The people a Global Capability Center most wants to keep are leaving for reasons a raise won’t solve. Fixing that has less to do with employer branding than with how the centre is built and run.
India’s Global Capability Center (GCC) sector passed a marker this year that quietly changes the talent equation. The 2026 NASSCOM–Zinnov landscape report counts 2,117 centres in the country, employing 2.36 million professionals, with more than 500 of the Forbes Global 2000 now running operations here. Growth on that scale used to be the headline. It isn’t any more. The sector’s own leadership has been clear that cost and location are no longer the point. Once those stop being the differentiator, the pressure moves to something harder: keeping the people who do the work.
And keeping them is getting harder in a specific way. Overall attrition across India’s GCCs is around 16% in 2026, roughly where it has sat for three years. The figure that deserves more attention is narrower. Churn among high performers is running above the average and climbing, and in AI and cloud roles average tenure has compressed to somewhere between 18 and 24 months. The centres are not losing everyone. They are losing the people they can least afford to replace.
What people actually leave for
The reason they leave is rarely the salary. NASSCOM’s GCC research found that among senior engineers exiting GCCs, 42% named limited career progression as their main reason, 28% cited a lack of ownership over what they were building, and only 19% pointed to base pay. Compensation has already been addressed on the market’s terms: more than 85% of GCCs now run differentiated pay structures, and top performers are seeing increases close to 1.8 times the average, with AI roles higher still. Paying well for scarce skills has become standard practice. It wins the offer. It does much less to hold someone two or three years in, once the novelty of the role has worn off and the ceiling above them has come into view.
An employee value proposition, minus the branding
Strip the marketing language off the term and an employee value proposition (EVP) is simply the full set of reasons a person joins an organisation, does their best work there, and stays. For a GCC, that has to run deeper than a careers page, because the workforce it depends on is young, mobile, and unusually well briefed on what rival centres offer down the road.
The parts that carry the most weight are also the ones hardest to manufacture: a visible route to more responsibility, learning that keeps a person’s skills from ageing out (many technical skills now go stale within two years), mentorship that is designed rather than left to chance, and real ownership of outcomes rather than the execution of decisions taken elsewhere. Across recent surveys of GCC professionals, career growth comes up as both the thing people value most and the area where they rate their employer lowest. Pay, and even work-life balance, trail well behind on that gap. The strongest people are asking for somewhere to grow, and telling their employers plainly that they aren’t getting it.
Why this is structural, not a communications gap
Here is where many GCCs misdiagnose the problem. They read a weak EVP as a storytelling failure and answer with a campaign: a refreshed careers site, a run of LinkedIn posts, an award submission. Then the brand-ambassador programme fades within a quarter and no one is quite sure why.
The cause is almost always the same. The experience inside the centre and the story told outside it have come apart, and people compare notes. What a candidate is promised and what a new joiner actually finds are, too often, two different things, which is why nearly one in three GCCs now worry about what the sector has started calling infant attrition: people leaving inside their first six to twelve months. That early exit is seldom about pay. It is the sound of a hiring promise meeting a different reality.
The career ceiling itself is built into the conventional operating model. In a standard captive centre, strategy is set at headquarters and the India team is left to carry it out. An engineer six to ten years into their career reads that arrangement correctly, as a limit on how far they can go, and no volume of external branding changes how the job feels on an ordinary Tuesday. A structural constraint on people’s growth cannot be campaigned away.
What closing the gap actually takes
The fix is an operating problem before it is a marketing one. It means building the machinery that careers run on: promotion frameworks transparent enough that people can locate themselves inside them, internal mobility that lets someone change function without changing employer, learning tied to real progression rather than an optional course catalogue, and leaders who are visible in the moments that count — the town hall, the awkward question, the layoff handled with some decency — because that is where employees decide whether the values on the wall mean anything.
None of this is exotic. It is expensive to build and harder to sustain, which is precisely why many centres never reach it. The technical mandate lands first; the talent infrastructure trails it by years. A centre can be running genuinely advanced engineering on top of people practices designed for a back-office delivery unit, and the space between the two is where good people go looking elsewhere.
Where the operating model earns its place
This is worth slowing down on, because the choice of model shapes how solvable the problem is. A managed GCC operating model draws a clean line between two things that usually get tangled together. The client keeps everything that should never sit anywhere else: the brand, the mandate, the intellectual property, the people, and the direction of the work. A partner takes on the parts that slow most builds down and that few parent companies have any reason to build from scratch — entity setup and compliance, the hiring engine, facilities, and the day-to-day running of the centre, including the career architecture itself. It is commonly structured so the whole operation can move in-house whenever the client chooses.
The connection to EVP is direct. The two things a young workforce weighs up, whether they belong to something with meaning and whether they can build a career where they are, tend to work against each other in the two default choices. A pure captive can offer identity and ownership but often hasn’t built the people-and-career machinery to support them. A traditional outsourcing arrangement has the operating muscle but hands employees a supplier’s badge instead of a mission. A managed model is one of the few ways to put both in the same place: the client’s brand and ownership in front, a mature talent engine underneath. For the individual deciding whether to stay another year, that combination is the difference between holding a job and building a career.
The point for anyone running a centre
The GCC talent market has grown past the things money can simply buy. Pay parity, a good building, a name people recognise: these now qualify a centre to compete; they no longer settle who wins. What separates the centres that keep their best people from the ones stuck rehiring is quieter and much harder to copy — whether someone can look up from their work and see room ahead of them.
India’s GCCs are on track for close to 2.8 million professionals by the end of the decade. The centres that treat careers as infrastructure, and that staff their operating model to actually deliver it, will hold the talent everyone else is still trying to hire.
At Versitae, we help enterprises set up and run Global Capability Centers where the employee value proposition is designed into how the centre operates rather than added on later. If you are planning a new centre, or rethinking one that isn’t holding its people, let’s talk.
