There are three ways to use tech
A good friend and advisor, Rohan Gupta, told me there are two ways to use tech:
- Use tech to enhance an existing business, like a mutual fund that lets you open an account online instead of filling a paper form.
- Use deep tech to build a product that couldn’t be built before, like self-driving cars. For this, you need the best people in the world in that domain, like PhDs from CMU.
But there’s another kind of startup:
3. A startup that combines existing tech in a novel way, like Uber. It didn’t invent new tech per se. Everything it used — smartphones, geolocation, cellular data networks, the cloud— already existed. Uber’s genius was in combining them in a new way , like all substances are built by combining protons, neutrons and electrons in different ways.
All three require a techie (eng or product) in the driver’s seat, as opposed to a traditional businessperson.
Let me clarify what I mean by a traditional businessperson: There are a few businesspeople like my aforementioned friend Rohan Gupta who deeply understand tech, the way it works, its culture and priorities. Rohan, for example, proposed starting a startup that uses blockchain, and he used it in a way that actually helps, as opposed to the tons of startups that use blockchain without benefit, because they have no clue when to use it. This demonstrates a better understanding of blockchain than most of the techies I’ve spoken to! I’m not talking about such businesspeople in this post, to be clear. Such a person is more than just a businessperson .
Coming back to the point, all three types of startups need a techie in the driver’s seat.
It’s obvious to see why with the second category of product — deep-tech like self-driving cars.
But it’s also important with the third kind, because only a techie is likely to think of combining existing tech components in a new way, such as to build Uber. In fact, software similar to Uber was built before Uber, but was sold to taxi fleets to manage their taxis. A traditional businessperson typically looks at tech and enhancing his existing business, like running his taxi fleet better. A techie doesn’t work within the framework of an existing business model, so he can come up with a new kind of business like Uber.
Even the first kind of company — using tech to enhance an existing business —requires a techie in the driver’s seat. Let me give you an example. A mutual fund I’d invested in made an app, and when their relationship manager (RM) came by, he suggested I tried it out. So I installed it and it asked me for my folio number. I don’t remember mine. I wanted to give the app my mobile number or PAN, the way I would give a call center operator, but the app wouldn’t accept it. I told the RM, who looked up my folio number, and I entered it. I had to choose a password, and the app didn’t integrate with 1Password, my password manager, so I typed a password. I accidentally signed in instead of created an account. It told me I don’t have an account. So I pressed Sign Up. It again asked me for my folio number. Why? I just told it. I don’t know my folio number to retype it. So I asked the RM again and entered the folio number and password. It created my account, but instead of letting me see how my investments are doing, it logged me out. I typed my information the third time to log in, and it logged in, but while it was loading, it crashed. When I reopened the app, it logged me out. Instead of typing my information the fourth time, I uninstalled the app.
This shows many things that typically go wrong when a traditional businessperson is in the driver’s seat. He thinks of an app as a checklist: “Does it let you invest? Does it let you redeem? Can you see how your investment is performing? Great, we’re done!”. He doesn’t understand user experience, that an app that technically does something but is too hard to use is as good as not having an app. He doesn’t know the importance of having a top-quality tech team in-house, and wouldn’t be willing to pay the market rate for such talent. They’ve outsourced their app. The outsourcer couldn’t care less about the success of the app, since they’re incented to deliver something that barely works, pocket their fee, and move on to the next client. The traditional businessperson doesn’t know what the bar is in tech, and he typically isn’t used to delighting customers the way techies are. He lives in a zero-sum world, used to negotiating everything down to the rupee, since if he pays a supplier a rupee more than he can get away with, he has a rupee less. He’s an expert in slicing the pie to keep a bigger piece for himself, not in making the pie bigger. Masters of zero-sum thinking can’t build anything new.
To build any new product that’s actually better for its target audience than existing alternatives, you need to have a tech / product person in the driver’s seat. This could mean the CEO, or a cofounder, or a CTO who’s an equal — he should have as much say in running the company as anyone else, and should invited to all meetings where critical decisions are being taken, as opposed to being treated as taking care of just one department.
 This is why I chose not to do a PhD. This kind of innovation is looked down upon, despite having a huge impact. I would recommend others skip a PhD, as well, since it has a narrow view of what innovation is.
Besides, for many people, a PhD amounts to going deep into one domain at the beginning of your career, which is premature optimisation. It’s better to start wide: learn a variety of things, like web frontend, backend and mobile. Work in different domains, like fintech, computer vision, retail. Work in a giant company like Microsoft, start a startup, try freelancing. Maybe you’ll want to continue to be an individual contributor, or maybe an eng manager, maybe a product manager, or maybe even a UX designer. Maybe you’ll like a particular domain, maybe not. As time goes by, you’ll understand what does and doesn’t make you happy, and that’s the time to go deep, not prematurely.
 Which brings up another point that you need T-shaped people in the senior management team. Just as a businessperson who is good only on the operational side, doing things as he’s done years, is not right, an engineer who just wants to write code would be a bad fit. Or a designer who wants to make it pretty at a superficial level rather than understanding the users, empathising with them, and solving their problems.