Week 0: Starting Out? Choose an Idea that Scales!

I’ve spent about three weeks evaluating ideas on what I want to build. More details on what it is over the coming weeks! For now though, the first order of business is explaining how I got to the point of building this website.

The backstory:

As many of you already know, I’ve worked in political consultancy for a few years now. We’ve helped many parties across states with data analytics and campaign support. It’s great work and I find it super interesting. It’s pretty well paid too. There is an issue though – it can’t ever compound!

What does that mean? It means that unlike a business where you add one client on top of another, in political consultancy you churn through clients. Once an election is done, you’re off to a new state with a new party, pretty much starting all work from scratch. Even if they loved your work, they’ll only really require your services after a few years when the next election rolls around. I’ve now done this across 9 different states, all of which have their own unique dynamics, but the work on a day to day basis remains largely the same. You use the same methodology to execute, just in a different state.

This, in essence, is the reason that SAAS (Software As A Service) companies get such high valuations in the market, while conventional businesses, especially in sectors where the client doesn’t need to come back for a repeat order anytime soon, get low valuations. If the SAAS company gets one client enrolled into a monthly subscription, they’ll continue to get revenue from that one person and can focus on adding more customers. Once their product is built, the cost of catering to an additional customer is tiny. A political consultancy on the other hand needs to keep getting new clients after every election cycle. Worse yet, even though it’s a high ticket price service (the client pays a large amount), the number of clients at any one point of time is in the low single digits. At best 3-5 states go to elections at one time, and you can only work with one party in a state (because obviously they’re all contesting against each other). I realized the limits of such a business model a few years ago. It could be a great business for your lifestyle, but it wouldn’t be the next big thing…

The lessons:

1. If you want to make it big, choose a business model that scales.

2. Ideally choose a business model with a low or marginal cost of delivery. This means businesses like software or media where once it’s built, it doesn’t cost much (or even anything) to supply it to another person, no matter how many customers you get, are by far the most profitable. Things like restaurants on the other hand have a high marginal cost of delivery, as for every customer you need to buy the raw ingredients, hire additional staff etc. If you don’t franchise or expand locations, even your capacity is limited in how many customers you can serve in a given time period given the number of tables and the kitchen capacity.

This made it obvious to me that no matter how much I liked working in politics, I should be doing something more scalable. And if the scalable thing works, you can always dive back into politics – as a politician this time around!

What industries are scalable?

  1. Technology and software, most definitely.
  2. Consumer products that turn into brands, for sure. Once it’s a brand you have repeat customers, and if it’s a consumable, they’re coming back to you very soon. (Think Gillette selling the first razor for cheap, then making profits on the blades. Or selling diapers. Or the model that fast fashion retailers follow, changing styles every season to get you to come back.
  3. Media, 100%. Though with the shortening attention spans of people and proliferation of free content on YouTube, Instagram, TikTok etc. it’s a painful market to get big in.
  4. Digital products – courses, ebooks, NFTs! Hell yeah!
  5. Branded foods even – think Kellogg’s (But, it can take decades to establish a brand. And generic food is too competitive to make a sizable profit.)

There are many other industries that are scalable and I’m not trying to make a comprehensive list, but when I evaluated what to work on there was another major consideration…

Low capital requirements.

Some people can think of working on business ideas that involve building a factory or buying land and machinery. That was the kind of thing I’d personally totally sidestep though. Personally don’t have that kind of capital, and I’d hate to raise and lose someone else’s money. If done right though, these businesses are great! Especially true for a country like India where access to capital is in itself a major moat! Very few people can get access to capital, that means if you build something capital intensive, you’ll have very low competition. (That’s why Ambani is safe with his huge capital outlay for installing Reliance Jio telecom towers, and why petroleum refining businesses are globally profitable).

No Humans!

The last consideration on my list: businesses which require few people. Not because I don’t like working with huge groups, but because it makes profitability trickier. If you need to hire more people as you expand, the chances of something going wrong at some point – like the current rise in interest rate leading to a drying up of VC funding – can lead to disaster!

A business where the revenue per employee is high is always a better work environment too in my experience. If you’re in a job think of how much profit the company makes and divide it by the number of people – that’s the highest wage the company can pay on average and still survive. If this number is low, that industry will only have low wages! That’s why retailers and shopkeepers can’t go out and pay crazy high wages – they operate on tiny margins. Here’s a great website breaking down revenue per employee for Starbucks and Microsoft – largely why some sectors pay better than others.

The Decision

With this much clear, it was obvious technology was the best industry to target. Great businesses can be made in many (even any) sectors, but tech is one of those rare sectors that checks off every box in the checklist. No wonder that’s the sector a lot of the world’s most successful companies are in.

The title of this post says Starting Out? Choose a Business that Scales! That isn’t my actual advice though. I think if you’re really just starting out in your career, the best idea might be to work at a normal job, or run a business that is profitable from the get go. I’m working on a multi-million dollar idea now, but I couldn’t have really done it even 3 years ago. Ideas like this take time to yield results, and unless you have the money to survive through the slog, you’ll likely desert the endeavor. So the best way to do something like this seems to be to build a capital reserve and safety net doing something else for a few years.

Salary and regular income is a tough drug to quit. It totally makes it leaving and going to entrepreneurship harder once you’ve experienced regular income and your lifestyle has gotten better than it was in college. The fact that your friends would be advancing in careers and getting salary hikes while you’re working essentially unpaid for the initial phase of a startup is also trickier the more experienced you become in a career. Even then, these factors are psychological. Actually not having the resources to stick with building a startup, jumping in, then being forced to go back to a job before the startup even has the time to test the market due to financial constraints is a real pain! Avoid it.

Coming up with an idea

As Michael Seibel, the Managing Director of YCombinator keeps saying, you’d be a lot better off if you didn’t start with a product in mind and instead started with a problem in mind. Then you evaluate what could fix the problem, and only then do you start thinking of a product. If you try to force a product on the market that it doesn’t want, it’s a recipe for disaster.

Step 1: Identify a problem or pain point people have.

Step 2: Brainstorm potential solutions.

Step 3: Test the solution (or even the main assumptions in the solution) in the cheapest manner possible. Build an MVP (Minimum Viable Product), or offer a service without building anything that tests if the solution is working for people, and only settle on a specific solution if it is actually solving the problem! Till you find this, keep iterating and trying new ideas to solve the problem.

Ideally, you’d yourself be one of the people facing the problem you’re trying to solve! Makes testing the solution a whole lot easier. If you aren’t in the target market for your product, talk to consumers and get feedback. Only invest yourself into a product idea once you know it at least kind of works.

Note: A very simple way of testing out if a solution works can be something as simple as making a landing page, or social media post asking people to sign up for your product (even when it doesn’t yet exist).

Story: The founders of DoorDash (a US food delivery startup) literally just put their phone number on a website and asked people to call for delivery in their college – zero tech! Once the phones started ringing, they knew there was demand. That’s when they built the tech.

Step 4: Develop the product further. Make sure you have product market fit.

Step 5: Scale! This is when you can go crazy with the marketing. Many startups fail because they burn money in marketing a product that doesn’t solve the problem they think it does, and even worse… a problem that the customers don’t even feel they need to solve!

So this is the general framework I used to come up with, and then test an idea. I’m confident I have the right problem identified. Still working on testing if I’m at the right solution.

It’s a wrap!

Thank you all for reading. I hope you find this series to be useful and entertaining. And I really hope it gives you something new to think about each week! Please do scroll to the bottom and subscribe if you haven’t done so already. Some content in the coming weeks will go exclusively to the newsletter audience.

Coming Up:

Next week we’ll cover what the idea is that I’m working on, how I hit upon it, and what the current progress on all of it is! PS: I’ll also be covering some of the tools I used to build this blog, and how much it cost to set up.


Refining is almost always profitable, but extraction isn’t. Countries like Saudi Arabia have oil waiting to gush out of the ground, so they can extract it really cheaply. Many US shale oil producers went bust when crude prices were low 5-7 years ago! Important lesson here: Even within the same industry, some parts of the value chain are much better than others. Sell shovels in the gold rush, don’t go prospecting for gold!