That happened in the year the Union Budget cut the Safe Harbour margin to 15.5%, replacing a range that had run from 17% to 24%, and lifted the turnover threshold from roughly USD 30 Mn to USD 200 Mn with automated approvals valid for 5 years. The reform applies nationally and therefore favours no city over another.
It happened in a year when every state in the country was bidding for those units with capital expenditure subsidies, stamp duty waivers, rent and lease assistance, power subsidies and employment subsidies. India’s emerging cities hold 375 GCC units between them and are actively buying more.
The reform landed, the subsidies landed, and the map did not move. Two-thirds of new GCCs went to Bengaluru or Hyderabad.
Bengaluru alone holds 1,080 of India’s 3,728 GCC units and employs 34% of the 2.36 Mn professionals who work in them. Nearly a third of the country’s GCC units sit in “one city.”
Every company that came and set up their GCC in Bengaluru ran the arithmetic and paid the premium anyway. The ones deciding right now are running the same arithmetic. And this is why, here are the 5 reasons they keep arriving at the same answer.
Two numbers get confused in this market, so it is worth settling them first. A GCC is an offshore unit established by a multinational corporation to perform strategic functions from India, leveraging knowledge-based talent, cost and operational efficiencies. It spans technology, engineering and operations, and includes the shared services centers of MNCs. A GCC unit is one city-level location that a GCC operates. A company that sets up in Bengaluru and later expands to Hyderabad counts as one GCC and two GCC units. Zinnov counts 2,117 GCCs in India operating 3,728 GCC units in FY26E. Every city figure below counts units.
Global 2000 GCC units in Bengaluru
Source: Nasscom-Zinnov GCC Landscape Report 2026
| Bengaluru | 320+ |
|---|---|
| Hyderabad | 165+ |
| NCR | 165+ |
| Mumbai | 165+ |
| Pune | 150+ |
Hyderabad holds 165 Global 2000 GCC units. So do NCR and Mumbai. Pune holds 150. Chennai, 130.
Bengaluru holds 320.
There are 506 Global 2000 companies with a GCC anywhere in India. Two out of every three of them have one in Bengaluru.
Now think about who these companies are. They can afford any city on the map. If the talent isn't there, they wait, or they pay, or they build it. Nothing on a cost sheet explains why two-thirds of them ended up at the same address.
And they didn't turn up together. 18% arrived in the last 5 years. The rest have been here longer, which means each new arrival was looking at what the ones before them had already built, and decided that was worth joining.
That is a long queue of very well-informed companies reaching the same answer, one at a time.
So if you're planning a GCC setup and you expect it to eventually own something that matters to the enterprise, not run a process, but own it, the question isn't really which city. It's whether you want to be the first company to find out whether a place can carry that, or the 321st to benefit from everyone who already did.
India hosts 583 mid-market GCCs. Across the six Tier-I cities they operate 740 units. Click a city to claim its share.
Source: Nasscom-Zinnov GCC Landscape Report 2026
| Bengaluru | 280+ |
|---|---|
| Hyderabad | 125+ |
| Pune | 105+ |
| Chennai | 95+ |
| NCR | 80+ |
| Mumbai | 55+ |
| Tier-I total | 740 |
India has 583 mid-market GCCs — set up by companies doing somewhere between a hundred million and a billion dollars a year. 280 of those mid-market GCCs are in Bengaluru. But the interesting thing is what a GCC means to a company this size.
For a mid-market company, setting up a GCC is a bet. They cannot outbid anyone for a senior architect. They cannot wait three years for a talent market to mature around them. And there is no budget for a second site that quietly corrects the first one.
Around two-thirds of mid-market GCCs from the Americas run a product or platform mandate. So these are not back offices. They are being asked to build, on a budget that allows one attempt.
So you've got mid-market GCCs carrying a serious mandate, on a budget that leaves no room for a mistake, choosing a city they can't easily un-choose.
When those are the stakes, you don't go somewhere and hope the market grows into you. You go where someone has already done your exact job. And in Bengaluru, someone has, 280 times over. The recruiters there have filled a three-hundred-seat brief for a mid-market company. The landlords have leased to one. The advisors have set one up.
That's what half of India's mid-market is actually buying when it picks Bengaluru. The people who staff and lease and advise in Bengaluru have solved a mid-market brief more than twice as often as anyone else in India. You get one attempt at this. Spend it where your problem is routine.
Hyderabad is second at 14% and Pune third at 13%. More GCC professionals work in Bengaluru than in Hyderabad and Pune combined.
Source: Nasscom-Zinnov GCC Landscape Report 2026
| Bengaluru | 34% |
|---|---|
| Hyderabad | 14% |
| Pune | 13% |
| Chennai | 12% |
| Mumbai | 11% |
| NCR | 10% |
| Emerging cities | 6% |
India's GCCs employ 2.36 Mn people, and more than a third of them are in Bengaluru. It's worth sitting with how far ahead that is.
Bengaluru has 34% of the country's GCC talent. Hyderabad, the closest, has 14%. Everyone else trails from there. That 34% is where most people stop. But the number that should interest you isn't how many people Bengaluru has. It's how long they've been there.
The shortage in India was never junior talent. Engineers who can execute, you'll find in half a dozen cities and onboard in a quarter. What's genuinely scarce is senior talent — the people who've run this kind of build before, who've already carried the responsibility and know where these projects break.
That pool is thin everywhere, and you can't hire your way to a senior team overnight. It builds up over years, and it builds up wherever the work has run the longest. Which brings you to Bengaluru.
The first global centre in India opened here in 1980s. The city has been doing this, without a break, for 40 odd years, so the talent here has been solving difficult problems the whole time since. That's the part any other city can't close the gap on. Match Bengaluru on headcount and you still haven't matched it on the years behind that headcount, and the years are what you're actually after.
If the centre you're building is there to run execution, keep the process clean, keep it efficient, keep it cost-effective, almost any major Indian city will do the job.
But if you're building for frontier work; AI, R&D, deep tech, the things the enterprise is betting on rather than just running, Bengaluru is the safe call. The senior talent is already there, which means you can land a Centre Head faster, and once that person is in, the product engineers and the rest of the senior bench come together far more quickly than they would anywhere you'd have to import them from.
That's the whole advantage in one line: the people who make a frontier centre work are already in the city, so you spend your time building the centre instead of convincing them to move to it.
1,080 of India's 3,728 GCC units operate from Bengaluru, 29% of the national total. Hyderabad holds 515 and NCR holds 490.
GCC units, FY26E · Source: Nasscom-Zinnov GCC Landscape Report 2026
Bengaluru has 1,080 GCC units. Hyderabad has 515. NCR has 490. Add the second and third cities together and you still haven't caught the first. Density like that changes how a centre actually runs day to day, in two ways that don't show up until you're operating.
The first is the one nobody expects, because it's usually the argument against Bengaluru: attrition. The city is expensive, people move, salaries climb because everyone is hiring from everyone. All true. But watch where the person who quits actually goes.
Your architect resigns on Friday, and in Bengaluru, she isn't leaving the city, she's walking across the road to one of a thousand other GCCs. The experience stays in town. When you backfill her, you're hiring someone doing the same walk in reverse: already in Bengaluru, already knowing the work.
Run that same resignation in any other city and she leaves the city entirely, the experience goes with her, and your replacement has to be relocated in or trained from scratch. Same exit, completely different cost. You'll lose people faster in Bengaluru and you'll replace them faster, and time-to-refill is the number that actually hits your roadmap, not the attrition rate everyone frets about.
The second is the ecosystem that grows up around a 1000 centres. Everything that serves a GCC has scaled to serve them here; the specialist recruiters who fill niche roles, the contractors who've built a hundred tech campuses, the legal and compliance firms, the Service Providers running everything from build-operate-transfer to embedded delivery pods. In a smaller city, a lot of that you assemble yourself, slowly, while your centre is trying to stand up. In Bengaluru it's already there, waiting to be plugged in.
So for a CXO the calculation is simple. A 1000 centres in one city means that when something goes wrong, someone leaves, a project needs a partner, a campus needs building, the answer is already next door. You're not the first company to need it. You're the thousand-and-first, and everything you need has already been built for the ones before you.
Source: Nasscom-Zinnov GCC Landscape Report 2026
Hover a city to see the real multiple
| Bengaluru | 245+ |
|---|---|
| Hyderabad | 120+ |
| Pune | 105+ |
| Chennai | 75+ |
| India total | 504 |
There are 504 PE-backed or PE-acquired GCCs in India now. Bengaluru has 245 of them, nearly half the country's total, and more than double the next city.
To see why that number matters, you have to understand how a PE-backed company thinks, because it's not like the others.
Most companies setting up a GCC have time. If the first year is slow, they absorb it. If the talent takes a while to come together, they wait. A PE-backed company has none of that room. The exit is already on the calendar, the fund knows, more or less, when it needs to sell, and that clock starts ticking the day the centre opens, not the day it finally hits its stride. Every month lost to a slow hire or a delayed setup is a month subtracted from the only window that matters.
So these are the least patient buyers in the market, and the least forgiving of a wrong call. They can't afford to pick a city and hope it works out, because there's no time to recover if it doesn't. And when you look at where they land, they land in Bengaluru, 245 of them, far ahead of anywhere else.
That's worth reading as a signal. These aren't companies following a trend or picking the city their competitors picked. They're firms whose entire discipline is putting money where the risk is lowest and the return is fastest, on a deadline, with real consequences for getting it wrong. When that crowd concentrates in one place, they're telling you something they've already paid to figure out.
So if your own mandate has a clock on it; a deadline, an integration target, a board that wants results by a certain date, you're in the same position they are, whether or not there's a fund behind you. The people who do this for a living, with the least room for error, have already run the analysis. Half of them chose Bengaluru. You can treat that as a shortcut past a calculation you'd otherwise have to make yourself.
Different companies, different pressures, same answer. And the answer is always about time. Bengaluru doesn't win because of the cost associated with running a centre. It wins because it's the fastest place to set up a GCC and the safest place to have it work. Other cities may ask you to build the ecosystem, wait for the talent, or carry the risk of being early. Bengaluru is where all of that has already been paid for, by the thousand companies that arrived before you.
That's the premium, and that's what it buys.
If your mandate is cost-led execution, a smaller city may serve you better and cheaper. If it's BFSI, Hyderabad has a real claim. If it's automotive, so does Pune. The right call depends entirely on what you're asking the centre to do, and getting that right, before you sign anything, is the difference between a GCC setup that compounds and one you spend years fixing.