The Committee That Approved Everything and Delivered Nothing

The quarterly review has ninety minutes on the calendar and four initiatives on the agenda the program office has prepared, each arriving with its own deck, its own sponsor, and its own color-coded risk register. The COO takes the chair, flanked by a CFO who will leave halfway through for another meeting and a CIO who will stay but spend most of the session replying to messages under the table. Toward the back, in the observer ring, the program director sits with the master schedule on his lap, while around the room the VPs who own the four initiatives have arranged themselves in a quiet order of precedence that anyone who has been in the company longer than a year can read at a glance.

The first deck goes up: the ERP consolidation, approved eight months ago and since then absorbing budget in a way that was not quite expected. The sponsor walks through a status whose indicators are green, though two of them were amber last time and the description of what changed is slightly too smooth; a few executives ask the kind of questions that suggest interest without requiring commitment, and the committee notes the program is on track. Nothing technically requires approval at this stage, yet the sponsor leaves the section having been granted what amounts to tacit continuation, which is the outcome the pre-meeting bilaterals were organized to secure.

The second deck is new: a customer data platform initiative the commercial team has been preparing for months, with a business case that promises, in about that order, revenue lift, cross-sell improvement, and analytics maturity. The case assumes the same data engineering team the ERP program is already using, a fact nobody in the room mentions. The deck moves through the usual beats, and the committee approves it with a light revision to the timeline.

The third deck comes from the CIO, who has been waiting for this slot since the last session and who has prepared the argument that the platform debt accumulating in the current estate will eventually render every other initiative on the agenda impossible. He is right, though it will never be provable in a way that matters for this room. The initiative is an infrastructure modernization everyone agrees is ultimately unavoidable, at a cost that is high and for a business value that is diffuse, and the committee grants approval, conditional on a phased funding release that nobody present has the authority to actually administer.

The fourth deck is the initiative nobody wanted to discuss first: the operating-model redesign, which will eventually touch every function in the room. The HR director, whose name is not on the deck but whose organization will absorb most of the change, has been told she is here to listen. The sponsor presents and the room approves, without anyone raising sequencing, or capacity implications for the three initiatives already on the board, or the fact that the operating-model redesign and the ERP consolidation are, in a sense nobody is willing to name out loud, competing for the same redesign window.

By the end of the day the minutes will have been signed and filed, recording four approvals, after which the program office will begin building a consolidated roadmap that shows all four programs on parallel tracks, color-coded, with milestones. Six months from now, two of the four will have quietly stalled, a third will have absorbed budget and attention the others needed, and the fourth will have been descoped to a shadow of its original design without anyone formally acknowledging the descope has occurred. The committee will reconvene, express concern about execution, and ask the program office to propose a remediation, and no one will suggest that the problem began in the room where the approvals were granted.

The scene just described is the normal output of a governance forum carefully designed for a purpose its participants would never explicitly endorse. Executive laziness has very little to do with it.

The architecture beneath the agenda

Committees of this kind are almost always described, at the level of formal charter, as decision-making bodies, language that is precise and misleading at the same time. They do make decisions, in the sense that the meeting produces a record of approvals, but the decisions are not the work; they are the artifact of the work. The work is the question of what the organization will actually do with its finite capacity, and that question is settled by which competing claims get surfaced, which get deferred, which get quietly accommodated, and which get written into the record as if they were compatible when they are not.

A steering committee that does its job well is, before anything else, a negotiation forum, and the negotiation is constitutive, the only way a governance body composed of people with competing portfolios, competing budgets, and competing priorities can produce a decision the organization is actually able to execute. Competing claims that are not surfaced and reconciled in the room do not vanish; they remain active, and they will be reconciled somewhere else, by somebody else, under conditions almost always worse than the ones the committee could have offered.

The architecture of most steering committees is calibrated against this function rather than toward it, beginning with the agenda itself, which is almost always organized by initiative rather than by trade-off. Each sponsor presents her own program, on her own time slot, with her own deck, so that the implicit structure of the meeting is one of adjacent non-overlap: the programs are reviewed as if they existed in separate capacity universes, with a cross-portfolio view, if it appears at all, reduced to an opening slide or a closing slide and never becoming the structure of the session itself. The committee’s time is consumed by initiative-by-initiative review, which is precisely the format least likely to produce a conversation about which initiative should yield to which.

Layer on the convention of pre-meeting alignment, which every organization of any maturity has cultivated in one form or another. Sponsors are expected to resolve disagreements with peers before the session, to bring only clean material, to avoid the kind of in-session friction that wastes executive time, and the convention itself, taken on its own, is reasonable, because no one wants to run a steering committee that continuously re-hashes first-order disputes. The same convention, however, removes the only mechanism through which genuine trade-off resolution could occur, because by the time the meeting is called to order every initiative has already been negotiated into a state of presentability, and the negotiations that produced that state were, by definition, bilateral. They did not include the full set of parties whose commitments would have to move for the approval to hold, neither the CFO, who would eventually have to redirect funding already committed elsewhere, nor the program director, who would eventually have to sequence the redesign windows. Pre-alignment produces an artifact of consensus that, once accepted by the committee, will not survive first contact with implementation.

Add, finally, the incentive structure of the sponsors themselves, each of whom is accountable for the success of her own initiative and judged, in the end, by whether her program delivered, and none of whom is judged by whether the portfolio as a whole delivered or by whether the approvals granted in a particular session produced a coherent plan. The sponsor who comes into the room and says her program should be deferred in favor of another is making a heroic move no one in the organization is structured to reward, whereas the sponsor who comes into the room and says her program is ready and requires only the approval the committee is already inclined to give is making the rational move, and the forum rewards the second behavior and, over time, selects for it.

Put the three patterns together: the agenda format discourages cross-portfolio comparison, the pre-meeting alignment convention removes the in-session space where such comparison could happen, and the sponsor incentive structure makes in-session negotiation irrational for any individual participant. What remains is a forum that processes approvals, that records decisions, and that does none of the negotiation work those approvals would have required in order to hold.

Where the unresolved migrates

What the committee leaves unresolved does not disappear, and it does not quiet; it migrates into the parts of the organization least equipped to absorb it.

Consider the capacity director, who runs a team of about forty specialized engineers and who happens to employ the only people in the organization with the depth to lead both the ERP program’s integration layer and the customer data platform’s core build. The committee approved both programs in the same session, without any conversation about which program would have priority over her team for the two quarters in which both builds would overlap, and she discovers the conflict three weeks after the session, when two VPs separately approach her with staffing requests that, added together, exceed her team’s available hours by a factor of roughly 1.4.

She could, in principle, escalate the conflict to the committee that created it, because the governance architecture formally permits this; in practice, there are reasons not to. The committee meets quarterly, and by the time it reconvenes both programs will be past the point at which a clean re-sequencing is possible, while the escalation itself would require her to ask two executives, both senior to her, to revisit approvals each has already internally declared, in what would effectively become a priority dispute between two peer functions that is not hers to adjudicate. The path of least political cost is to negotiate a local solution, so she pulls her lead integration architect off a third project that was supposed to be her team’s contribution to the operating-model redesign, stretches her data engineering capacity across the two priority programs, and lets the redesign contribution slip into a future that will not be recorded anywhere except in the operating-model program’s internal dependency log, which the steering committee will never see.

Consider now the program lead for the customer data platform, who has been given a twelve-month timeline that, when she looks at it closely, assumes a level of business adoption support another approved program has already committed. The commercial team that will have to sponsor the platform’s adoption is, at the same time, the team asked to lead the commercial workstream of the operating-model redesign, and both programs have been told they have the commercial team’s full commitment. She does the arithmetic and concludes that full commitment to both, in overlapping quarters, is not possible. Her two options are to surface the conflict formally, which would require a governance escalation her sponsor would prefer she not initiate, because the sponsor’s own political position depends on the platform’s approved timeline holding, or to manage through it. She manages through it: she resets her internal expectations, revises her milestones to emphasize activity rather than adoption outcomes, and prepares the kind of status report in which green is sustained through a careful selection of what to measure, so that the steering committee, nine months later, will receive a program that is on schedule by its own internal definition and off schedule by any external one.

Consider, finally, the program director himself, the observer at the back of the room, who has watched the four approvals and has already, in the week after the session, begun to sketch the consolidated roadmap. As he sketches, he discovers what the committee did not: that the four approvals, in their current form, require about 1.7x the capacity the organization actually has for the overlapping quarters. When he takes this finding to the COO who chaired the session, the response is polite and unambiguous; the committee approved four programs, and it is the program office’s responsibility to sequence them in a way that is feasible. He has been asked, in effect, to re-do the negotiation the committee did not conduct, without the authority the committee had, without the sponsors in the room, without any governance mandate other than the implicit one that he find a way to make the roadmap work. He will produce a roadmap in which one of the four programs has been quietly pushed out by two quarters, the program whose sponsor has the least political capital in the organization, and it will be circulated, noted, and filed, without returning to the committee for re-approval because nobody wants it to.

Three scenes, and they are only samples, because in any large transformation portfolio the pattern repeats dozens of times: specialized teams become the arbiters of priority, program leads become the custodians of timelines that were never achievable, and program directors become the shadow committee that produces the feasibility the real committee did not. The work of governance migrates, not because it disappears but because it is no longer being done by people who were empowered to do it, and is instead being done by people who are carrying, on their own shoulders, the trade-offs the forum chose not to adjudicate.

There is a further consequence, easy to miss: the migration is invisible in the governance record. The minutes of the committee continue to show the four approvals, the program status reports continue to show, by any indicator anyone has agreed to track, that the programs are progressing, and the moment at which the operating-model program slipped two quarters is not recorded anywhere the committee would naturally see, just as the moment at which the platform’s milestones were rewritten to emphasize activity rather than adoption does not appear as a descope, and the moment at which the integration architect was pulled off the redesign is recorded, if at all, in an internal dependency log that has the distribution list of about four people. What the committee experiences, six months later, is a general and mysterious sense that execution is underperforming, while what has actually happened is that the negotiations the committee avoided have been conducted by other people, less well equipped, under worse conditions, and their cumulative result has produced the fragmentation the committee now wishes to remediate.

The people in the room cannot name what is happening

There are plenty of smart people in the room, so it is reasonable to ask why the pattern persists.

Part of the answer is that the people best positioned to name what is happening, the program managers and program directors, are trained in a discipline that does not treat naming it as their work. PMO certifications, transformation methodology courses, and the various flavors of project management credentialing are all organized around a particular model of the program manager’s role, in which she is the person who prepares the deck, presents the status, escalates the risks, and gains the approvals, serving as an interface between the delivery layer and the governance layer. The skills she is trained in are real and nontrivial, covering stakeholder communication, executive presentation, issue synthesis, scope management, and risk articulation, all of it codified with considerable thought into what the discipline calls good practice.

What the discipline does not train, almost at all, is negotiation within a governance forum, because the framework assumes, tacitly, that the governance forum is an authoritative decision-making body. If that assumption holds, the program manager’s job is to give the forum the information it needs, accept its decisions, and implement accordingly, whereas if the assumption does not hold, and the forum is instead a negotiation venue whose decisions bind only to the extent that the underlying trade-offs have actually been resolved, then the program manager would need an entirely different set of skills. She would need to know how to surface a competing claim without looking as though she were attacking another sponsor’s program, how to test in real time whether an approval reflects a genuine resource commitment or a social courtesy, and how to propose an explicit trade-off conversation in a forum designed to avoid them, all without violating the norms that governed her admission to the room in the first place. None of these skills appear in standard training, and their absence is not accidental; they are excluded by the same assumption that makes them necessary.

The asymmetry runs the other way as well, because the executives in the room are, in general, also not trained to adjudicate trade-offs in a committee format, having acquired their training implicitly, through years of executive meetings in which the surface norms rewarded diplomacy, efficiency, and the avoidance of in-session conflict. The executives who rose to the committee are, on average, the ones who learned those norms best and who, by selection, are least likely to disrupt the forum’s architecture, because to disrupt it would mean violating the set of behaviors that credentialed them in the first place.

The program director sits in the hardest position of all, because he can see the pattern, which his job requires him to see, while his job also does not include a mandate to challenge the forum. He was not appointed to tell the committee that its architecture is producing decisions it cannot deliver; he was appointed to implement what the committee decided, so that raising the structural question exceeds the scope of his role, while not raising it means absorbing the consequences of the forum’s refusal to resolve what it was built to resolve. Most program directors manage this by privately carrying the shadow work, publicly delivering the reports that make the programs look defensible, and waiting for the day they will be blamed for an outcome whose architecture they did not design, and many of them have long since stopped expecting the committee to see what is happening, an expectation that, once it dies, is not usually revived.

There is a more uncomfortable observation here, which is that the entire ecosystem around the steering committee, the PMO staff, the consultants who support it, the methodology training, the governance frameworks, has grown up around the assumption that what the committee is doing is decision-making, and every artifact the ecosystem produces reinforces the assumption: the templates encode it, the rhythms encode it, the vocabulary encodes it. Anyone who points out that the forum is not, in fact, making durable decisions has to swim against the current of every process that has been built to support the forum’s self-description, and most people do not bother, even though the pattern is not hidden and is openly visible to anyone willing to trace the gap between what is approved and what is actually delivered. It persists because the governance architecture that would require naming it has not been built, and no one whose role it would be to build it has been asked to.

A committee that admits what it actually does

The question of what would have to change is, on the surface, a question of meeting design, though that reading is misleading, because it is, more fundamentally, a question of whether the forum is prepared to admit that approval and commitment are different things and that its work is the closing of the gap between them. Once the question is posed that way, the implications unfold in a specific direction.

Preparation would have to be organized around interests rather than around presentations. In most organizations today, the week before a steering committee is absorbed by the production of decks, with the work of pre-alignment running in parallel, through bilateral conversations whose purpose is to ensure each presentation arrives in the room without unresolved disagreement. A forum organized around negotiation would invert this, because the week before would be spent not on finalizing decks but on mapping, explicitly and with some rigor, what each stakeholder is protecting, which resource claims, which priority positions, which organizational interests, so that the mapping itself would form the basis of the agenda, rather than the presentations. The meeting would then be built around the live engagement of those interests, rather than around their pre-resolution.

Approval itself would have to be treated as incomplete until the forum has recorded, alongside each decision, what the decision requires from every other initiative in the portfolio. The current convention treats approval as an endorsement that arrives at the end of a presentation, whereas a different convention would treat it as the beginning of the conversation rather than the end, so that the sponsor whose initiative has been approved would be asked, before the committee moves on, to identify what existing work she is willing to stop or to slow in order to create the capacity the approval assumes. She would not be allowed to nod silently, the sponsors of every adjacent initiative would be asked to confirm or to contest her assessment, and the record would reflect both the approval and, alongside it, the agreed cost.

Over time, the distinction between approval and commitment would develop into a visible grammar of the forum, because right now the two words are used as if they were interchangeable, whereas in a redesigned forum they would be visibly different and the difference would matter: approval as the committee’s willingness to record that the initiative should proceed, commitment as each sponsor’s willingness to carry the real cost the approval implied, with the forum remaining open on any given item until both had been made explicit. Most approvals, in the current architecture, would fail this test on first pass, which is precisely the point, because approvals that failed the test would either be renegotiated into a form that could pass it or would not be granted at all, and the committee would make fewer decisions, though the decisions it made would hold.

After the meeting, the governance architecture would have to monitor whether the conditions assumed at the moment of approval actually materialized, with the committee itself owning this monitoring as a matter of governance architecture rather than performance reporting. When a capacity assumption turned out to be incorrect, or when a priority conflict emerged that the approval had not anticipated, the committee would own the renegotiation and would not delegate it to the program director, who would otherwise carry the burden invisibly, which means that the committee that made the approval would also own its implications, up to and including the obligation to reconvene ahead of the quarterly cycle if the implications were material.

None of this is theoretical, because organizations that have built governance of this kind exist, though they are uncommon, and the design paths by which they arrived vary. A few have done it deliberately, often after a specific failure large enough to make the conventional architecture untenable, while a few have evolved into it over long periods, through the patient work of program directors and committee chairs who refused to accept the frictionless forum as the operating norm, and none of them describe what they have built as governance innovation; they describe it as the ordinary exercise of running the forum with the honesty its function requires.

What is common to all of them is that the forum has stopped pretending its job is throughput and has accepted, at the level of architecture and not only of language, that its job is the resolution of the trade-offs every multi-initiative portfolio generates, with the understanding that the resolution is often uncomfortable, often incomplete, and often slower than the organization would prefer. A governance body whose decisions hold is a governance body that has done the work of negotiation in the room, and the work cannot be outsourced to the week before or to the quarter after.

The distance the minutes do not show

A committee’s most consequential output is not the decisions it records. It is the negotiation it either conducts or defers, because when the negotiation happens in the room the decisions hold, and when it does not the decisions fragment, after which the organization spends the following months managing the consequences of fragmentation under the polite name of execution difficulty, while the forum that produced the fragmentation reconvenes to discuss remediation.

The gap this produces is quantifiable, in the particular sense that anyone who has audited a portfolio retrospectively has seen it: the distance between what the steering committee approved and what the organization actually absorbed. In one case of my own experience, that distance was close to forty percent of the original portfolio ambition, measured in milestones that had been formally approved but operationally never delivered, and the gap was not mysterious; it had been produced, item by item, by the forum’s unwillingness to engage the trade-offs each approval implied. The sponsors had felt they had approved programs, but the organization had, in effect, approved nothing of the sort; it had approved a set of intentions whose compatibility had never been tested.

Most practitioners will recognize the pattern when it is described to them, often with a small grimace, and the design question that remains is whether the forum is prepared to treat the negotiation as its function, rather than as a pre-meeting courtesy to be completed before the real work of approval begins. Most forums have defaulted to the second posture, because it is easier and the immediate costs are lower, even though the longer-run cost, the steady divergence between what is approved and what is delivered, is absorbed by the organization at a distance from the committee and in a form the committee does not see.

Changing the posture is neither fast nor quiet, and requires the committee chair to accept that sessions will run longer and produce fewer approvals, at least at first; it requires sponsors to accept that their initiatives will be contested in the room rather than waved through after a bilateral pre-alignment; and it requires program directors to accept a role that extends beyond implementation into the design of the forum itself. None of these changes are attractive to the people who would have to make them, but they are the only path by which the committee’s output will come to match its charter.

The minutes of the session described at the beginning will continue to show four approvals, and they will be accurate. What the minutes will not show is the negotiation that did not happen, the sequencing that was deferred, the capacity that was assumed to exist, and the programs that, six months later, will have quietly yielded to one another in ways nobody in the room was asked to endorse. That is the output of the forum as currently designed, and the question is no longer whether the pattern will recur; the question is whether the organization will keep pretending not to see it.