solutions instead of complaints, and the honest note about missing a daughter's first years.

30:15 → 32:43 · 2 min ▶ Watch clip

Glossary

The jargon, unpacked

Brokerage

The intermediary's cut on an Indian property transaction: customarily one month's rent from the owner and one from the tenant on an 11-month lease, or 4-6% of the value of a sale.

Product-market fit

The point at which a product demonstrably solves a problem the market will pay for; NoBroker's early technology choices were made purely to reach it faster.

City-level profitability

When an individual city's revenue covers the cost of running it, ahead of any company-wide profit. NoBroker first hit it in 2017.

Encumbrance

A legal or financial claim registered against a property; clearing it is part of the buy-side paperwork the platform moved online alongside sale deeds.

Empanelment

Getting onto a platform's approved-vendor list, and how small packers, cleaners and painters reach demand they would otherwise have chased through Yellow Pages.

Microservices

Splitting one large application into small independently deployable services; the shape NoBroker's original Java build evolved into as the product lines multiplied.

Society management

Software for running a gated apartment complex (visitor entry, finances and the rest) which NoBroker added once it was already serving the residents inside.

Unicorn

A private company valued at over a billion dollars; Gupta's caveat is that it describes the company's valuation, not money the founders have earned.

Connections

If this resonated, go here next

EP 78 · Fintech

Full transcript

The whole conversation, searchable

Auto-generated captions, lightly cleaned. Click a timestamp to open that moment on YouTube.