Follow one decision. A center wants to change how a process it has run for four years actually works: a sequencing change, well inside its competence, invisible to any customer, worth perhaps two days of saved effort every week, forever. The team lead agrees in an afternoon. Her director agrees the same week. Then the decision leaves the building. It travels to a process owner in another time zone, who asks for a summary. The summary waits eleven days for a slot on a governance call. The call defers it, pending one more stakeholder. Somewhere around week six, the decision comes back approved, unchanged, and slightly embarrassed.
Nothing in that journey was unreasonable. Every person along the route did their job. And yet a center full of capable people just spent six weeks obtaining permission to do something it was already trusted to run. Multiply that journey by every improvement, every hire above a certain band, every vendor above a certain amount, every exception to every policy, and you have the second surface where a center's strain shows: the work moved to the center, and the authority to run it did not.
The arrangement began honestly, and it helps to remember that, because the honesty is what makes it so durable. When a center is new, it is unproven by definition. The enterprise is moving real work to people it has mostly never met, in a country its leadership may have visited twice. Approval is how trust gets built under those conditions. Every decision that travels up and comes back correct is a deposit in an account, and in the early years, the deposits are the point. I have sat in those early governance calls, and they were good for us. They taught the enterprise our judgment, and they taught us the enterprise.
The trouble is what happens next, which is usually nothing. Trust earned is supposed to become authority granted, and often it does not, because there is no mechanism through which it would. The approval rights were written into charters, delegation matrices, and system workflows in the founding years, by people solving the founding problem. Trust, meanwhile, accumulated informally, in the judgment of individuals who watched the center perform. The formal thing persisted because formal things do. The informal thing never got written down. Ten years on, a center can hold a decade of flawless deposits in an account nobody remembers how to draw from.
So the work keeps moving in, because moving work is easy and the center keeps earning it. The accountability moves with the work, because accountability travels with delivery automatically. And the decision rights stay where they were on day one. The result is the specific, quietly corrosive condition this essay is named for: a center answerable for outcomes it cannot fully decide, led by leaders who hold the title without the levers.
Consider what the title without the levers does to a good leader, because this is where the second signal from the first essay in this series comes from: strong people leaving without quite saying why.
A leader hired to run a thousand-person operation discovers, over her first two quarters, the actual perimeter of her role. She owns the outcome of the annual plan but a committee elsewhere owns the budget that funds it. She owns the talent strategy but approvals above a certain band travel. She owns the delivery commitments but the process standards that determine how delivery works are set in rooms she can request to join. Each individual constraint has a rationale. Together they describe a job that is accountable for everything and decisive about little, and the best leaders are exactly the ones who cannot live inside that description for long. They do the math quietly and leave gracefully, and the exit interview records a better opportunity, which is true, and incomplete.
I watched this discovery happen to more than one leader during my years inside, and the moment it lands is always quieter than you would expect. It is rarely a confrontation. It is a Tuesday, somewhere in the second quarter, when a leader realizes the meeting she is preparing for is one she would simply have decided in her previous role. The preparation itself is the tell: the pre-reads, the pre-alignments, the deck versioned nine times. She has begun spending her judgment on persuasion instead of decisions, and she is good enough to notice the difference. The ones who stay longest are often the ones who stop noticing, which is its own kind of loss.
Then there is the matrix. Dual reporting is a fact of life in any global enterprise, and done deliberately it is harmless. Done by accretion, it produces the question I have heard asked in one form or another in every center I have known: when my two bosses disagree, who actually counts? Where the answer is clear, the matrix is furniture. Where it is unclear, every disagreement between distant executives gets settled locally, by a person three levels down, choosing which instruction to quietly deprioritize. That person is doing political work that the design should have done, and doing it without the standing to do it safely.
Centers do not simply endure this arrangement. They adapt to it, and the adaptation deserves its own attention, because it looks like health and behaves like debt.
What grows up around a slow formal channel is a fast informal one: the exceptions economy. A center leader who has spent a decade building relationships learns which calls to make, which sponsor to text, which approval can be treated as retroactive if the outcome is good. Work moves. Dates hold. From the outside, the center looks empowered, because things happen at a speed the delegation matrix cannot explain. From the inside, everyone senior knows the truth: the speed is personal, secured by specific individuals spending specific relationship capital, transaction by transaction.
Two things are wrong with this, and both are structural. First, an exceptions economy prices out the next generation. The workaround routes are owned by the people who built them; a newly promoted director inherits the formal channel at its full six-week length, plus the unspoken expectation of moving at the informal channel's speed. Second, and this is where authority quietly becomes a continuity problem, the center's real decision rights now live in a handful of relationships. When those individuals leave, the delegation matrix does not change by a single line, and the center's effective authority drops overnight. Nothing on any chart records the loss. Everything in the following year reflects it.
Here is what makes this surface especially interesting right now: at the level of the ecosystem, the authority is actually moving. Nasscom and Zinnov's FY2026 landscape report documents what it calls the migration of enterprise authority to India-based leaders as one of the defining structural shifts of the year. The trailing five years produced more than 6,500 global roles held from India by 2024, a count that grew at roughly forty percent annually, and Zinnov's analysis of the largest centers finds nearly two-thirds of their heads now carrying dual mandates, running India operations while leading a global portfolio in product, engineering, or an equivalent domain.
The same report's framing carries a sentence worth quoting exactly: "The ecosystem is ahead. The org chart is behind." That is the industry's own flagship study describing the gap this essay describes. At the headline level, authority is migrating to India. Inside any individual center, the founding delegation matrix may not have been touched since the year the center opened. Both facts are true at once, which is exactly why the averages are dangerous. A center's leadership can read the ecosystem story, recognize their center in it, and never check whether their own charter has moved an inch.
The tells, inside a single center, are unglamorous and specific.
A decision the center could make in a day takes six weeks to travel.
Budget and hiring authority that exists on paper and expires in practice.
A person with two bosses and no settled answer to which one counts.
The obvious cost is speed, and it is real: decisions that queue across time zones arrive late, and late decisions compound. But speed is the smallest of the costs, and the easiest to mistake for the whole problem, which is why so many centers respond with escalation matrices and fast lanes. Those treat the queue. They do not treat the reason the queue exists.
The deeper cost is what the arrangement teaches. An organization learns from where its decisions get made. A center whose calls travel upward for years learns, in its bones, to package rather than decide: to prepare the perfect recommendation deck instead of making the call and owning it. This becomes a skill, then a culture, then a ceiling. When a global mandate finally lands, the enterprise looks for leaders practiced at deciding under accountability and finds a center full of people expert at seeking approval brilliantly, and concludes, unfairly, that the talent is not ready. The arrangement produced the very evidence used to justify the arrangement.
And there is a cost the enterprise pays without noticing: judgment it already bought and does not use. The center's leaders are closest to the work, the market, and the people. Every decision that travels past them to someone farther away is made with less information by design. Over a thousand decisions a year, that is not a governance preference. It is a standing tax on decision quality, levied by geography, paid in outcomes.
None of this indicts the founding arrangement. Approval was the right architecture for an unproven center, and the centers that skipped it usually paid for the shortcut. The indictment, if there is one, is of the silence afterward: the absence of any moment at which the enterprise asks what its own trust, demonstrably earned, should now formally become.
The question for a leadership team, on this surface, is therefore precise. Not whether the center is trusted; the work flowing in answers that annually. The question is whether the trust has been converted: whether the delegation matrix, the budget authority, the hiring bands, and the matrix reporting lines have been deliberately revisited since the center proved itself, or whether the center is still operating under the paperwork of its own probation. Most leaders can answer that in one honest minute. What they usually find is that nobody ever ended the probation, because nobody's job was to notice it was over.
Conversion, where I have seen it done well, is unglamorous and explicit. It is a working session, not a ceremony: the delegation matrix on the table, line by line, with one question asked of each line. What evidence would change this, and has that evidence already arrived? Routine decisions come home first, because they are the volume. Budget and hiring authority follow, with thresholds set to the center's present scale rather than its founding one. The matrix reporting lines get the one sentence they always needed: who ultimately decides when the two disagree. None of it requires courage so much as it requires an owner, someone whose job, for one quarter, is the paperwork of ending a probation the center passed years ago.
Nasscom & Zinnov, GCC Value Orbit: From Delivery Engine to Enterprise Nerve Centre, India GCC Landscape Report FY2026, July 2026. zinnov.com
Nasscom & Zinnov, India GCC Landscape Report: The 5-Year Journey, 2025 (global roles held from India and their five-year growth). nasscom.in
Zinnov, The 5 Shifts That Defined India's GCC Story in 2025, 2026 (dual mandates among Mega GCC heads). zinnov.com