The Indian GCC story in 2026: what the numbers actually say

India's Global Capability Centres, the engineering and operations offices that multinationals set up on Indian soil, have been expanding for two decades. The original pitch was straightforward: access English-speaking technical talent at a significant wage differential from headquarters. Hire engineers, run your helpdesk and QA, process your back-office work, and pocket the savings. The model worked, and for a long time the GCC story was essentially a labour arbitrage story.

That story shifted sometime in the early 2020s. Wages rose, the junior-level differential compressed, and companies started asking for more. The industry renamed itself from 'captive centres' to 'Global Capability Centres', a name change that announces ambition without proving it. The question worth asking in 2026 is whether the new name now reflects a real shift in what these centres actually do, and what the answer means for the 1.9 million engineers who work inside them.

The answer is: mostly yes, unevenly distributed. Here is what the data shows.

The numbers in India GCC 2026

India has roughly 1,700 GCCs. They employ approximately 1.9 million people directly. Revenue from GCC operations reached $64.6 billion in FY2024, and projections put that figure at $99 to $105 billion by 2030. The count is expected to grow to 2,100 to 2,400 by the end of the decade, with the workforce expanding to between 2.8 and 3.5 million.

What the headline numbers miss is the pace of the most recent window. In the first two quarters of 2025 alone, 49 new GCCs opened in India, roughly one new centre every three days for six months. GCCs accounted for 38% of total office leasing across India's top seven cities in 2025, at 31.3 million square feet, the highest proportion on record. The sector is not just growing; it is growing faster than comparable periods earlier in the decade.

Bengaluru alone absorbed 11.9 million square feet of GCC leasing in 2024, ahead of every other city in India. Walmart Global Tech leased nearly 465,000 square feet in Chennai in 2025 to expand engineering operations, on top of an earlier Bengaluru expansion. These are not satellite offices hosting a handful of engineers. They are large-campus commitments representing decade-long bets.

What GCCs are actually building

The phrase 'from cost centre to innovation hub' appears in nearly every GCC market report, which makes it easy to dismiss. The underlying data is harder to argue with.

Over 80% of GCCs are actively scaling GenAI projects, according to the EY-NASSCOM 2025 survey. More than 185 dedicated AI/ML Centres of Excellence have been established across India's GCC base. Engineering R&D-focused GCC operations have grown 1.3 times faster than the overall sector growth rate. In 2025, more than half of GCCs were already investing in agentic AI systems: not pilots, but production deployments.

The specific examples are more convincing than aggregate statistics. Goldman Sachs' Bengaluru centre builds the risk and portfolio analytics systems that feed live trading desks globally. JPMorgan's India GCC is leading AI automation in financial products, in some functions ahead of its New York counterpart on specific product lines. Hitachi Energy's India centre owns industrial AI for grid management. Medtronic's India GCC does AI work in medical device software that goes through FDA regulatory validation.

None of that is back-office processing. The shift happened because two conditions were met simultaneously: the senior engineering talent pool in India deepened enough to support genuine R&D mandates, and the wage differential at the junior level compressed enough that companies needed higher-value work to justify continued expansion. Both trends reinforce each other.

Bengaluru at the centre

Of India's 1,700+ GCCs, more than 870 are in Bengaluru. The city accounts for approximately 36% of all GCC talent nationally, around 660,000 engineers. Karnataka as a whole accounts for roughly 35% of India's total GCC workforce.

This concentration creates a specific employer ecology in the city. Three distinct archetypes compete for the same senior engineering talent: IT services firms (Infosys, Wipro, TCS, HCL), GCCs, and Indian SaaS and deep-tech companies. The three have very different propositions at the 8 to 12 year experience band, which is where the competition matters most.

IT services is losing relevance as a preferred destination for engineers who want to build products. Structured careers and large campuses matter less to engineers who have watched colleagues at GCCs and startups build systems that ship globally. The real competition for senior engineers in Bengaluru today is between GCCs and Indian product companies, and the terms of that competition are not what either side's HR teams usually frame them as.

Why engineers leave GCCs

GCCs have a well-documented attrition problem at the senior level. Voluntary attrition across the GCC sector overall fell to a record low of 12.6% in 2024, lower than comparable figures for IT services and roughly in line with mature product companies. But that headline obscures the specific problem: senior engineers in AI/ML and platform engineering roles are leaving at 22 to 30% annually, with spikes even higher at GCCs that have not updated their work profile from the cost-centre era.

The standard narrative is that Indian startups are paying more. NASSCOM's 2025 GCC India Report challenges that framing. Among senior engineers who left a GCC role for another position, the distribution of primary reasons looked like this:

The hierarchy matters. A salary problem has a straightforward fix: raise salaries. A career progression and ownership problem is structural. It reflects the fundamental architecture of how a GCC relates to its parent. GCCs own components, not products. A Goldman Sachs engineer in Bengaluru might build a risk model that powers live trading globally, but the trading product that model feeds into is owned from New York. Strategic decisions happen there. The Bengaluru centre executes, even when it executes at very high quality.

That dynamic frustrates engineers who are capable of more. At the 8 to 12 year mark, many GCC engineers can lead a system end to end, make architectural decisions, and drive an outcome rather than implement a specification. The ones who leave for startups are rarely leaving for a compensation increment. They are leaving because the startup offers the product ownership the GCC cannot.

GCCs are responding, with mixed results. Variable pay rose to 16.1% of fixed salary in 2025, up from 14.8% in 2024. Several GCCs have introduced India-first product mandates, where a product is owned entirely from India rather than partially staffed from here. A smaller number have built internal startup programmes. Whether these changes move fast enough to hold the senior tier is an open question, and the attrition data at GCCs that have made structural changes versus those that have not is telling.

What this means for Indian SaaS founders

The GCC expansion creates a direct hiring challenge for Indian SaaS founders. They are competing for experienced engineers against organisations with MNC compensation structures, established brands, and stable employment. The GCC offer is real: meaningful technical problems, global exposure, job security, and increasingly interesting AI mandates. A startup needs to offer something genuinely different, not just moderately better on one dimension.

But the GCC boom also creates something useful. It is training a generation of Indian engineers on global-scale systems. JPMorgan GCC engineers build payment infrastructure that processes hundreds of millions of transactions. Walmart GCC engineers have worked on supply chains at continental scale. Medtronic GCC engineers have shipped certified medical AI. That experience base does not stay inside GCC walls forever, and when those engineers are ready for ownership, the right Indian startup is their natural next stop.

That proposition is genuinely distinctive, but it has to be real. Engineers with eight years of GCC experience can spot the difference between a startup that says ownership on a slide and one that actually means it. The tell is usually in how technical decisions get made, who holds the architecture, and whether the engineering team genuinely controls the product roadmap.

SaaS and deep-tech companies are projected to capture 60% of Bengaluru's venture capital inflow through 2026. That capital is chasing the same mid-senior engineers that GCCs have trained. The founders who compete on career rather than just compensation, and can make that promise credibly, hire better.

There is a second implication that receives less attention: GCCs are creating a cohort of Indian enterprise buyers who are technically sophisticated. A senior engineer who has spent a decade at a GCC building distributed systems is a completely different kind of customer for an Indian SaaS company than a traditional IT procurement manager. They read documentation carefully. They will notice a poorly-designed API. They will test your security claims rather than accepting them at face value. This raises the bar for any vendor selling to the enterprise segment in India, and rewards the companies that build seriously.

The 2030 projections and what they require

The consensus forecast: 2.8 to 3.5 million GCC employees by 2030, revenue at $99 to $105 billion, the sector contributing roughly 2% of India's GDP. If those numbers land, India's GCC ecosystem will be larger than the IT services industry at its employment peak.

Those projections require two things to stay true simultaneously. First, India's supply of senior engineering talent has to keep growing in quality. The IITs and IIMs produce graduates who are globally competitive; the broader engineering university base is improving but inconsistently. Demand from GCCs, hyperscalers, Indian unicorns, and international product companies all arriving at once is creating genuine tightness in the senior talent market, and that pressure does not resolve on its own.

Second, GCCs have to keep upgrading the nature of their work fast enough to hold the engineers they have trained. The transformation from cost centre to capability centre is real at the right tail: Goldman Sachs, JPMorgan, Medtronic, Hitachi Energy. But the median GCC has further to travel. Engineers in Bengaluru, Hyderabad, and Pune can tell the difference between a centre doing genuine R&D and one that has updated its name while running the same processes from 2012. The attrition data reflects exactly this, with the reformed GCCs showing meaningfully lower senior attrition than the unreformed ones.

The 2030 projections assume the right tail keeps pulling the median up. On current trajectory, with 49 new centres in six months, 80% scaling GenAI, and the structural competitive pressure that comes from competing for engineers who have real options, that assumption holds. The window for GCCs that have not yet made the transition is narrowing, and the talent market will not wait for them to catch up.