For most of the last two decades, "Global Capability Centre" meant hundreds or thousands of people, a multi-year build, a real estate team. That's not what's driving the fastest-growing part of this story anymore.
The fastest-growing pattern right now is the opposite: smaller, senior-heavy teams 5 people, 15 people, 25 people built around one specific capability, hired through an Employer of Record or a lean local entity, running in weeks instead of years. The label the sector has landed on is "nano" or "micro" GCC, and it exists because the old assumption that a serious India presence requires serious scale stopped being true.
What's actually driving this, and it isn't cost.
The instinct is to assume this shift is about squeezing costs at the smallest possible scale. That's part of the story, but it undersells what's happening. Companies building nano and micro teams in India aren't doing it to shave a budget line. They're doing it because a focused, senior team of 10 or 15 people, with direct ownership of a specific problem, ends up shipping more than a much larger team with diffuse responsibility ever would.
That's the same instinct driving India's broader Emerging Enterprise GCC segment now more than 610 centres employing over 462,000 professionals, growing at a projected 14% a year through 2030 where operators describe the model as "lean, high-impact" rather than large-scale. Senior-first hiring instead of a junior pyramid. A specific capability AI, data engineering, a product line instead of broad functional coverage. The nano and micro GCC is that same instinct taken to its smallest useful size.
A focused, senior team of ten or fifteen, with direct ownership of one problem, ships more than a much larger team with diffuse responsibility ever would.
Three things made this viable now, when it wasn't a few years ago. The infrastructure to hire small and fast exists as a service an Employer of Record can onboard a first hire in as little as 48 hours, no entity of your own required. The economics work at small scale: reported cost comparisons for India-based technical talent routinely show 40% to 70% savings against equivalent US-based hiring, and a 5-person team captures nearly the same per-head economics as a 50-person one. And the talent itself has kept moving upmarket India produces roughly 2.5 million STEM graduates a year and has built one of the largest AI-capable technical workforces outside the US and China.
Where the big advisory firms still aren't looking.
Search for guidance on this and you'll land on pages from CBRE, PwC, EY, Deloitte, and KPMG current, well-produced, and built almost entirely around enterprise-scale entry. CBRE cites an 80% Fortune 500 client base; PwC's language centers on "multinational banks." None of them meaningfully address a company hiring 5 or 15 people through an EOR or a lean entity. That's not a flaw in their content, it's simply not who they're built to serve. Which means the fastest-growing, most accessible way into India right now is also one of the least-covered by the advisory content most companies find first.
The honest starting question isn't "how many people can we afford to hire." It's "what specific capability do we actually need built, and by whom." That reframes the whole decision: match the size of the commitment to the size of the question you're trying to answer, not to a headcount number that made sense for a different kind of company.
Once you've sized the ambition, the real decision is EOR or your own entity.
Deciding that a nano or micro-scale team is the right approach only gets you to the starting line. The next decision: hire through an Employer of Record, or set up a lean entity of your own matters more than most companies expect, because unwinding the wrong choice later costs real time and money, even at a team of 10.
An Employer of Record legally employs your India-based team on your behalf payroll, statutory compliance, benefits while you direct the work day to day. No entity registration, no local HR function, no multi-month setup. A first hire can be onboarded in anywhere from 48 hours to a few weeks, typically priced as a flat monthly fee per employee or a percentage of payroll. What it doesn't give you is the same structural control a company gets from its own legal presence, equity, benefits designed exactly as you'd want, a team identity that isn't mediated by a third party's employment relationship. For a short-duration pilot, or a team under roughly 10 people, that trade-off is usually the right one.
A nano or micro GCC typically 10 to 25 people means setting up your own lean entity rather than employing through a third party. Reported setup costs vary by scope and provider, generally landing somewhere between roughly $15,000 and $100,000 to incorporate and register, plus an ongoing annual compliance overhead in the range of $8,000 to $20,000. What you get for that cost is direct control over equity and compensation design, over IP and data handling, and critically, over retention: a person hired directly, with a real local entity behind them, is building a career with your company. A person employed through an EOR, however well the relationship is run, sits one layer removed from that.
Mapping the choice to what you're actually trying to learn.
If you don't yet know whether the capability belongs in India at all, start with EOR. The entire point of this phase is to answer that question with real people doing real work, without a legal entity to unwind if the answer turns out to be no. If you already know the answer is yes and you're building for the next few years, a nano or micro GCC is usually the better foundation, even at a small size, because the retention and control benefits compound from day one. If you're in between, a hybrid path works well in practice: keep hiring through the EOR while entity registration runs in parallel, then migrate the team once the entity clears, typically around month six.
The cost comparison between the two models turns out not to be the deciding factor; the 40% to 70% savings against US-based hiring shows up almost identically either way. The real difference is control and durability, and it's worth treating it as that decision rather than defaulting to whichever model a vendor happens to sell.
The most common misstep isn't picking EOR when a micro GCC would have been better, or the reverse both are recoverable. It's not thinking about the choice at all, and ending up with a structure that fits a vendor's business model better than it fits yours. A team clearly headed toward 20 people in one year shouldn't spend that whole year on an EOR arrangement priced for uncertainty it no longer has.
A few common questions.
Is a 5-to-10-person team really enough to call a "GCC"? The label matters less than the function. What defines these centres isn't headcount, it's a dedicated, India-based team with real ownership of a specific capability, managed with the same intent as a much larger centre, just sized to the actual need.
Can we run EOR and a micro GCC at the same time? Yes, and it's a common transition pattern new hires go through the EOR while the entity is being registered, then the whole team migrates once it's active. Getting that transition timed correctly, without disrupting contracts or momentum, is specific enough to your situation that it's worth mapping properly rather than winging it.
Is EOR more expensive in the long run than setting up our own entity? The crossover point moves around depending on provider pricing, retention patterns, and how fast the team is actually growing. Treating it as a fixed headcount or timeline rule of thumb is exactly how companies end up switching models later than they should have.
Does starting small limit how big this can eventually become? Not inherently but whether it stays a sequencing choice or quietly becomes a ceiling depends on whether anyone planned the transition points in advance, or is just waiting to see what happens.
Espresso Shot Consulting works with companies specifically at this scale, sizing a nano or micro GCC or EOR arrangement, and the transition between them, to match a real growth goal rather than a generic template built for a much larger company.
