5 Key Takeaways
- India's top five IT firms have reduced their bench by 25% (about 75,000 professionals), marking the end of the traditional 'hire and deploy' staffing model.
- The primary driver of this contraction is slow post-pandemic growth, not just automation, with companies shifting to local hiring and a 'skill and bill' approach.
- Demand for AI, generative AI, data science, and cloud skills has jumped 30-40%, while traditional mid-level delivery roles have fallen 20-30%.
- Generic software engineers now face longer reassignment times (60-90 days) and lower salary hikes, whereas AI specialists command 20-40% premiums and faster offers.
- The era of a permanent reserve bench as a safety net is over; professionals must rapidly upskill to remain relevant in a leaner, more specialized industry.
The Quiet Reset: Why India's IT Giants Are Running Out of Spare Hands
A 75,000-person contraction marks the end of a decades-old staffing philosophy — and for millions of engineers, the rules of career survival have changed permanently.
The numbers are startling. Over the last two years, India's top five IT services companies have quietly shed roughly 75,000 professionals from their "bench" — the reserve pool of employees waiting between projects. What was once a sprawling force of nearly 300,000 unassigned workers has shrunk to about 225,000, a drop of 25 percent. This is not a temporary blip. It marks the end of a decades-old staffing philosophy, and for millions of engineers, the rules of career survival have changed permanently.
To understand the magnitude of this shift, you first have to understand what the bench really meant. For years, IT giants like Tata Consultancy Services (TCS), Infosys, Wipro, HCLTech, and Tech Mahindra operated like standing armies. They hired thousands of engineers in anticipation of future contracts, stockpiling talent so they could deploy teams instantly when a client signed a deal. If a project ended, employees didn't leave; they returned to the bench, drawing full salaries while the company found their next assignment. This model provided a massive cushion against demand fluctuations. It was the industry's shock absorber.
That shock absorber is now being dismantled. The proportion of unassigned employees relative to the total workforce has narrowed sharply. Pareekh Jain, CEO of EIIRTrend, points to the new reality:
Data from TeamLease Digital paints an even starker contrast, estimating the current range at 8 to 12 percent, down from peaks of 20 to 30 percent in previous years. The buffer is gone.
The End of "Just-in-Case" Hiring
Historically, maintaining a fat bench was a symbol of strength. It signaled that a company was aggressive, ready to pounce on growth, and willing to absorb short-term costs for long-term speed. This strategy worked brilliantly during periods of rapid expansion, when the industry was clocking double-digit growth and the pipeline of large outsourcing deals seemed endless.
Today, that logic has been inverted. Companies that once felt comfortable with 4 to 5 percent of their workforce idling on the bench are now driving toward astoundingly lower targets. Some are pushing to keep unassigned staff at just 1 to 1.5 percent of their total headcount. The operational discipline is ruthless. TCS, for instance, has reportedly capped an employee's bench duration at around 35 days annually. If a professional remains unassigned beyond that window, performance evaluations are triggered, and those who cannot be allocated are asked to exit. The message is clear: the company is no longer a guaranteed parking lot for talent.
What makes this contraction permanent rather than cyclical is a fundamental recognition of unpredictability. In the old model, companies could roughly guess the type of talent they would need. That is no longer possible. Gaurav Vasu, founder of UnearthInsight, dissects the broken logic:
In the current climate, forecasting demand with that level of precision is a fantasy.
Growth Slumps, Not Just Robots
Whenever the IT sector tightens its belt, automation and artificial intelligence are quickly branded as the job-killing culprits. However, in this specific contraction of the bench, technology isn't the primary driver — slow growth is. The post-pandemic surge of digital transformation created a sugar high of hiring, and when that demand normalized, companies were left with a workforce bloat they could no longer justify.
"Low growth is the bigger factor in bench reduction today," confirms Pareekh Jain. He notes a critical geographical nuance that reduces the need for a homegrown reserve army: companies have significantly increased local hiring in their client markets over the last five to six years. If a US or European client needs a team, the IT firm is increasingly likely to hire locally rather than flying someone over from a bench in India. When growth eventually returns, Jain argues, firms may not need to rebuild their domestic bench to historic levels because the deployment model has already shifted.
This doesn't mean technology is irrelevant to the story; it has simply reshuffled the deck of who gets to work. The erosion of the bench is not a uniform, blunt-force layoff. It represents a violent recalibration of skills. Neeti Sharma, CEO of TeamLease Digital, reveals a stark divergence in demand. Over the same two-year period that saw the bench collapse, demand for traditional mid-level delivery roles fell by roughly 20 to 30 percent. These are the classic IT service jobs: managing teams, maintaining legacy code, running testing protocols, and overseeing operations. Simultaneously, the hunger for skills in artificial intelligence, generative AI, data science, and cloud technologies has jumped by 30 to 40 percent across these same firms.
The bench isn't just shrinking; it's being completely reshaped. The "excess" is concentrated in legacy skills that cannot be easily retooled for the AI boom, while the "demand" is for a type of engineer who rarely sits around waiting for a project.
The Cost of Being Generic
The new hierarchy of talent is vividly illustrated by changing compensation and placement timelines. In the past, a generic software engineer with 8 to 12 years of experience who found themselves on the bench could expect to be reassigned within 30 to 45 days. Today, that waiting period has stretched to 60 to 90 days, according to TeamLease Digital data. For a professional drawing a monthly salary in the lakhs, sitting idle for two to three months makes them a visible cost center that companies are eager to eliminate.
Salary premiums tell the same story. During the hiring frenzy of the 2022-23 fiscal year, a professional switching companies could command a 25 to 35 percent hike for non-specialist roles. Those days are over. Premiums for lateral hiring in non-AI roles have collapsed to a modest 10 to 20 percent. The market no longer rewards generic experience.
Premium commanded by professionals with specialized generative AI skills, depending on seniority. If you can build a large language model or architect a cloud-native AI pipeline, you are not sitting on a bench — you are naming your price.
The internal corporate ladder is also being sawn down. The archetypal "people manager" — the mid-level leader whose primary job was to oversee the work of 30 to 50 engineers — is being redefined. This role isn't disappearing entirely, but Vasu notes a sharp pivot in its responsibilities. The focus is shifting away from headcount oversight and toward revenue expansion and profitability management. In simple terms, managers will not be judged on the size of their teams but on the margin and growth those teams generate. Carrying extra, unassigned individuals becomes a direct drag on a manager's metrics, disincentivizing the hoarding of bench resources.
The View from the Bottom and the Top
The structural erosion of the bench creates a barbell effect in the job market. At the entry level, the news is grim. Kapil Joshi, CEO of IT staffing at Quess Corp, reports that hiring at the entry level has declined by around 30 to 35 percent. This is directly linked to the elimination of the bench; companies that historically absorbed massive batches of fresh graduates and trained them on the bench for months can no longer afford the carrying cost. If a fresher cannot be billed to a client on day one, their utility is questioned.
Interestingly, leadership hiring tells a different story. Global capability centers (GCCs) and IT firms are still fighting for senior talent. Joshi notes that the share of leadership roles in the hiring mix has actually increased, moving from around 15 percent in 2024 to approximately 20 percent in 2025. But, critically, the nature of these leaders has transformed. More than 50 percent of the demand for these senior jobs is now driven by emerging skills. Companies aren't looking for operational generals; they are hunting for visionaries who understand platform engineering, AI strategy, and cloud architecture.
A Leaner, Colder Future
The message from this 75,000-person contraction is unambiguous: the era of the permanent reserve force in Indian IT is over. The sector that built its empire on the model of "hire and deploy" is pivoting to "skill and bill." The traditional social contract, where large corporations absorbed the industry's talent risk by parking people on the bench, is being ripped up.
For the workforce, the implications are profound. The concept of a "safe" legacy career in a large IT services firm is now a mirage. The buffer that once protected employees from the friction of the market has been removed. The 60-to-90-day window of idleness on the bench has become a ticking clock, a final notice that forces professionals to either rapidly upskill into AI and cloud roles or face an exit.
The Indian IT industry is not collapsing; it is purifying. The fat has been trimmed, and the muscle that remains is being rewired for speed and specialization. When growth finally rebounds, the headcount will likely recover, but the bench — that comfortable, salaried in-between space used by millions to ride out project droughts — will not return in anything resembling its old form.
No comments:
Post a Comment