GCC or a Vendor: What Standing Up a Capability Centre Takes

A Global Capability Centre, or GCC, is a wholly owned subsidiary a company sets up in India to run engineering, finance, analytics or support work directly — its own entity, its own payroll, its own lease. It sits in the same conversation as staff augmentation or a dedicated offshore team because it solves the same problem — access to India's engineering labour pool — with a different ownership structure. The two aren't always substitutes. Below a certain headcount and above a certain cost tolerance, they are.

We sell the vendor side of this: it-staff-augmentation, dedicated-development-team and offshore-software-development engagements run out of India, plus build-operate-transfer arrangements for buyers who want to start as a vendor relationship and convert to a captive entity later. We have an obvious interest in this answer leaning toward 'use a vendor'. What follows is the arithmetic a GCC advisory firm would also show you — including the cases where it points the other way.

What a GCC actually buys you that a vendor doesn't

The difference is legal and practical. A GCC employs its engineers directly, on its own company registration, under Indian labour law, in a facility it leases or owns. Intellectual property sits with the parent company from day one, with no assignment clause to negotiate. The roadmap, the hiring bar, the tooling, the on-call rotation — all of it is set by the parent, not negotiated with an account manager. A vendor engagement — staff augmentation, a dedicated team, or a fully outsourced project under it-outsourcing — gives you access to the same labour market without the entity, the lease or the compliance calendar. You get people, not a subsidiary.

The practical consequence is control versus overhead. A GCC owns the relationship with every engineer it hires — no vendor markup, no bench risk sitting on someone else's balance sheet, no renegotiation when a contract comes up for renewal. It also owns every statutory filing, every labour dispute, every lease renewal and every line of the India P&L. Build-operate-transfer exists specifically to let a buyer defer that second part: a vendor builds and runs the team under a vendor contract, and the buyer takes the option to convert it into its own entity once the team, and the business case, are proven.

The headcount line where an entity starts paying for itself

Entities carry fixed costs that don't scale down: a managing director, a company secretary, statutory audit, a registered office, HR and finance headcount to run payroll and compliance even for ten engineers. Deloitte's work on GCC economics in India treats a standalone entity as viable once it is supporting a genuinely India-based function, not a handful of roles bolted onto a holding structure (Deloitte India, 2023). In practice, advisory firms size that threshold in the low hundreds of FTEs — a team small enough to run as a vendor engagement usually doesn't clear the fixed-cost bar an entity needs to make sense.

Do the arithmetic yourself rather than take a vendor's word for it: take the entity's fixed annual running cost — leadership, compliance, finance, HR, facilities overhead that doesn't flex with headcount — and divide it by the per-head saving a captive entity gives you over a vendor markup. If that breakeven headcount is above your actual hiring plan for the next three years, the entity loses money relative to a vendor every year you're below it. If your hiring plan clears it within year one or two, the entity wins from year two onward.

The headcount threshold above is illustrative, not a single published breakeven number — multiply your own entity's quoted fixed costs by the vendor markup you'd otherwise pay to find your own figure; don't borrow ours.

What the setup itself costs and how long it takes

Setting up a GCC in India means incorporating a company, registering for GST, PF, ESI and professional tax, opening banking, signing or building a facility, and hiring a local leadership team before a single engineer starts. JLL's guidance on GCC site selection puts the realistic timeline from decision to first hire at six to nine months once the location and legal structure are chosen (JLL, 2023). ANSR's GCC playbook frames first-year setup cost — legal, compliance, facility fit-out and leadership hiring, before any engineering headcount — in the low millions of dollars, which only amortises sensibly against a multi-year, multi-hundred-head plan (ANSR, 2022).

Incorporation itself is the fast part. The World Bank's last Doing Business assessment put the time to register a company in India at under three weeks (World Bank, 2020). The slow part is everything after: the first leadership hire, the facility lease, the first compliance audit cycle, and the twelve months it typically takes a new entity to stop feeling like a start-up and start feeling like an operating site. None of that clock starts until the legal decision is made, and none of it can be compressed by writing a bigger cheque.

The hiring brand you don't have yet

India's GCC market is not an empty field you're entering first. Nasscom's most recent strategic review counts more than 1,700 GCCs already operating in India, employing close to two million people, competing for the same senior engineers, DevOps leads and engineering managers you'll want to hire (Nasscom, 2024). Those incumbents have ten, twenty, thirty years of local hiring brand, employee referral networks and campus relationships. A new entity starts with none of that — it starts as an unknown name on a job board, hiring into roles like devops-engineers, cloud-engineers or full-stack-developers against competitors candidates already recognise.

A vendor doesn't have this problem because it isn't trying to build a hiring brand from zero for one client's logo — it already runs a pipeline across role categories like react-developers, python-developers, java-developers and qa-engineers, sourced against its own reputation, not yours. That's a genuine trade-off, not a sales line: the GCC gets a dedicated, loyal, fully branded team eventually; the vendor gets you staffed next month. If your hiring plan depends on employer branding to hit headcount on schedule, budget a year or two of weaker offers and slower closes before that brand exists.

Running an India site from 8,000 miles away

A GCC doesn't outsource its own operations — someone still has to run the help desk, patch the servers, manage the on-call rotation and keep the network up, the same functions we sell as it-help-desk-outsourcing, system-administration and devops-services. A captive entity either builds those functions in-house, which means more direct hires and more management layers before any product work starts, or it quietly re-hires a vendor for exactly those functions anyway, which narrows the actual difference between a GCC and a vendor engagement down to the engineering headcount alone.

The management load that doesn't show up in the headcount plan is statutory: PF and ESI filings, Shops and Establishment renewals, state-level labour law variance if you expand beyond one city, an annual audit, a company secretary's calendar, and a lease that runs regardless of utilisation. None of this needs daily attention from the parent company, but all of it needs someone accountable, awake, in the right time zone, who isn't an engineer. Skipping that role is the most common GCC failure mode: a site that technically exists but has no one driving it, which behaves like an offshore team with worse support and a bigger fixed cost.

Where a GCC wins outright

None of this is an argument against GCCs — it's an argument for doing the arithmetic before setting one up. A GCC wins cleanly when the headcount plan clears the entity breakeven within two to three years, when the horizon is five years or more, when the work involves product IP you don't want to assign or licence to a vendor contract, and when the parent is large enough to second a general manager rather than hire one cold. Scale is why the market looks the way it does: Nasscom's 1,700-plus GCCs exist because enough companies cleared that bar (Nasscom, 2024), not because an entity is inherently better than a vendor relationship.

• Headcount plan clears entity breakeven within two to three years
• Horizon of five years or more, not a single product cycle
• Work involves IP you won't licence or assign to a vendor
• Parent can second a general manager, not hire one cold
• Already running multiple offshore functions that could consolidate under one roof

The honest comparison, and the middle path

In the US search data we track for this topic, queries about outsourcing or staff augmentation run roughly nine times the volume of queries about setting up a GCC — about ninety a month against ten. That split is a reasonable proxy for where most buyers actually sit: below the headcount and horizon a GCC needs, even if the GCC conversation gets more attention in trade press. Most of the dozen vendor pages a buyer reads before landing here are written by people selling one side of this decision; we're selling the other side, which is exactly why we're naming the cases where it's the wrong one.

The 90-to-10 split is our own search data for the US market, not a published industry figure — treat it as directional, not a market size.

The middle path is build-operate-transfer: start the team as a vendor engagement — a dedicated-development-team or a contract-staffing arrangement — hit the headcount and tenure numbers that would justify an entity, then convert. It costs more than going straight to a GCC if you were always going to clear that threshold, and it costs far less than standing up an entity that never gets there. If your honest three-year hiring plan is under a hundred engineers, if you don't yet have a general manager you trust to run an India site, or if the work doesn't need IP sitting on your own balance sheet from day one, the answer is a vendor, possibly us, and the GCC conversation is premature. If the plan clears those numbers, talk to a GCC advisory firm, not a staffing vendor, and build the entity.

Sahil Dubey

Group Managing Director

Runs engineering and compliance across the group. Twenty years in infrastructure, most of it on the operations side of systems that were already in production when he inherited them.

• CISA
• AWS Solutions Architect – Professional
• Microsoft Certified: Azure Administrator Associate