It is two in the morning in Bengaluru and an alert has fired for a payments system that belongs, on paper, to a team in Frankfurt. The engineer on shift can see that something is wrong. What she cannot see is which upstream service changed, who owns it, or whether the threshold that fired was set for this environment or inherited from a datacentre in another country three years ago.
This is not a story about a badly run centre. It is a story about how quickly the Indian side of these organisations has grown, and how slowly the operational scaffolding around them has caught up.
How Big Has This Actually Become?
The NASSCOM-Zinnov India GCC Landscape 2026 report puts the country at 2,117 global capability centres operating across 3,728 individual sites, employing around 2.36 million professionals and generating 98.4 billion dollars in revenue in FY2026. Two years earlier that revenue figure was 64.6 billion dollars. JLL’s GCC Office Guide 2026 records 31.3 million square feet of office space leased by GCCs across the top seven cities in 2025, the highest share of total absorption yet recorded.
Those numbers describe an estate, not just a headcount. Every one of those sites carries networks, applications, cloud accounts and endpoints that somebody has to keep running.
The same NASSCOM-Zinnov research records something less discussed. GCC hiring volumes fell by roughly 28 percent between the first and second halves of FY2026, with redeployment and AI-supported capability building taking the place of recruitment. The estate is still growing. The number of hands per system is not.
Why Does Inherited Tooling Become a Problem?
Most GCCs do not choose their monitoring stack. It arrives with the mandate, configured by a team at headquarters for the environment headquarters was running at the time.
That works while the Indian centre is executing work defined elsewhere. It stops working when the centre starts owning services outright, which is exactly the shift that has been under way for the last few years. The India team now carries responsibility for uptime, but the thresholds, dashboards and escalation paths were designed around a different topology and a different working day.
Three consequences follow, and they are visible in most centres that have crossed a few hundred engineers.
Alert thresholds reflect the parent environment. A latency limit tuned for a European datacentre behaves differently against traffic patterns originating in Asia, and nobody has the authority to retune it. Engineers learn which alerts to ignore, which is the beginning of a longer problem.
The dependency map lives at headquarters. When something breaks, the Indian team can see the symptom but not the chain. Escalation happens not because the problem is severe but because the information needed to judge severity sits somewhere else, on someone else’s clock.
Local systems fall outside the picture entirely. The India centre almost always runs things the parent does not track. Local network links, campus infrastructure, regional cloud regions, vendor connections. These get watched through a second set of tools, or a spreadsheet, or not at all.
What Changes When Expansion Moves to Tier-2 Cities?
The Union Budget 2025-26 announced a national guidance framework to help states attract GCCs to emerging tier-2 cities, and states have moved ahead with their own programmes while that framework has been under consultation. Karnataka, Maharashtra, Telangana, Gujarat and Tamil Nadu are all competing for the same expansion.
For operations teams this means something specific. A single large campus in Bengaluru becomes a primary site plus two or three smaller ones in cities with less mature connectivity and a thinner local vendor bench. The failure modes multiply, and they stop being uniform. What worked as an informal arrangement when everyone sat on the same floor does not survive distribution across four locations.
This is the point at which IT infrastructure monitoring has to be treated as a decision the India operation makes rather than a configuration it receives. Not necessarily a different product, but certainly a different set of answers about what is watched, by whom, and against which thresholds.
Where Does the Gap Show Up First?
Four signals are worth looking for, and none of them require an audit to spot.
The time to acknowledge an incident is stable but the time to resolve it is climbing. That gap is usually information, not effort. Your engineers know something is wrong and cannot establish what it affects.
Escalations to the parent team are rising faster than incident volume. That indicates a decision-authority problem as much as a technical one.
Somebody on the team maintains a private tracker. A spreadsheet of local circuits, a script that polls a set of hosts, a WhatsApp group that gets used before the ticketing system does. These exist because the official tooling does not cover something, and they are a precise map of the gap.
Nobody can produce a current list of what the India centre actually owns. If that question takes a week to answer, the monitoring picture is unlikely to be better.
What Should a GCC Leader Ask?
Ask who can change a threshold, and how long that takes. If the answer involves a ticket to another country, your alerting will drift out of usefulness and your team will start filtering it out.
Ask what the India site owns that headquarters does not monitor. Write the list down. In most centres it is longer than expected, and it usually includes the things that fail most often.
Ask whether your service dependencies are documented anywhere that an engineer on shift can reach at two in the morning. Documentation that requires a business-hours conversation is not documentation for operations purposes.
Ask what happens at the second and third sites. If the answer is that the same people will handle it remotely, plan for that assumption to break within a year.
The Real Constraint Is Not Talent
India’s advantage in this market has never been in doubt, and the scale figures make the case on their own. The constraint is that operational maturity has been assumed to travel with the mandate, and it does not. It has to be built locally, by the people who will be woken up by it.
The centres that will handle the next phase well are the ones treating their operations tooling as something they own and shape, rather than something that was installed for them.
