So Why Does It Feel Like Something Broke?
I’ve been sitting with this for a few weeks now.
The numbers look fine on the surface. Revenue up. Deal wins being announced. Management commentary on earnings calls is optimistic. And yet something about Indian IT feels structurally different this year — not broken, but quietly, irreversibly shifted.
The thing that keeps nagging me is this: NASSCOM expects the industry to hit around $315 billion in FY26, up 6.1 per cent. Headcount growth this year is 2.3 per cent. The industry is growing at nearly three times the rate it’s adding people.
That has never happened before at this scale. And nobody is talking about it plainly enough.
For three decades, this sector’s entire logic was built on a simple human chain — more work comes in, you hire more engineers, you bill more hours, revenue goes up. That chain is what created six million direct jobs. That chain is what turned IT into the most reliable middle-class career ladder in India for an entire generation. Every year, roughly 15 lakh engineering graduates came out of colleges across the country, and the IT sector absorbed a huge chunk of them. Campus placement from TCS or Infosys wasn’t just a job offer — it was a family milestone.
That chain is breaking. Not snapping dramatically. Just… quietly fraying.
The work being automated first isn’t glamorous software development. It’s the stuff that actually employed the most people — testing scripts, documentation, BPO workflows, customer support, repetitive report generation. Per a Wealth Insight (June 2026) analysis, a bank that earlier needed 2,000 people for customer support could eventually run the same workload with 800 to 1,000 people plus AI copilots. A project that needs 100 people today may need 50 in three to four years. Same output. Smaller team.
And here’s the part investors aren’t fully pricing in yet: this compression shows up first in contract renewals. A $100 million deal comes back to the table at $80 million because the client knows 20 per cent of the work got automated. The relationship holds. The revenue doesn’t. That’s a new kind of pressure this industry has never really faced at scale.
I want to be honest about the other side though, because the pure bearish take on Indian IT misses something important.
These companies have been sitting inside client systems — banks, hospitals, manufacturers, governments — for 15 to 20 years. They know the legacy architecture. They know where the bodies are buried. An AI-native startup from San Francisco cannot walk in and replicate that institutional knowledge in a year or two. The engagement model is changing from “send us your engineers” to “come solve our business problem using AI” — and that second conversation? Indian IT firms are actually well-placed to have it, if they move fast enough.
The ones I’m watching are the ones actually changing how they price work. Not per-seat, not per-hour. Outcomes. Service tokens. Transformation fees. Tech Mahindra is trialling this. TCS has been vocal about repositioning as a full-stack AI services player. These aren’t just earnings-call buzzwords — the underlying contract structure is genuinely shifting.
So what’s the actual trade here?
Revenue per employee is now the number that matters most. In FY26, TCS, Infosys, HCLTech and Tech Mahindra all reportedly saw it improve by 3–4 per cent. That sounds small. But across workforces of 2–6 lakh people, even a 3 per cent shift in productivity rewrites the entire cost structure — hiring plans, bench ratios, fresher absorption, margin guidance.
The companies expanding that number while also winning AI-native contracts are the ones building for the next decade. The ones growing headcount the old way while putting “AI” on their investor slides are the ones I’d be skeptical of.
The sector isn’t dying. But the valuation framework from the last cycle is. And there’s a real difference between an IT company that’s repositioning and one that’s just buying time.