Location strategy is really a talent-supply decision

Location decisions tend to be owned by the people who model rent, incentives, and cost per seat, and on those axes a tier-2 city often wins cleanly. What that model leaves out is supply. If the capability the centre needs is thin on the ground there, the saving on real estate is spent several times over on time-to-fill, on relocation to attract seniority the local pool can't provide, and on the roles that simply sit open while the ramp waits.

The pattern is most acute for the senior and specialist seats. Volume roles travel well; deep judgment and scarce skills tend to concentrate in a few markets, and a location chosen for rent can be structurally unable to supply them at all. A seat that's cheaper to house but takes twice as long to fill, or never fills at the seniority required, was not the cheaper seat.

The concentration isn't anecdotal. India Brand Equity Foundation reports that tier-1 cities house roughly 94% of the country's GCCs, and the Zinnov–Nasscom 2026 landscape counts about 2,117 centres employing some 2.36 million people. The talent has clustered because the capability clustered first. A cost model that treats the map as open ground, any city will do if the rent is right, is arguing with a decade of where the work actually went.

The better sequence inverts the usual one. Begin with the capability the centre must own, the specific skills and seniority the charter demands, and ask where that talent already exists at depth. Then let cost act as a tie-breaker between viable supply markets, rather than the opening filter that quietly rules out the only places the talent lives. Sometimes the answer is still the lower-cost city; often it's a split, with the deep roles in a primary market and the scalable ones in a cheaper satellite.

Reading where a capability actually lives is a specific exercise, not a vibe. It means looking past headcount totals to the density of the exact layer you need: how many people in that market have already done the senior version of this role, at a comparable scale, in a comparable sector. A city can have a large technology workforce and still be shallow on, say, quantitative risk leadership or medical-affairs heads, because depth is skill-specific and seniority-specific, and the aggregate number hides both.

This is why location belongs partly to whoever owns talent, and not solely to whoever owns cost. The pharma and BFSI centres that get it right treat the map as a supply question first and a spreadsheet second, and they build multi-site strategies that put each role where its market actually is, because one cheap location can't supply the whole charter.

The tier-2 case is real, and it's getting realer. EY's work on how eastern and tier-2 India can lead the next GCC wave describes genuine, government-backed momentum: infrastructure, incentives, and a widening base of graduate talent are pulling volume and mid-level work into cities beyond Bengaluru and Hyderabad. For the scalable layer of a charter, engineers, analysts, operations roles hired in numbers, a tier-2 hub can be an excellent decision on both cost and supply.

The trade changes at the top of the org chart. The senior and specialist bench, the functional leaders, the scarce-skill principals, the people who have already run the thing at scale, is far more concentrated, and it concentrates in tier-1. A tier-2 location can supply the base of the pyramid comfortably while being structurally unable to supply its apex from the local pool. That's simply the shape of how deep, senior capability accumulates: slowly, in the markets that had the work first.

The AI layer makes the concentration vivid. ANSR reports that Fortune 500 GCCs in India already hold more than 126,600 AI-aligned professionals, about 22.5% of the country's AI talent pool, and that pool sits overwhelmingly in the established hubs. Against a Deloitte–Nasscom projection of AI-talent demand exceeding 1.25 million by 2027 with only around 16% of IT professionals AI-skilled, the senior end of that pool is the scarcest resource in the market. You don't stumble onto it in a location chosen for rent.

Most location models optimise for the average seat. The centre lives or dies on the scarce one. A charter can be 90% staffable in a given city and still fail if the remaining 10%, the function head, the principal architect, the leader who sets the standard everyone else works to, simply isn't available there. Ramp is gated by the seat that won't fill. That's what makes senior availability the binding constraint: it's the part of the map with the least give.

The macro backdrop tightens the constraint further. ManpowerGroup's Q2 2026 survey put India's Net Employment Outlook at +68%, the strongest since 2008, while 82% of employers reported difficulty finding the skills they need. Strong hiring intent meeting thin senior supply is precisely the condition under which the leadership seat becomes the bottleneck, and choosing a location that's shallow at that layer converts a market-wide shortage into a self-inflicted one.

In our own mandates, the location that reads “cheaper” on the cost model frequently carries a materially longer time-to-viable-shortlist at the senior layer, often measured in extra months rather than weeks, because the corrective is relocation, and relocation is slow and expensive. The honest way to price a location is to price its scarcest seat, then decide whether you build that capability locally, borrow it from a deeper market, or split the charter so the scarce role sits where it can actually be filled.

The tool that replaces the cost-first spreadsheet is a talent availability map: the charter broken into role layers, each layer scored on how deep the supply runs in each candidate market. It goes past headcount totals to ask, for every layer, how many people in this location have already done this exact role at this scale and seniority, and how that depth compares across the markets on the shortlist.

The map is built from signal the aggregate reports miss. Nasscom ranks India first globally on AI-skill concentration, but a national ranking tells you nothing about whether the specific senior AI leader you need is available in Pune versus Bengaluru. That resolution comes from live evidence: where comparable senior roles are actually being filled, how long they take, how many viable candidates surface per mandate, and how much of the bench is home-grown versus borrowed. It's a search firm's funnel, read as a supply instrument, that turns a national average into a city-and-sector answer.

Read that way, the map does three things the cost model can't. It flags the layers where a location is deep enough to hire at pace. It exposes the layers where the local pool is too thin and relocation or a second site is the honest answer. And it tells you which single seat, usually a senior one, should decide the whole location, because it's the one the market can least easily supply.

The whole argument resolves into a sequence you can actually run. First, name the capability, the specific skills and seniority the charter must own, layer by layer. Second, map supply depth for each layer across the candidate markets, using live hiring evidence rather than headcount totals. Third, let cost enter, but only as the tie-breaker between markets that already clear the supply test, never as the opening filter that rules out the places the talent lives.

That sequence usually produces a multi-site answer rather than a single winner, and that's the point. The scalable base of the charter goes where cost and supply both work, often a tier-2 hub with a widening pool. The specialist and senior layers go where their thin, tier-1-concentrated supply actually sits. One cheap location pretending to serve the whole charter is the failure mode; a deliberate split that puts each role where its market is, is the design.

So the governance move is to stop letting the cost model open the conversation. Bring it in third, after the capability is named and the supply is mapped. A centre sited that way pays a little more on rent and saves it many times over on the seats that would otherwise sit open, because it was built where the talent already is.