Strategy & Transformation Office

TL;DR. A transformation office is either the best seat in a large company or a slide-deck graveyard. It depends entirely on the setup. One row on the right is survivable; three is terminal.

The Strategy & Transformation Office exists because cross-cutting change has no natural owner. When an initiative touches three business units, two functions, and the IT estate, nobody with P&L will pay the coordination cost out of their budget for someone else's KPI. Put the CEO in charge directly and they become the bottleneck — every trade-off queues in their calendar and dies there. So the organisation creates a function: a named executive, a small team, a mandate to get the thing done across silos. The logic is clean. The track record is not.

McKinsey has been publishing the 70% transformation failure rate for close to two decades. The number hasn't budged, despite the proliferation of the function created to fix it. Separately, roughly 75% of PMOs fail inside three years, while McKinsey's own guidance puts the realistic horizon for a transformation at three to five years. The gap between how long the job takes and how long the structure that runs it lasts is the first clue that something structural is wrong.

The asymmetric ledger

The central problem with the transformation office is what happens to its P&L of successes and failures over time.

When a transformation project works, it works because a BU head, a CIO, or a COO drove it home. The office helped — unblocked a dependency, arbitrated a budget fight, forced a decision — but the thing that actually shipped has a line owner's name on it. At the next company offsite, that line owner tells the story: "we migrated the Northern region to the new platform." The office is not in that sentence, and shouldn't be. The owner earned it.

When a transformation project fails — and most do — the failure has no line owner. The BU head is on the next job. The CIO has moved to another priority. The office, by design, is the only entity left holding the file. In the post-mortem, the failure is attributed to "the transformation" — which is the office's name.

This is not a temporary bug. It is a structural feature of the setup. The office's ledger accumulates only losses, not because it loses more than it wins, but because its wins don't stay on its books. Give it five years and the ledger reads as a handful of unattributable contributions to line successes plus a stack of failed programmes that were never really the office's to execute. The internal reputation bends toward "whatever the transformation office touches doesn't land", and the people running it start updating their LinkedIn.

The execution gap

The office cannot execute anything directly. That is not a flaw of a particular team — it is the shape of the role.

It isn't the domain expert on the new core platform. It can't rewrite the supply chain. It doesn't know which regions to consolidate. Whatever it "owns" has to be delivered by the people whose territory it is crossing — the CIO for the IT stack, the COO for operations, the BU heads for their patch.

Those people read the transformation office as a tax on their autonomy, paid for someone else's KPI. They have ten ways to slow-walk it without ever saying no: under-staff the working group, escalate every trade-off, demand more discovery, redefine scope, loop in legal, reopen the business case. None of these are insubordinate. All of them are reasonable. In aggregate they move the roadmap at the speed of the least enthusiastic dependency.

The office's response to this is rational and disastrous: it starts measuring what it can measure. Meetings held. Workstreams open. RAG-status reports. Risk logs. Decision registers. The ritual of programme governance replaces the substance of delivery. McKinsey and BCG will tell you a good transformation office is a decision-forcing function with authority over scarce resources and the mandate to hold line leaders accountable. In practice, most offices never get that authority exercised visibly, and without visible exercise, everyone in the building reads the mandate as optional.

Theatre as a stable equilibrium

Once an office has drifted into this state, it is extraordinarily stable. It can't be shut down, because closing it is a narrative defeat for the CEO who set it up. It can't be escalated, because the BU heads have been practising slow-walk for eighteen months and are now very good at it. It can't deliver, because delivery would require authority it never exercised and now can't. So it produces slides. Nothing ships. Nobody is accountable. Everyone is busy.

This is the McKinsey 70% number, viewed from the inside. The function that was supposed to fix the failure rate became a new way to record it.

The steelman

The defence of the office has three strong points, and the critique is not honest without engaging them.

First, credit migration is the system working. The office is the seed crystal. Without it, most of those successful transformations would never have started — nobody with a P&L pays the coordination cost for cross-cutting change out of their own budget. The fact that the BU head tells the story at the offsite is not theft. It is the office doing what it should: catalysing a thing, then letting the line own the delivery. Fair.

Second, failure absorption is the job. Somebody has to be structurally dispensable so permanent leadership can keep trying. General counsels eat blame for killed deals. Interim CEOs eat blame for unpopular restructures. Transformation offices eat blame for the change that didn't land. That isn't inefficiency — it is shock absorption. Also fair.

Third, the alternatives are worse. Line ownership with a CEO sponsor makes the CEO the bottleneck. "Run it as a campaign" gets negotiated to death the moment the first BU head notices the timeline is optional. In distressed, regulated, or post-M&A contexts, a single horizontally-authorised executive isn't a preference — it is a requirement. Private equity keeps deploying interim Chief Transformation Officers into portfolio companies, and PE has the sharpest P&L incentives in corporate life. If this were purely theatre, the buyers with the shortest patience would have stopped paying.

These are real. They do not dissolve the structural problem. What they describe is the narrow configuration under which the office earns its keep — and outside that configuration, it decays into the equilibrium above. The existence of the office is not the question. The setup is.

When the job is worth taking

The table at the top is a checklist. Three rows do most of the work.

A real CEO mandate is one that has already been exercised. Ask in interview: tell me about the last time the office made a trade-off call that a BU head disagreed with — what happened? If the answer stumbles, the mandate exists on paper, and on paper is where it will stay.

A distressed context hands the office its teeth. In a turnaround, a consent-order remediation, a PE acceleration, or a post-M&A integration, the board watches weekly and the alternative is worse for everyone. In a healthy, profitable, comfortable enterprise with "an ambition to change", the office is decorative by default and will spend every week fighting for relevance it won't get.

Leavers' next roles tell you what the office actually is. Last three people who left — did they take COO, BU head, or divisional MD roles? The office is a launch pad. Did they leave the company? It is a graveyard, and the people still in it know.

When the left column holds, this is the best single job in a large company. You sit at the crossroads of strategy, finance, operations, and technology. You broker trade-offs at exec level. You build a network across every BU in eighteen months that a line role would take a decade to match. The same structural pressure that makes the office hard to succeed from the company's point of view is exactly what makes it valuable from yours — because you are forced to decide, in public, under a CEO who is visibly backing you, with a named landing zone waiting on the other side.

Where I land

The transformation office isn't a scam, and the structural critique doesn't imply it shouldn't exist. It should — sometimes, in narrow contexts, with a short clock and a named outcome and a CEO who has already shown they will back it against the line.

Most offices are not that. Most are permanent functions with plural mandates and committee reporting lines, set up inside comfortable enterprises where nothing is actually burning. They decay into theatre because the structure rewards it, and the people who run them become unemployable by the end — not by malice, but by design.

If you are joining one, read the table. One row on the right is survivable. Three is a scar on your CV in eighteen months.

If you are running a company and you are about to set one up, ask instead whether what you need is a campaign with a sunset clause, a named executive, and a promotion waiting on the other side. If the answer is yes, build that and call it whatever you like. If the answer is no — if what you actually want is the appearance of transformation without the cost of the trade-offs — spare yourself the expense. The office won't solve that problem. It will just give it a name.