Why a GCC build playbook beats a six-month evaluation
The slowest part of standing up a Global Capability Center is rarely the hiring. It is the deliberation — the months spent debating cities, entity structures, and headcount before a single recruiter is briefed. By the time a decision lands, the market has moved and the budget cycle has closed.
A productized GCC build playbook replaces that drift with a fixed sequence. The work splits cleanly into three phases — Assess, Build, Operate — and a disciplined team runs them in roughly 90 days. Maturity, as Zinnov puts it, is now a day-one choice: you decide what your center owns before you hire anyone, not three years in. The cadence below is deliberately boring, because predictable is what gets a pod live.
Phase 1 — Assess (weeks 1–2)
The first two weeks are about decisions, not recruiting. The goal is a signed brief that your CTO, your CFO, and the build partner can all point to when scope starts to creep.
- ICP fit and mandate. What will this team own? Not "extra capacity" — a real product surface, platform layer, or global function. The Cognite turnaround is the cautionary tale: a dispersed extension team of one or two people per global squad delivered no clear ROI until India leaders owned whole product pillars end to end. Decide the mandate before the math.
- City. Bangalore, Hyderabad, or Chennai for depth and density; a Tier-2 city like Coimbatore or Ahmedabad as a margin and retention lever, running roughly 25–30% cheaper with lower attrition.
- Site-leader profile. Write the spec for the anchor hire now — a real engineering leader, increasingly dual-mandated to run the India site and own a global function.
- The signed brief. Mandate, headcount band, comp ranges, target start dates, and success metrics, agreed in writing.
Get the mandate right here and the rest of the playbook compounds. Get it wrong and you will be re-recruiting in month four.
Phase 2 — Build (weeks 2–5)
Building overlaps the back half of Assess. Two things make the timeline work: hiring the leader first, and running on an employer-of-record (EOR) from day one so you do not wait on entity incorporation to make offers.
Site leader first
The site leader is the single highest-leverage decision in the build. A GCC succeeds when it owns something real and is led by someone senior — so the anchor hire comes before the pod, not after. The leader then helps close the team, sets the bar in interviews, and becomes the reason strong engineers say yes. See hiring a GCC site leader for the full profile and scorecard.
The first three hires
Discipline on the first three hires sets the ceiling for everyone who follows. They are your culture carriers and your interview loop, so resist the urge to fill seats fast:
- Hire 1 — the site leader. Owns the mandate and the bar.
- Hire 2 — a senior IC or tech lead. Proves the work can be done to your standard in India and anchors the engineering culture.
- Hire 3 — a second senior, or a function-specific lead (platform, QA, support) depending on the mandate.
India's hiring market is fast and competitive — interview-to-offer ratios run around 7:1 and strong candidates often hold three or four offers at once — so a partner with a live recruiting engine matters. The realistic shape: first hire within about four weeks, a full pod of 5–15 by roughly week five. EOR from day one means those people are employed, paid, and compliant immediately, with the entity question deferred until it actually pays for itself.
Phase 3 — Operate (week 6+)
From week six the pod is live and the model shifts. The build partner runs HR, payroll, compliance, and facilities; you direct the work. This is the managed-GCC arrangement — a branded team you control without the administrative load — and it is the difference between a second home for engineering and a vendor relationship. How it works walks through the operating split in detail.
30-60-90 onboarding
Every hire runs the same ramp so output is predictable:
- Day 30: environment set up, first PRs merged, embedded in standups and the team's rituals.
- Day 60: owning a service or feature area independently; on-call and review rotations begun.
- Day 90: fully productive against the same expectations as a peer in your home office — measured on outcomes, not hours.
Governance cadence
A light, fixed rhythm keeps the center aligned without smothering it:
- Weekly: site leader and home-office manager 1:1, covering delivery and blockers.
- Monthly: hiring funnel, attrition, and ramp reviewed against the signed brief.
- Quarterly: mandate review — is the team owning more, and is comp still competitive?
The metrics that tell you it is working
Track a short list and the rest takes care of itself.
| Metric | Target | Why it matters |
|---|---|---|
| Time-to-pod | < 6–8 weeks | Speed is the whole point of a playbook; slow builds bleed budget and momentum. |
| Annual attrition | < ~12% | Below the IT-services benchmark. Ownership and a real mandate retain; a ticket queue does not. |
| Net retention of the engagement | NRR > 120% | A healthy build expands. When the work lands, you add scope and headcount rather than wind down. |
When Build-Operate-Transfer belongs in the plan
For most first-time builders, a managed GCC is the right operating state indefinitely. But if owning the entity outright is a strategic goal — for IP, valuation, or control — bake Build-Operate-Transfer into the playbook from the start rather than bolting it on later. BOT deals fail at the transfer, not the build, so the time to negotiate pre-agreed triggers, step-down fees, day-one novation terms, and a fixed transfer price is in the brief, before anyone is hired. Our guide to Build-Operate-Transfer covers how to de-risk it.
Maturity is a day-one choice. Decide what your center owns — and whether you will own the center — before you make the first hire.
The economics make the case either way: a representative 10-person pod runs roughly $744K a year all-in in India versus about $2.58M in the US — a ~71% gross saving. We work that math line by line in what a GCC in India costs. Cost is the CFO's confirmation, not the reason. The reason is a senior, owned team shipping on your roadmap within a single quarter.
Frequently asked questions
How long does a GCC build actually take?
With a productized playbook, about 90 days end to end: roughly two weeks to assess and sign the brief, three weeks to build, and live operations from week six. The first hire typically lands within about four weeks and a full pod of 5 to 15 by around week five.
Who should be the first hire in a GCC?
The site leader, always. A GCC succeeds when it owns something real and is led by someone senior, so the anchor hire comes before the pod, sets the interview bar, and helps recruit the rest of the team.
Do I need to set up an Indian entity before hiring?
No. Running on an employer-of-record from day one lets you make compliant offers immediately and defer incorporation until it pays for itself, often only when you are heading toward a Build-Operate-Transfer.
What metrics show a GCC build is working?
Three signals: time-to-pod under six to eight weeks, annual attrition below roughly 12 percent, and an expanding engagement with net retention above 120 percent as the team earns more scope. Track those and the rest follows.
What is the difference between Operate and Transfer?
In Operate, the partner runs HR, payroll, compliance, and facilities while you direct the work, indefinitely if you like. Transfer, the T in BOT, moves the entity and team to you, and should be negotiated up front because BOT deals fail at the transfer, not the build.