Why this matters
Most organisations describe their priority problem as a volume problem. There is too much to do and not enough people to do it. The framing is comfortable because it points toward remedies nobody has to argue about — more headcount, more hours, better tools, tighter project reporting. It is also, in most cases, wrong about where the difficulty actually sits.
Priorities accumulate not because leaders lack discipline but because the cost structure of adding one is deeply asymmetric. At the moment of decision, adding a priority is close to free. It requires a sponsor, an announcement and a slide. The cost lands somewhere else entirely — later, further down, and distributed across people who were not in the room when the commitment was made. Nothing in the decision itself registers that cost, so nothing in the decision resists it.
The consequence is a gap that opens without anyone noticing. The organisation continues to believe it is pursuing eleven things. In practice it is pursuing four of them properly, three intermittently, and holding the remaining four in a state that is neither active nor closed. Leadership sees a reporting layer that shows movement on all eleven, because a status report asks whether work is progressing, not whether it is receiving the attention it was promised. The real allocation decision — which of the eleven gets the good hours of the good people this week — has been made repeatedly, informally, by individuals acting reasonably within their own line of sight.
This matters for three reasons that compound. The first is arithmetic: under-resourced work does not fail cleanly, it takes longer, and long-running work consumes coordination effort out of proportion to its progress. The second is credibility: each announcement that joins a visible queue of unfinished ones reduces the weight of the next, and that erosion is not recoverable by communicating harder. The third is the least visible and probably the most damaging. An organisation that never formally concludes anything cannot learn anything, because learning requires a moment at which someone asks what a piece of work produced — and that moment only exists if the work has an ending.
How the list grows, and why nothing leaves it
The mechanics of accumulation are less about ambition than about asymmetry. Adding a priority is an act with a sponsor, a rationale and visible momentum. Removing one requires someone to state publicly that something previously committed to is no longer worth doing. That statement carries a cost the addition never did: it implies the earlier decision was wrong, or that conditions were misjudged, or that the sponsor has lost an argument. Faced with that cost, organisations rarely stop work. They let it go quiet.
Consider an organisation where a process improvement project launched with visible sponsorship. Eighteen months on, it has no active workstream, no owner, and no record of having been closed. The obvious reading is that it failed and people moved on. What complicates that reading is that nobody involved describes it as having failed. Each person believes it was paused pending something else, and each names a different something. What the organisation lacks is not commitment or competence. It lacks any mechanism by which an initiative formally ends. Work does not get cancelled; it becomes dormant, and dormancy is invisible.
Dormancy is worth sitting with, because dormant work is not free. It holds budget lines nobody has released, calendar slots that recur out of habit, and a share of the attention of whoever still half-believes they own it. It also occupies a place in the organisation’s sense of what it is doing, which is why leadership teams are so often surprised by how little capacity is genuinely uncommitted.
A second mechanism sits alongside the first. Priorities arrive from multiple legitimate sources — a board, a regulator, a large customer, a functional plan, a competitor’s move. Each is individually defensible. What is missing is anyone whose job is the aggregate: individual decisions are owned, the total is not. In practice, organisations often find that the question “what does this displace?” has no natural owner, and so it is not asked.
There is also a developmental dimension. Larry Greiner’s work in the Harvard Business Review on how organisations change as they grow made the argument that the arrangements which resolve one phase of an organisation’s development tend to become the constraint in the next — that a major solution in one period reappears as a major problem in a later one. Priority management is a clear case. In a smaller organisation, the list is self-correcting: everyone can see everything, conflicts surface in a single room, and the constraint is obvious because it is visible. That mechanism does not break as the organisation grows. It simply stops covering the territory, and because it was never a formal mechanism, nobody notices when it stops working.
What actually runs out
Capacity is almost always discussed in hours, and hours are usually the least binding of the constraints in play. Three others bind sooner.
The first is senior attention. A cross-cutting priority does not merely need people to work on it; it needs someone senior enough to unblock it when it collides with something else. That unblocking capacity is genuinely finite and does not scale with headcount. An organisation can staff twelve initiatives and still only be able to arbitrate for four of them.
The second is coordination. Research from Donald Sull, Rebecca Homkes and Charles Sull, published in the Harvard Business Review in 2015 and based on a multi-year study of more than 250 companies and several thousand managers, found that only 9 per cent of managers said they could rely on colleagues in other functions and units all of the time, with around half saying they could rely on them most of the time. That figure is worth reading carefully, because it describes a supply, not an attitude. Every cross-cutting priority draws on the same limited stock of cross-unit reliability. Adding the ninth initiative does not divide that stock into nine equal parts; it draws on it in a way that makes each existing commitment slightly less dependable. The same research located the central execution difficulty in cross-unit coordination rather than in vertical alignment, and found alignment processes in most firms to be broadly sound — which reverses the direction most organisations look when priorities start failing.
Jay Galbraith’s work on organisation design, published in Interfaces, supplies part of the reason. His argument was that task uncertainty drives organisational form, because uncertainty determines how much information must be moved and processed. Each additional priority raises the volume of information that has to cross boundaries — and boundaries are where information moves slowest and degrades most.
The third constraint is the least discussed: the willingness to keep absorbing change. Work by Nelson Repenning and John Sterman of MIT Sloan, published in the California Management Review, examined why process improvement programmes so often stall, and found that managers under pressure default to working harder rather than working smarter, in part because those who prevent problems receive no visible credit for problems that never occurred. The implication for a crowded priority list is uncomfortable. When commitments exceed capacity, what gets dropped is not the least important work. It is the work whose absence is hardest to see — the preventive, the structural, the slow. That work is quiet when it is done and quiet when it is not, so it loses every contest for attention it enters.
Underneath all three sits a question of what people are actually rewarded for. Steven Kerr’s paper in the Academy of Management Journal on the divergence between what organisations reward and what they say they hope for observed that people seek information about which activities are rewarded and then do those things, sometimes to the exclusion of activities that are not. Applied here, this suggests the real priority order is not the one in the plan. It is the one legible in what gets noticed, chased and recognised — and where those two orders differ, the second one wins without ever being debated.
What this looks like from inside is rarely dramatic. Decisions get revisited rather than made. Status reporting shows progress on everything and completion of little. Senior calendars fragment into short slots. Requests described as quick absorb the time that had been reserved for declared priorities. None of this reads as a crisis, which is precisely why it persists.
Where the pattern differs, and what it is usually mistaken for
The pattern is general, but its presentation is not.
Size and maturity change how conflict surfaces. In a founder-led organisation, competing priorities collide visibly and are resolved in the room, often within minutes and sometimes badly. At a few hundred people the collisions still happen, but they no longer happen anywhere in particular, and resolution occurs through the absence of resolution — one thing simply receives less.
Structure changes what the conflict looks like. In matrixed or multi-site organisations, priority conflict tends to present as local variation rather than open disagreement, because a site or a function can absorb an unwanted priority by deprioritising it locally without ever contesting it centrally. The centre sees compliance, and the variation surfaces later, usually by accident.
Sector changes what capacity means. In professional and knowledge-based organisations, capacity is a small number of individuals with expertise that cannot be substituted, so the constraint is a person rather than a budget. In production-based organisations capacity appears measurable, which creates a different failure: the constraint is read as a scheduling problem and handled as one, long after it has become something else.
Three misconceptions recur.
The first is that this is a discipline problem — that leaders simply need to say no more often. The evidence points elsewhere. Work by Paul Rogers and Marcia Blenko, published in the Harvard Business Review and drawn from a survey of executives at more than 350 global companies, found decision-making stalling at organisational boundaries: between centre and business unit, between one function and another, between global and local. Priority conflict concentrates at exactly those seams. That is a structural location, not a character trait, and a leader with excellent judgement standing on one side of such a seam cannot resolve it alone.
The second is that prioritisation is the missing act. Most organisations prioritise, sometimes elaborately. What they lack is any means of making a prioritisation binding on the following week, when new information arrives with its own sponsor and its own urgency. A decision that binds nothing is not a decision; it is a preference, and preferences lose to arriving pressure.
The third is that the failure rate is known. Figures claiming that a fixed proportion of strategies or change programmes fail circulate widely. Mark Hughes, at the University of Brighton, examined five separate published instances of the 70 per cent change-failure claim in the Journal of Change Management and found no valid or reliable empirical evidence supporting any of them. The honest position is that there is no trustworthy base rate for this at all — and that anyone quoting one should be asked, first, where it came from.
Two tensions have no clean resolution, and it is worth naming them as tensions rather than pretending otherwise. A tightly held priority list is a bet that the environment will not change in ways that make the excluded options valuable — sometimes a good bet, never a free one. And when the front line quietly reprioritises, the same behaviour reads as a control failure from above and as sensible adaptation from below. Both readings can be accurate at once, which is why the argument recurs and rarely concludes.
What the research shows
Read together, the available research agrees on more than it disagrees on, but the agreement is narrower than it first appears.
The clearest point of convergence is directional. Sull and colleagues located the execution difficulty in horizontal coordination rather than vertical alignment. Galbraith’s design work reaches the same territory from a different starting point, treating cross-boundary information load as the thing that determines what organisational form can sustain. Michael Beer and Russell Eisenstat, writing in what is now the MIT Sloan Management Review, identified a set of mutually reinforcing barriers to strategy implementation — among them unclear strategy and conflicting priorities, and poor vertical communication — and made the further observation that these barriers are frequently invisible to senior leadership, because the people positioned to see them are not the people who say so. Three independent lines of work, then, point away from the place organisations habitually look first.
The divergence is about mechanism, and it is real. Kerr, and Repenning and Sterman, locate the driver in incentives and in how attention is allocated under pressure. Galbraith locates it in structure and information-processing demand. Greiner locates it in developmental stage — arrangements that were adequate becoming inadequate without anything having broken. These are not variants of one explanation. They imply different things about whether a priority problem is designed in, incentivised in, or grown into, and an organisation that assumes the wrong one will look in the wrong place for a long time.
What remains genuinely unresolved is more striking. The literature is far stronger on why priority conflict occurs than on how much of it an organisation can carry. There is no credible figure for the number of concurrent strategic priorities any organisation can sustain, and given how heavily that number would depend on interdependence, boundary structure and the availability of senior arbitration, it is not obvious that a general figure could exist. Hughes’s work is a useful corrective here: the appetite for a number in this territory has historically outrun the evidence for one. In practice, organisations often find that the more useful question is not how many priorities they hold but how many of them require the same scarce people to cooperate across the same boundary at the same time.
The same pattern appears across many kinds of organisation; see Other Sectors for how it is read elsewhere.
Questions worth considering
- What would have to be true for a priority in this organisation to be formally ended, and has that ever happened?
- Who, if anyone, holds a view of the total commitment rather than of an individual initiative?
- If the priority order were inferred purely from what gets chased, escalated and recognised, would it match the stated one?
- Which current commitments depend on the same small group of people cooperating across the same boundary?
- What work has been quietly set down without a decision, and who knows about it?
- When something is dropped locally, does that information travel anywhere?
- What would the organisation lose if it were wrong about how much uncommitted capacity it has?
This guide sits within Strategy Execution. A related guide: Why the Same Standard Produces Different Results by Site.
Your team is facing similar challenges?
Subscribe to The Riddle Ledger
Perspectives are sent as they are published, with no promotional email, and you can unsubscribe at any time.