What GCC Talent Really Costs in India: A Realistic Guide

Most India GCC business cases are wrong on the day they are approved — not because India disappointed, but because the model compared the wrong numbers. The compare-two-salaries method that dominates board decks omits the three variables that actually move offshore economics: the load factor, the attrition tax and the ramp curve. This analysis rebuilds the cost model the way operators — rather than advocates — build it, with the charts that make each distortion visible.

The idea in brief. India’s cost advantage is real: typically 40–60% below Western fully loaded equivalents, a range consistent across published compensation studies. But it is pyramid-shaped (vast at entry level, thin at the elite top), it must be computed on fully loaded cost (1.4–1.8× fixed salary), and it is protected or destroyed by two operational variables — attrition and time-to-productivity — that dwarf salary-band differences. Model cost per retained, productive employee, and the business case becomes both honest and, usually, still compelling.

Begin with the shape, not the average

India’s salary curve is steeper than Western curves — a structural fact with strategic consequences. At the base, the world’s largest supply of capable graduates (over a million engineering degrees a year, per AISHE) keeps entry-level costs remarkably low. At the top, a globally traded elite — principal engineers, AI researchers, proven centre heads — prices with reference to global markets, not local ones. Between them sits the contested middle where every GCC, consultancy and startup fishes.

The planning implication: any business case built on a single blended rate misrepresents reality in both directions — it understates the bargain at the base and the price of the top. Deloitte’s global shared-services survey series has documented for years that top-quartile operators plan at role-family granularity; the bottom quartile plans in averages and discovers the pyramid later, expensively.

The load factor: from salary to truth

Fixed salary is the beginning of cost, not the measure of it. A credible per-person model adds each block below — and the chart shows why skipping any of them flatters the case:

• Statutory employment costs — provident fund, gratuity accrual, insurances; commonly 12–18% on fixed pay under India’s employment framework.
• Variable pay and equity — standard from mid-level upward; offers without meaningful bonus or stock price themselves out of the top of the funnel.
• Workspace and technology — real estate, equipment, software seats, security, connectivity.
• Enabling functions — HR, finance, IT support, facilities; in-house or partner fees.
• The recruiting engine itself — agency fees (often 8–16% of annual pay per agency hire), assessment platforms, employer branding. Perennially forgotten; invariably paid.

Across city and seniority mixes, disciplined operators land on a fully loaded multiplier of 1.4× to 1.8× fixed salary. Build your own stack from your own quotes — but if your model’s multiplier starts with a 1.1, it is a brochure, not a plan.

The attrition tax — the variable that dominates

Replacing a mid-level engineer costs far more than a recruiting fee: months of vacancy, months of successor ramp, knowledge loss, and load on the remaining team. Model it explicitly and a striking arithmetic appears:

Read the right-hand bar carefully: at 25% attrition, the same team costs roughly 38% more per retained, productive employee than at 10% — before counting any morale or quality effects. That relationship explains most of the gap between GCCs that celebrate their economics and those that quietly renegotiate them; it also reprices “soft” investments — manager training, career ladders, engagement — as line items with hard returns. Our attrition deep-dive reconstructs the underlying rates from public disclosures; the planning range for a well-run centre is 10–15%, which means the difference between competent and careless management is, in cost terms, the entire India advantage at the margin.

The ramp curve — the silent variable

An unfilled seat costs nothing; a filled, unproductive seat costs double — salary out, output absent. Time-to-productivity in a new centre ranges from six weeks to six months for identical roles, depending almost entirely on onboarding design and knowledge-transfer discipline (our cross-time-zone onboarding article gives the design). Business cases assume day-one productivity; operations deliver a curve. The gap between assumption and curve is where first-year economics go to die — and it is a designable variable, not a fixed cost. Add a ramp column to the model: months to standard output, by role family, with an owner accountable for shortening it.

Where companies overpay — and underpay

Overpayment is usually panic-shaped: losing a bidding war and buying the next candidate at any price; importing HQ salary bands unexamined (“it’s still cheaper than home” is how bands inflate 30% above local market); reflexively matching counter-offers, which teaches your own compensation structure to leak (the counter-offer evidence is collected in our notice-period analysis). Underpayment is subtler and often costlier: a weak employer brand forces cash premiums a stronger brand would never pay; a discount centre-head hire produces an expensive centre (the mis-hire economics get their own article); and under-investing in selection quality buys attrition and mediocrity at scale — the most expensive savings available, as the selection-validity literature (Schmidt & Hunter, 1998) has quantified for decades.

Sensitivities the model must carry

Intellectual honesty demands ranges, not points. Compensation cycles: the 2021–22 boom repriced hot skills 20%+ in quarters — visible in the investor disclosures of TCS, Infosys and peers — and cooled just as fast; build ±15% bands on contested roles. City and domain variance: role-level differences within a city exceed averages between cities (hence role-city mapping in our location analysis). Currency: a multi-year model silently assuming a static rupee is making an unpriced currency bet; the RBI’s published history suggests planning with a gentle depreciation assumption rather than none.

The metric that keeps everyone honest

One number reconciles all of the above: cost per retained, productive employee-year — fully loaded cost, divided by expected productive months, adjusted for attrition probability. It is harder to compute than a salary ratio and worth the trouble: it makes retention investments legible to finance, exposes false economies in selection and onboarding, and lets two cities or two operating models be compared without rhetoric. Centres managed on this metric make visibly different decisions — senior-weighted founding teams, designed onboarding, proactive band corrections — and their unit economics show it. Centres managed on cost-per-head rediscover, annually, why the cheap decision was expensive.

A modelling agenda

1. Build the pyramid: role-family salary bands per target city, from live data, refreshed quarterly.
2. Apply your true load build-up; challenge any multiplier below 1.4×.
3. Add attrition and ramp as explicit, sensitivity-tested variables — never footnotes.
4. Adopt cost per retained employee as the governing metric from day one; report it to the board alongside headcount.
5. Re-baseline annually; the market will have moved, and honest models move with it.

Offer dynamics: where models meet the market

Cost models are static; offers are negotiated. Four dynamics of the Indian compensation conversation that models should anticipate rather than discover:

• The CTC vocabulary. Indian offers speak in CTC — “cost to company” — an annual figure bundling fixed pay, variable pay, statutory contributions and sometimes one-time components. Candidates compare CTCs; your finance team models fixed-plus-load. Translate deliberately in both directions, and present offers with the fixed/variable split explicit — sophisticated candidates discount opaque CTCs, correctly.
• The hike convention. Moves are conventionally negotiated as percentage hikes on current CTC — the market’s anchor, whatever your banding logic. A candidate at an under-market employer expects 25–40% to move; one already at market expects less but anchors harder. Bands survive this convention only when offers are framed against role value, with the hike as an outcome rather than the unit of negotiation.
• Variable pay is a signalling instrument. A high-variable structure reads as risk at junior levels (where cash flow matters) and as confidence at senior levels (where upside matters). Mirroring seniority in the mix — modest variable early, meaningful variable plus equity at leadership — prices better than a uniform formula.
• Equity literacy is bimodal. Product-company and startup alumni price options and RSUs fluently; services-industry candidates often discount them to zero. Where equity is a real part of the offer, the education is part of the recruiting — a one-page, numbers-worked explanation routinely moves acceptance decisions at no cash cost.

The common thread: India’s compensation market is highly literate and convention-bound at once. Offers designed inside the conventions — CTC-translated, hike-aware, seniority-mirrored — buy the same talent for visibly less than offers that fight the market’s grammar.

A worked model: the 100-person centre

Numbers discipline arguments, so here is the skeleton of a realistic 100-person centre model — composite values, indicative of the shape rather than any single market quote:

Apply the load stack (say 1.55× on this mix), then run the two operational scenarios: at 12% attrition with designed onboarding, the model’s cost per retained productive employee lands roughly 45–50% below the Western baseline — the promise, delivered. Rerun it at 22% attrition with laissez-faire onboarding and the identical salary structure delivers barely 30% — and that before the unmodelled costs (missed deadlines, HQ trust erosion) that attrition drags behind it. Same city, same pay bands, same people on day one: a fifteen-point economic swing produced entirely by management choices. This is the arithmetic that should decide where the next dollar of centre budget goes — and it usually says “retention machinery,” not “salary escalation.”

One more worked observation: in this model, halving agency dependence via referrals and an in-house sourcing engine (realistic by year two — see the founding-cohort analysis) saves roughly as much as shaving 4% off every salary — without the quality and morale costs that salary-shaving inflicts. The cheapest euros in the model are hiding in the recruiting engine, not the pay bands.

Questions finance teams ask

“What exchange-rate assumption should we use?” A gentle depreciation trend rather than a static rate — consult the RBI’s published series and your treasury’s view. Historically, moderate rupee depreciation has partially amplified dollar-based savings, but building a case that requires currency tailwinds is speculation wearing a spreadsheet.

“How do these costs trend over time?” Indian tech wage inflation runs above Western rates in normal years — high single digits, spiking in booms (the FY22 disclosures of the listed majors show the extreme). The advantage erodes slowly at aggregate level and is refreshed continuously by the graduate pipeline; model a modest annual convergence and let the productivity curve — seasoned teams, owned mandates — offset it, as it has for the sector historically.

“Are the savings real after all overheads — honestly?” At competent scale, yes: the 40–60% range survives full loading in every credible study and in our engagement experience. What does not survive is careless execution — the attrition scenario above is the honest answer’s other half.

“When does the centre break even against setup costs?” Typical pattern: modest net savings in year one (setup, sub-scale overheads, learning curve), full run-rate economics from months 18–30. Boards briefed on the J-curve fund it calmly; boards promised instant arbitrage cancel at the dip — usually just before the curve turns.

“What single number should govern the centre?” Cost per retained, productive employee-year — reported quarterly beside headcount and attrition. It is the one metric that makes every trade-off in this article legible in money.

Beyond CTC: what Indian talent actually values

A cost model that sees only cash mis-prices the market, because several of the strongest levers in Indian compensation are cheap in money and expensive only in intention:

• Learning as compensation. In survey after survey of Indian tech professionals, growth and skill development rank at or near the top of employer-choice criteria — above incremental cash for a large fraction of the market. A funded certification path, conference budget or genuine AI-upskilling program (article 30) buys loyalty that a matching cash amount cannot.
• Title trajectory. Career-ladder visibility — published levels, honest promotion cycles — is a compensation instrument in a market where the next role is the point. The build-vs-buy analysis (article 25) shows the same lever from the employer’s side.
• Brand on the CV. Working for a respected global name carries durable career value here; an unknown entrant (article 23) effectively pays a brand discount in cash until its reputation capitalises. Budget the employer-brand build as the compensation investment it is.
• Flexibility with integrity. Hybrid arrangements that do not quietly penalise users have become a baseline expectation in the post-2020 market — and a differentiator where competitors enforce rigid mandates.
• Insurance that includes parents. A specific, telling detail: family health cover extending to parents is among the most-valued benefits in Indian packages — a modest premium with outsized signalling in a family-centred culture.

None of this replaces fair cash; all of it changes what fair cash needs to be. The complete offer — money, growth, brand, flexibility, family — is the unit the market actually prices.

Methodology & data notes

The 40–60% advantage range, statutory-cost percentages and load multipliers reflect commonly published compensation-study findings and HexGn’s market observation; the attrition-cost chart is a transparent model (recruiting fee + vacancy months + ramp months, at stated attrition rates) rather than a survey result. Where figures are marked illustrative, the claim is the relationship, not the point value. Primary sources below publish current-year numbers.

References & further reading

• NASSCOM — industry compensation and workforce statistics
• Deloitte — Global Shared Services and Outsourcing survey series
• AISHE — graduate-supply statistics
• TCS and Infosys investor disclosures — published attrition and wage-cycle data
• Reserve Bank of India — exchange-rate series
• Schmidt & Hunter (1998) — selection validity and its economic utility

HexGn maintains live compensation and supply intelligence across India’s hubs and models centre economics exactly this way — because the business case you approve should be the one you experience.