The Annual Planning Cycle Is Not the Problem — the Missing Ceiling Is
A portfolio board without a ceiling is a committee for saying yes more often.
The Spreadsheet That Arrives Every September
Every September, in most large organisations, a spreadsheet goes out. It has one row per proposed initiative and a column for each of the next twelve months, and it asks divisional heads to state what they intend to do next year, what it will cost, and what it will return. By late October it comes back, three times larger than anyone can afford. By December it has been cut, usually by a process nobody would defend in daylight, and in January the organisation begins executing a plan that was already out of date on the day it was approved.
We all know this. The textbooks know it too, and for the last few years they have proposed the obvious cure: replace the annual cycle with continuous portfolio management. Review the pipeline monthly. Re-rank on a rolling basis. Treat the portfolio like a fund manager treats a fund, rebalancing as conditions change rather than waiting for the year-end. The portfolio management literature, the PPM tool vendors, and a growing number of conference keynotes all point the same way.
My position is that this prescription is right about the disease and wrong about the cure — or rather, right about the cure and silent about the dosage. The annual cycle does not persist because organisations are slow to read the literature. It persists because it does something the continuous model, as usually implemented, quietly stops doing. Until we name that function and rebuild it, the move to continuous management does not produce a more responsive portfolio. It produces a portfolio that never says no.
What the Annual Cycle Actually Does
Strip away the ritual and the annual planning cycle performs exactly one irreplaceable act: it forces the organisation, once, to place every demand beside every other demand and compare them against a fixed sum of money and people.
That is worth dwelling on, because almost everything else the cycle does is done badly. The cost estimates are soft. The benefit figures are written to clear the hurdle rate rather than to describe the world. The sequencing is fiction, because nobody in October knows which regulatory letter will land in March. But the comparison is real. For a few weeks, a finance director and a handful of executives sit in a room with a total they cannot exceed, and they are obliged to rank things. The ranking is political, inconsistent and frequently wrong. It is still a ranking, and it still produces a list of initiatives that will not be funded.
That list is the product. Not the plan — the plan decays within a quarter. The list of things the organisation has agreed not to do is the only durable output of the whole exercise, and it is the thing that continuous models most often fail to reproduce.
The annual cycle is not valuable because it plans. It is valuable because it is the one moment in the year when demand is made to meet a ceiling and something is visibly refused.
Consider what happens in the intervening eleven months, under a conventional annual regime. New demands arrive — a regulatory change, a competitor’s move, a chief executive’s enthusiasm after a conference. Each is handled as an exception. An exception has no peer group; it is assessed on its own merits, against no alternative, and the question asked is “is this a good idea?” rather than “what will we stop in order to do this?” Almost every good idea passes that test. So the portfolio accretes through the year, and the September spreadsheet is partly an act of discovering what the organisation has already committed to without noticing.
The annual cycle, then, is a bad solution to a real problem: it provides the trade-off moment, but only once, and the organisation spends the rest of the year undermining it.
What the Continuous Model Leaves Out
The continuous model, as the textbooks and tool vendors describe it, proposes to make the trade-off moment permanent. A standing portfolio board meets monthly. The pipeline is scored against strategic criteria. Initiatives are admitted, re-prioritised or stopped on a rolling basis. In principle, the exception process disappears, because every new demand enters the same queue as everything else and is ranked against it.
In practice, the pattern I have observed most often is that the monthly portfolio board becomes an admissions committee. It meets; it reviews the new requests; it scores them; it approves most of them. What it does not do — because nothing forces it to — is remove anything. The ceiling that was so vivid in the annual process, the single number on the finance director’s page, has dissolved into a rolling forecast that can always be revised next month. Stopping a live initiative is a hard, visible, personally costly act, and a monthly meeting with no fixed envelope has every incentive to defer it.
The numbers tell the story. A composite that will be recognisable to anyone who has run a portfolio office in a large bank, insurer or utility: an organisation moves from an annual cycle to a monthly portfolio board in the first quarter of a year. The January portfolio holds 84 active initiatives against a change budget of roughly £60 million and a shared pool of about 220 business analysts, solution architects and test staff. By September the board has met eight times. It has approved 37 new initiatives and formally stopped 3. The active count is 118. The budget has been revised upward twice, but by less than the demand, so the real constraint — people — is now silent. Average analyst allocation across the active portfolio has reached 140 per cent, which is to say that a meaningful fraction of the 118 initiatives are being staffed by the same individuals on paper and by nobody in fact. The portfolio is, by the board’s own dashboard, almost entirely green.
Nothing in that sequence is a failure of the monthly cadence. The board met on time, scored consistently and minuted its decisions. What failed is that continuity was introduced without a ceiling, and a portfolio board without a ceiling is a committee for saying yes more often.
The Objection Worth Taking Seriously
The strongest reply to this is not that continuous management works fine — its advocates are usually candid that it is hard — but that the ceiling problem is a governance defect, not a cadence defect, and the fix is simply to give the monthly board a fixed capacity envelope. Put the 220 people and the £60 million on the first slide every month, require that admissions be matched by removals, and the continuous model does everything the annual cycle did, only twelve times as often. Why keep the September spreadsheet at all?
This is the right argument, and it concedes most of the point. Once you insist that the monthly board operates inside a fixed envelope, you have reintroduced the annual cycle’s one irreplaceable act — you have just relocated it. But the objection underestimates what it takes to set that envelope, and who has the authority to do it.
A capacity envelope is not a portfolio decision. It is a corporate decision about how much of the organisation’s money and attention will be spent on change rather than on running the business, and it is made — can only be made — by the executive committee and the board, in the context of the financial year, the dividend, the regulator’s capital expectations and the rest. A monthly portfolio board of divisional directors does not have standing to move that number, and if it is allowed to, it will, upward, every time demand exceeds it. The envelope has to be set by somebody senior to the people spending it, on a cadence tied to the organisation’s financial commitments. That cadence is annual, because the financial commitments are annual. This is not inertia. It is the structure of how a listed company or a public body is held to account.
So the objection is correct that the problem is governance, and correct that a fixed envelope is the fix. Where it goes wrong is in assuming the envelope can live inside the continuous process. It cannot. It has to be handed to the continuous process by a body that does not sit monthly.
Two Cadences, Not One
The conclusion I have come to is that the debate has been framed as a choice when it is a division of labour. The organisations that manage their portfolios well — and there are fewer than the conference programmes suggest — run two cadences with a sharp boundary between them.
The annual cadence decides the envelope and almost nothing else. Once a year, the executive sets the change budget, confirms the capacity of the constrained shared resources, and agrees a small number of mandated commitments — the regulatory deadlines, the board-level strategic bets — that are fixed for the year. It does not approve a list of 84 initiatives. It approves a sum of money, a count of scarce people, and a handful of non-negotiables. The September spreadsheet shrinks from a planning instrument to a sizing instrument: it still collects demand, but its purpose is to calibrate the envelope, not to pre-allocate it.
The monthly cadence spends the envelope. The portfolio board admits, re-sequences and stops initiatives, but against a number it cannot change. Its first agenda item, every month, is the envelope and the current commitment against it; its second is the list of what must stop or slow before anything new starts. The rule that gives it teeth is simple to state and uncomfortable to apply: no admission without a named removal or a named deferral, signed by the sponsor who loses.
| Annual cadence | Monthly cadence | |||
|---|---|---|---|---|
| Decides | Change envelope; constrained resource capacity; mandated commitments | Which initiatives are live, sequenced, paused or stopped | ||
| Owned by | Executive committee, with the board | Portfolio board of divisional and functional directors | ||
| Fixed for | The financial year | One month | ||
| Output | A ceiling and a short list of non-negotiables | A live portfolio that fits inside the ceiling | ||
| Failure mode if absent | Envelope drifts upward; constraint becomes invisible | Exceptions accrete; portfolio overloads between annual resets |
Return to the composite. The same organisation, a year on, has the executive set the envelope in the autumn — £58 million, 220 shared staff, six mandated regulatory and strategic commitments — and hands it to a monthly board that can spend but not stretch it. The board admits 29 initiatives across the year. It stops or formally defers 31. The active count peaks at 91 and ends the year at 79. Analyst allocation never exceeds 105 per cent on paper, which in practice means people are working on what the dashboard says they are working on. The portfolio is no longer entirely green, because for the first time the status reports are describing something real. And the September spreadsheet, which nobody has abolished, is a third of its former length, because divisions have learned that demand submitted in the autumn is sized, not granted, and the real contest happens monthly against a number that does not move.
What the Textbooks Leave Out
What the literature omits is not a technique. It is a piece of organisational realism: the trade-off only happens when a ceiling is present, and a ceiling only holds when it is set by people who do not have to live inside it. The annual cycle supplied that ceiling, badly and once. The continuous model, as usually sold, removes the ceiling in the act of removing the cycle, and then wonders why the portfolio swells.
We should stop asking whether to plan annually or continuously. We should ask who sets the number, how often, and whether the people spending it are allowed to touch it. Get that right and the cadence of the portfolio board is a detail. Get it wrong and no cadence will save you — the monthly board will simply discover, twelve times a year instead of once, that it has promised more than the organisation can deliver.