In IT Firms, AI is Boosting Revenue Per Employee—at the Cost of Employees

When India’s IT bellwether Tata Consultancy Services (TCS) reported its full-year results last month, the number that stood out was not the topline. Dollar revenue had, in fact, declined 0.5% year-on-year to $30 billion. 

It was revenue per employee, which grew 3.5% from the previous year to $51,360. At the same time, the company had shed 23,460 employees over the course of the year, ending March with a headcount of 584,519. 

Meanwhile, the company's annualised revenue from AI services crossed $2.3 billion. Its operating margin, excluding one-offs from the impact of the labour codes, hit 25.3% in FY26—the highest in four years. 

That combination—revenue per employee rising even as absolute revenue softens—is now the defining story of India's $300-billion IT services industry as it digests the first full financial year of widespread AI deployment. 

Across IT services giants and mid-tier players like Coforge, the most consequential disruption from artificial intelligence is not a torrent of new client revenue. It is a structural compression of internal costs and an improvement in employee productivity.

Four of India’s top IT firms—TCS, Infosys, HCLTech, and Tech Mahindra—posted a 3–4% rise in revenue per employee in FY26. However, the number sits uncomfortably alongside a net reduction of nearly 7,000 employees across tier-1 IT firms.

“A very traditional metric for IT firms was revenue per headcount, and usually IT companies were able to increase their revenue by adding to the headcount. Now, the strategy is slowly changing. For example, TCS laid off employees and reduced hiring. Still, their efficiency and margins are improving because of AI solutions, which they are bringing to their internal usage,” Uddheshya Goel, a financial advisor at Empirica India, tells AIM.

He adds that this could be a trend in the coming quarters, wherein IT companies increase the revenue per employee without actually increasing the headcount, thanks to AI productivity gains.

For three decades, the IT services industry had operated on a single, reliable equation: more clients meant more projects, more projects meant they need more resources. But that assumption is now being seriously tested.

“Beyond delivering measurable, scalable AI impact for our clients, we are also transforming internally towards an autonomous, zero-friction enterprise by reimagining core operations and employee experiences,” C Vijayakumar, CEO, HCLTech, said in the recent quarterly earnings call.

SDLC Efficiency

As AI tools allow IT firms to complete work faster and with fewer people, the software development lifecycle is becoming more efficient. Now, clients are beginning to push back on pricing. If software testing that once required 10 engineers and eight weeks can be done by two engineers and an AI agent in 10 days, the logic of the old invoice is hard to defend. 

“I think most of the enhancements in models are really driving more and more velocity and efficiency in the SDLC lifecycle. I think that piece could go through a little higher deflation based on the model outcomes,” Vijayakumar remarked.

Coforge is also embedding AI across internal delivery operations. The company reports productivity gains of 25–35% across its software development lifecycle through AI-assisted delivery models. Code generation and review processes have improved efficiency by 40–60%, while legacy modernisation projects are being completed up to 10x faster.

Gaurav Vasu, Founder and CEO of cognitive intelligence platform UnearthInsights, says mid-tier IT firms, because of their smaller scale, are better positioned to optimise people costs, improve revenue per employee, and grow revenues faster as they are able to pivot to AI quickly. However, he adds that comparisons with large IT firms are not entirely fair given the latter’s broader geographic presence and wider range of service offerings.

The Jobs Equation

If rising margins and revenue per employee represent the upside of the AI productivity story, the downside is showing up in the staffing data. 

According to technology staffing and solutions provider TeamLease Digital, AI is expected to automate 30–35% of entry-level tasks in IT services over the next 3–5 years, though only about 15–20% of roles face full displacement. Roles in testing, L1 support, and basic coding are the most exposed. This will compress, rather than eliminate, the traditional pyramid model. 

Senior-to-junior ratios are already shifting from 1:8–10 to 1:4–6, with increasing demand for multi-skilled engineers, indicating a move toward a more skill-intensive and less volume-driven workforce structure.

“Pricing explains everything. Indian IT bills clients hourly and per person, so more work requires more hires. Meta and Microsoft don't work that way; their products serve millions without adding headcount. But that's changing. AI is shrinking billable hours and forcing Indian IT toward contracts where clients pay for outcomes, not people. As that shift spreads, hiring will fall with it. 

The firms making that transition fastest will cut headcount first,” observes Neeti Sharma, CEO of TeamLease Digital.

The IT industry is experiencing a dual, simultaneous disruption. Internally, AI is compressing delivery costs, raising revenue per employee and expanding margins. Externally, it is beginning to compress the prices clients are willing to pay, threatening topline growth. 

The firms navigating this most successfully will be those deploying the internal windfall to fund the transition toward higher-value AI-led services.