There is a number that Indian IT companies increasingly like to talk about: revenue per employee. It is a useful number because it tells us how much money a company is generating from every person on its payroll. And right now, that number is rising across the country’s biggest IT services companies.
The obvious explanation is AI.
Generative AI, coding assistants, automation and agentsare supposed to be doing what Indian IT has historically needed armies of people to do. So, if revenue is rising while headcount is flat or falling, it is tempting to call this the beginning of a new, AI-driven IT services model.
But the numbers tell aslightly less exciting story.
During April-June FY27,TCS, Infosys, HCLTech, Wipro and Tech Mahindra together generated $25.6 billion in revenue. All five reported higher dollar revenue per employee than a year earlier, although the reasons were not exactly the same.
TCS, for example,added 9,279 employees during the quarter, taking its total workforce to 5,93,798. Yet its headcount was still 19,271 lower than it was a year ago. Revenue, meanwhile, rose to $7,624 million from $7,421 million.
That pushed quarterly revenue per employee to roughly $12,840, compared with about $12,105 a year earlier.
Infosys told a similar story, although at a smaller scale. Revenue increased to $5,082 million from $4,941 million, while its headcount stood at 3,28,062, only slightly above the 3,23,788 employees it had a year ago. It also cut 532 employees sequentially.
Its quarterly revenue per employee came in at roughly $15,491, compared with about $15,260 a year earlier.
And this is where the AI argument becomes interesting. Infosys says AI-led services accounted for8.2% of revenue in the quarter.
HCLTech has also been talking up AI-linked growth, even as it cut 3,292 employees during the quarter to end with 2,23,889 people.
Wipro added 888 employees, while Tech Mahindra reduced its workforce by 863 sequentially. Their revenue per employee also increased.
Put together,it looks like Indian IT is beginning to get more revenueout of fewer or, at least, slower-growing workforces.
Join Neysa AI & AIM Innovator Series to Learn What Breaks When AI Moves to Production
As enterprise AI adoption accelerates, a new category of AI infrastructure, often described as a Neocloud, is emerging between these two approaches, promising a different balance between cost, operational control, and deployment speed.AIM, in partnership with Neysa, is hosting the next edition of the AIM Innovator Series to explore this topic.Click here to find out more and register now.
What’s Really the Case?
The industry would understandably like tocall this an AI productivity story.
But there is another explanation that is far less futuristic. Companies have been cutting costs.
They have been improving utilisation, moving work offshore, rationalising teams, reducing bench strength and becoming more selective about hiring. AI is now being added to that process, making some of these changes easier and faster.
That distinction matters.Data from IT industry analyst firm Unearthinsight shows that revenue per employee has not suddenly exploded since AI arrived. In fact, over the longer term, the gains have beenfairly modest.
Between FY19 and FY26, Infosys’ revenue per employee rose from $51,722 to $61,346, a CAGR of 2.47%. TCS went from $49,290 to $51,353, a CAGR of just 0.59%. HCLTech moved from $62,567 to $64,548, while Tech Mahindra went from $41,051 to $43,252.
Wipro actually went in the opposite direction, falling from $45,144 in FY19 to $40,763 in FY26.
These are not numbers thatscream sudden technological revolution.
Gaurav Vasu, Founder and Chief Executive of UnearthInsight, puts the split rather bluntly.He estimates that around 60% of the productivity improvement can be attributed to workforce rationalisation and optimisation, while the remaining 40% comes from AI-driven efficiency gains.
That does not make the AI story irrelevant.It just makes it less neat.
AI can make a smaller workforce more productive. But a smaller workforce itself will mechanically push revenue per employee higher if revenue remains stable or grows.
The two things are happening together.
And that is probably what makesthe current phase of Indian IT interesting. AI does not necessarily have to replace a large number of employees overnight to change the economics of the industry. It can quietly alter how many people are required to deliver the same amount of work.
India Doesn’t Have Enough GPUs: Here’s Why | Jarvislabs.ai Founder On AI Infrastructure
In this episode of Simulated Reality, we sit down with Vishnu Subramanian, Founder ofJarvislabs.ai, to unpack the infrastructure challenges shaping India’s AI future- from the rising cost and availability of GPUs to data centres, compute economics, open-
What Happens Next?
The bigger question iswhat happens when AI gets better.
Praveen Bhadada, CEO and Managing Director of NEOVAY Global, describes thecurrent phase as a transition from improving utilisation to bringing in ‘non-linearity’. That is the part the industry has been waiting for.
For decades, Indian IT services has largely operated on a simple equation: more clients and more work require more people.Revenue growth was therefore closely tied to headcount growth.
AI challenges that equation. If one engineer can eventually do the work that previously required two, or one team can handle the output of several teams, revenue and headcount no longer have to move together.
But we are not quite there yet.
AI revenue is still a relatively small part of the overall businessfor most large Indian IT companies. Even where AI-led services are growing quickly, they sit alongside huge legacy businesses built around application maintenance, infrastructure, consulting and managed services.
So when a company cuts 3,000 people and its revenue per employee rises, it is difficult to say that AI alone caused it.
There is also something else happening underneath the numbers. Indian IT companies are becoming more comfortable with the idea that they may not need to keep adding employees at the same pace as before.
That shift may have started with cost pressure, weak demand and margin concerns, butAI gives companies a much more powerful tool to make it work.
And this is perhaps the more important change.The question now is whether Indian IT can keep growing without adding people at the same rate.
That is a much harder metric to fake. For now, the answer seems to be moving in that direction.
Ready to Enter the Code Kitchen?
Think you’ve got what it takes to code under pressure?Code Kitchen is India’s coding reality show, where developers take on real-world engineering challenges, climb the leaderboard, and compete for a ₹10 lakh prize pool. Along the way, you’ll showcase your skills to top tech companies and industry leaders.
Apply now and get cooking.
