There’s a Lot of Momentum in Deep Tech Compared to Five Years Ago, Says Hemant Mohapatra

As investors, we always want more and more founders to show up so we can deploy more high-quality capital behind the highest-quality founders. So, the answer is yes and no.

When I joined Lightspeed in 2018, I probably saw one or two deep tech or frontier-tech companies a month. Now, in just one week, I’ve seen a company trying to cure depression with a headset, a brain-computer interface company installing headsets on animals to translate what they’re feeling to human owners, a robotics company and a nuclear-fusion company.

So yes, things are coming up. When we invested in Pixxel Space, it was one of just 20–25 companies in the entire region doing anything in space tech. Now there are over 300 companies, a 10x improvement in four years. Compared to China or the US, we still have a long way to go, but compared to India five years ago, there’s a lot of momentum.

Companies should flip where their market is. If your customers are Indian defence companies or local buyers, there’s no sense flipping to the US. But if your buyers are in the US, you need to be in that market.

Pixxel Space flipped to the US not because they could raise more money there, but because their clientele was based there. They had to build a team to sell into that clientele, both government and non-government.

It’s very case by case. I’m only aware of what’s happening in our portfolio. But yes, I’d say the primary reason to flip isn’t funding, it’s customers.

Our fund structure is the same across categories. We don’t change it unless there’s a regulatory reason, like having to operate as an RIA [registered investment adviser] to invest in public markets.

Even for deep tech, once companies come out of the lab and begin commercialising, they scale up very quickly. They may spend 3–4 years in R&D, but once something starts working, it better be truly de novo and compelling. Then the scale-up can be rapid.

It’s very nuanced. Some parts of deep tech should be funded by government grants or universities. For example, building the GPS ecosystem in the US or the space ecosystem in India was government-funded.

Only a limited subset of deep-tech start-ups is ready for commercialisation, and venture capital (VC) is a good fit for maybe just 1–2% of founders. VC money is meant for building massive global companies, not small profitable ones.

There are many other funding paths: loans, friends and family, grants or customer revenues. Venture capital gives you, say, $3mn today that you might otherwise earn in three years and asks you to compress your journey into quarters. That’s not for everyone.

We’re moving into a world where differentiated, net-new technologies matter more because software has become commoditised. Writing code is easy now, it wasn’t 20 years ago.

So yes, building IP is important. But we must avoid being just an IP house. Value accrues when you own the IP and sell, brand and distribute it yourself.

Take Arm, for instance. They had amazing IP, but most people know Intel or AMD, not Arm. Why? Because those companies controlled branding and customer relationships. So yes, create IP but also build the full stack around it.

It’s not just about design. That’s like saying, ‘We have coders, why haven’t we built a new Google?’ We do a lot of core research, but do we have the ability to productise that research?

Can we build teams that know how to design, brand, position and sell globally? People in the US grow up selling lemonade on the roadside. Indian kids grow up doing math classes. We’re great at analytics, not sales.

Many SaaS [software-as-a-service] founders here go for open source or product-led growth because they’re not comfortable selling. But selling is a critical skill for building big businesses.

This can’t be solved piecemeal. Government or big companies have to put a stake in the ground and say, ‘We’re doing this.’

The Indian government making all 10,000 scientific publications free is a bold move. It’s never been done at a country level.

People want to return. Every month, I get pings from top-tier folks at Google, Tesla, Nvidia asking how to move back. But if they don’t see strong core research opportunities here, they won’t return.

It’s a chicken-and-egg problem. We need principal investigators to return so they can guide younger talent. But they won’t come unless there’s meaningful work to do. Someone has to break that cycle.