“Why the Next Decade of Venture Returns Will Be Built on Atoms, Not Just Bits”
“Why the Next Decade of Venture Returns Will Be Built on Atoms, Not Just Bits”
Hi, I'm Shibil. I work at Speciale Invest, a Pre-Seed/Seed VC firm in India focused on deep tech—the kind of startups solving genuinely hard problems in science and technology.
My path here was anything but linear. I was doing biophysics research at IISER and then at IISc when I got the itch to build something. Started a Femtech company. Couldn't scale it. Joined TeOra, a Singapore biotech startup, to run their India operations. Somewhere in that chaos of building, failing, learning, and a fair bit of luck—I landed at Speciale.
I wrote my first deeptech thesis and fundraising guide two years ago. Reading them now feels like looking at old photographs—the bones are right, but the world has moved. Technologies that seemed five years out are shipping. Sectors that felt uninvestable are suddenly crowded. The entire landscape of what's possible and what's investable has shifted underneath us.
So this is an update. New frameworks. New sectors. New conviction.
If you're building in deeptech: shibil.ahammad@specialeinvest.com — I'd love to hear what you're working on.
If you want to follow along: LinkedIn | Medium | InSync Newsletter
Let me start with a number that should make you pay attention:
Deeptech delivered 16-17% net IRR versus 10% for traditional tech—sustained over two decades. McKinsey analysis of 115 deeptech funds vs. 1,572 traditional tech funds (2003-2020)- Link in the ref section
BCG's more recent analysis confirms it: 26% weighted average IRR for deeptech versus 21% for traditional VC.- Link in ref section
So ,the category everyone said was "too slow, too capital-intensive, too risky"—has outperformed software? By 60-70%?
This completely inverts the conventional wisdom I grew up with as an investor. The easy bets aren't actually easy. The hard bets aren't actually that hard. Something fundamental is happening here, and this report is my take on this
Let me explain why.