One person, every gate
Trace the transformation programme back through its approvals and a single name recurs. The operations director proposed it, shaped the business case, selected the consultancy that validated it, chairs the steering committee that governs it and will present it for final sign-off next month. Each gate the proposal has passed was real, and each was controlled by its author. The board believes the case has been tested five times. It has been asserted five times, with increasing polish.
How advocacy absorbs stewardship
None of this requires bad faith. Organisations resource conviction, so executives learn to advocate hard, and the sponsor role then hands the advocate the machinery of scrutiny. From there the incentives do quiet, cumulative work. Challenge from below starts to read as disloyalty to a person rather than scrutiny of a proposal, so it softens. Information flows upward through the sponsor, who selects, frames and sequences it, honestly but not neutrally. And as months pass, the programme fuses with the sponsor's standing: by the time doubts would be most valuable, expressing them has become an act with a career price, for the sponsor most of all.
The predictable results
- Core assumptions travel from first draft to final approval untested, because every reviewer assumed a previous gate had done it.
- Bad news attenuates on its way up, each layer softening it slightly, so the board hears a programme in permanent good health until it is in visible crisis.
- Kill criteria are never defined, because asking the sponsor to define them feels like doubting the sponsor.
- The board ends up approving a person's credibility rather than a case, which is why the same board is later surprised.
- The post-failure review discovers that several people had material doubts throughout, and that none of them had a channel that did not run through the sponsor.
Separating the roles without killing the energy
The fix is structural, not personal, and it works best when it is installed as standard governance rather than aimed at an individual. Name a decision owner who is not the proposer, so the person accountable for the verdict is different from the person selling it. Commission the case against with real resource and the same data access as the case for; an afternoon of token red-teaming does not count. Require kill criteria in the original approval paper, written by the sponsor: the conditions under which they themselves would stop the programme. Rotate authorship of progress reporting so the narrative is not singly owned. And make the sponsor's incentive symmetrical, so a well-judged stop is rewarded as visibly as a launch. Sponsors defend programmes to the last slide partly because stopping one has never once been career-positive in their organisation.
Challenge that does not pass through the sponsor
Internal challengers all have one limitation in common: a relationship with the sponsor that outlasts the decision. Selected senior operators from the Global Board have none. They can test the case on its merits, from experience of running comparable programmes, without managing anyone's standing, and their input arrives as a confidential report rather than as a colleague's dissent. For the board, that is an independent reading of the one document nobody inside the building can review neutrally. For a good sponsor, it is oddly liberating: external validation that survives scrutiny is worth more than internal applause that was never going to withhold it.
