When adding people stops adding money

Summary: Indian IT is entering a structural shift where growth is no longer driven simply by hiring more engineers. This story explores how AI is reshaping the industry from a labour-arbitrage model into a more complex business built around productivity, domain expertise and AI-led enterprise transformation.

Summary: Indian IT is entering a structural shift where growth is no longer driven simply by hiring more engineers. This story explores how AI is reshaping the industry from a labour-arbitrage model into a more complex business built around productivity, domain expertise and AI-led enterprise transformation. For almost three decades, Indian IT had an amazingly powerful business model: take work from the US or Europe, deliver it from India at a lower cost, hire more engineers, bill more hours, earn the spread and keep scaling. That model built a $300 billion-plus Indian technology industry, created nearly six million direct jobs, and made TCS, Infosys, HCLTech, Wipro and Tech Mahindra global names. But FY26 is the first year this model has started to look structurally different. The old equation was simple: more people meant more revenue. Now revenue is growing faster than headcount. NASSCOM expects India’s technology industry to touch around $315 billion in FY26, up 6.1 per cent, while headcount growth is only around 2.3 per cent. Truth is, the industry is still growing, but it is no longer growing mainly by adding people. AI is accelerating this shift much faster than most people expected. It will be very dramatic to say that Indian IT companies are being disrupted by AI. What is actually happening is that the lower-end work within IT is gradually being repriced across contracts. Testing, maintenance, basic coding, support, documentation, repetitive BPO work, report generation and simple workflow automation are exactly the kinds of tasks where AI can significantly reduce human effort. A project that earlier needed 100 people may still need 100 people today, but over the next few years, it may need 70, then 50, with the remaining team using AI tools to produce the same output faster. A bank that earlier needed 2,000 people for customer support may eventually run the same workload with 800 to 1,000 people plus AI copilots. That is why revenue per employee has suddenly become the number everyone is watching. In FY26, TCS, Infosys, HCLTech and Tech Mahindra reportedly saw revenue per employee improve by around 3-4 per cent. Sounds small, but for companies with 2-6 lakh employees, even a 3 per cent shift in productivity changes hiring plans, margin calculations and the entire pyramid structure underneath. The bigger worry, though, i

This article was originally published on June 01, 2026.