Why Portfolios Drift
Money already spent gets no vote.
The portfolio nobody chose
Walk into almost any organisation and ask to see its portfolio. You will be handed a list. It will be long. It will be sorted by something — start date, sponsor, business unit, the colour of the status light — and somewhere on it will be every initiative that anyone, at any point in the last several years, managed to get funded and has not yet been forced to admit is finished.
That list is not a portfolio. It is a sediment.
Nobody sat down and chose it. No single decision produced it. It is the accumulated residue of hundreds of small, locally reasonable approvals, each made when money was a little looser, each defended ever since by the simple fact of already existing. The organisation did not decide to run these forty things rather than some better forty. It decided to run the first one, and then the next, and then the next, and the rest is just what survived.
This is the central problem of portfolio life, and it is worth being precise about, because it does not look like a problem. It looks like activity. The teams are busy. The reports are produced. The governance forum meets on the second Tuesday of the month and approves the minutes of the previous second Tuesday. Everything is moving. The trouble is that nobody can tell you, with a straight face, that this particular set of moving things is the best set the organisation could be funding with the money it has. They can only tell you that it is the set that happens to exist.
“A portfolio that nobody chose is not neutral. It is a series of decisions made by inertia and defended by sunk cost.”
How the residue forms
Drift is not a failure of intelligence. It is the natural result of two habits that, on their own, look like virtues.
The first is additive intake. New ideas arrive constantly — from the strategy refresh, from the regulator, from the loudest customer, from the executive who came back from a conference with a thought. Each is assessed on its own merits, in its own meeting, against the question is this worth doing? And very often the honest answer is yes, it is. So it is added. The portfolio grows by accretion, one defensible yes at a time. What is almost never asked, at the moment of the yes, is the only question that actually matters at portfolio level: is this worth doing more than the thing it will take resources from? That question has no natural owner in an additive process, so it does not get asked, and the list gets longer.
The second habit is the absence of an exit. Starting work is an event. It has a sponsor, a business case, a launch, a kickoff with pastries. Stopping work is a non-event — or worse, an admission. There is no ceremony for it, no template, no person whose job it is to propose it. Initiatives therefore do not leave the portfolio when they stop earning their place; they leave it only when they are completely finished, or when someone with enough authority and enough courage decides to take the reputational hit of killing something. Since finishing is rare and courage is rationed, almost nothing ever leaves. The portfolio is a venue you can check into but, in practice, cannot check out of.
Put additive intake together with no exit and you get drift as a matter of arithmetic. Things come in faster than they go out. The pool of work expands while the pool of money and people does not. And because the new arrivals are assessed against an abstract bar rather than against the incumbents, the incumbents are never made to defend their place. They simply have it, by right of having had it yesterday.
The final ingredient is the most human: momentum gets mistaken for commitment. A piece of work that has been running for two years feels like something the organisation has decided to do. In truth, it may be something the organisation decided to start two years ago and has never once revisited. The decision was made under conditions — a strategy, a market, a set of assumptions — that have since changed beyond recognition. But the work carries on, propelled by its own history, and every month that it continues makes it feel more decided, not less.
What drift costs
It would be one thing if a drifted portfolio were merely untidy. It is not untidy. It is expensive, in ways that compound.
The first cost is dilution. A limited pool of money and skilled people, spread across too many things, is spread thin everywhere. The good initiatives are slowed because the people they need are also assigned to three other things; the weak initiatives are kept alive on a maintenance drip that produces nothing but keeps a claim on the budget. Nothing is properly resourced because everything is somewhat resourced. The organisation feels this as a permanent slowness — vast effort, little movement — and usually answers it by adding governance, which slows things further.
The second cost is the zombie. Every drifted portfolio contains initiatives that are neither alive nor dead: they have lost their sponsor, or their rationale, or their market, but no one has noticed or no one has acted. They consume budget, occupy people, generate status reports, and produce nothing anyone would miss. They are kept alive not by value but by the absence of a process for ending them. A portfolio can carry a surprising number of these before anyone notices, because each one is individually small and collectively invisible.
The third cost is strategic disconnection. Ask, of any large drifted portfolio, what proportion of its spend can be traced cleanly to a current strategic priority, and the answer — if anyone is honest enough to compute it — is sobering. A great deal of the money is funding things that made sense under a previous strategy, or no strategy, or a strategy that was quietly abandoned. The link between what we say matters and what we actually fund has been severed, slowly, by accretion. The organisation has a strategy on a slide and a portfolio on a spreadsheet, and the two have not been on speaking terms for years.
The deepest cost of drift is invisible: it is the good work that was never funded because the money was already committed to work that had stopped deserving it.
This opportunity cost is the silent killer. While the organisation congratulates itself on being busy, genuinely higher-value investments — the ones that could move the strategy forward — never even make it onto the list because capacity is already locked up.
That cost of omission never appears on any report. Every euro tied up in an incumbent that no longer earns it is a euro unavailable for the new bet that might. The drifted portfolio is not merely carrying dead weight; it is crowding out its own future — declining next year’s best work by default, not by judgement.
The tells
Drift announces itself, if you know the sounds. A few are so common they are almost diagnostic.
There is the initiative that only needs another ten percent to finish — and has needed exactly another ten percent for three consecutive years. The figure never changes, because it was never a measurement; it was the smallest number that sounded survivable.
There is the status report that has been green for eighteen months and turns red on the single day it is finally cancelled — proving, in one stroke, that the colour was never describing the work. It was describing the reporter’s confidence that no one would look closely.
There is the strategic initiative that nobody can link to the strategy. Everyone agrees it is strategic. The word appears in its title. When asked which specific strategic objective it advances, the room becomes thoughtful, and then someone changes the subject.
There is the sponsor whose bonus depends on the initiative continuing, who is therefore the last person in the building able to assess, calmly, whether it should. And there is its cousin: the initiative too senior to question, sponsored by someone too important to be told no, which has become less a piece of work than a monument.
These are funny, in the bleak way that true things about working life are funny. But the humour is not the point. Each tell is a symptom of the same underlying disease: a portfolio in which the act of choosing has been replaced by the inertia of having chosen. The reports stay green not because lying is fun but because there is no safe mechanism for a piece of work to admit it has stopped being worth it. The ten percent never lands because no one is asking the only question that would expose it — would we start this today, knowing what we now know?
Doing the right things
There is a comfortable belief, widespread and wrong, that the way to fix a struggling portfolio is to deliver better. Tighten the project management. Improve the estimates. Install a better tool. Run the standups more crisply. All of this is doing things right, and all of it is worth doing — inside an initiative that deserves to exist.
But no amount of execution excellence rescues a portfolio that is funding the wrong things. You can deliver a doomed initiative beautifully, on time and under budget, and at the end you will have a beautifully delivered thing nobody needed. The crispest standups in the world cannot make a zombie worth its budget. Precision in execution is wasted, and sometimes worse than wasted, when it is applied to work that should not be running at all — because it makes that work look healthy, which makes it harder to stop.
“You can execute the wrong portfolio flawlessly. The flawlessness is the trap.”
At the level of a single initiative, the question is are we doing this thing right? At the level of the portfolio, the question is entirely different: are these the right things to be doing at all? The two questions are not in tension, but they are not the same, and they are not solved by the same means. The portfolio’s lever is not execution quality. It is choosing — and, far harder, re-choosing. The single most valuable act available to the person accountable for a portfolio is not making a running initiative go faster. It is recognising that a running initiative no longer earns its place, and moving its money to one that does.
That act is rare, and the reasons it is rare are the subject of much of this book. It is uncomfortable. It is political. It looks, to the untrained eye, like failure rather than discipline. But it is the entire game. A portfolio managed well is not one in which everything is delivered superbly. It is one in which, at every moment, the money is pointed at the best available set of things — and is moved, without sentiment, when the best available set changes.
The promise
If drift is the natural state, then the cure cannot be a one-off cleanup. Every organisation has, at some point, run the great rationalisation — the painful quarter where a task force reviews everything, kills a third of it, and declares the portfolio fixed. And every organisation has watched it drift straight back, because the forces that produced the sediment in the first place — additive intake, no exit, momentum mistaken for commitment — were never changed. They were merely interrupted. The cleanup treats the symptom. The drift resumes the following Tuesday.
The alternative is not a bigger cleanup. It is to stop treating the portfolio as a thing that gets decided and start treating it as a thing that gets continuously re-decided. A portfolio is not a plan you make and then execute; it is a standing question you answer, and keep answering: given everything we now know, and the limited money we have, what is the best set of things to be funding — and what should stop? Asked once, that question produces a clean-up. Asked on a rhythm, for ever, it produces a portfolio that cannot drift, because drift is exactly what regular re-choosing prevents.
That is the promise of this book. Not a better way to deliver, and not a one-time purge, but a way of running a portfolio as a living set of choices — surfaced, valued, funded, coordinated, and re-judged on a steady beat, with the money moving to follow the value wherever it goes. And, crucially, a way of doing this that is genuinely one method whatever the scale: the same discipline serves a single owner steering a handful of products and a global enterprise governing billions through tiered boards. What changes between them is the speed, the number of people in the room, and the weight of the paperwork. What does not change is the question.
The next chapter sets out that method — its single model, its unit of management, its two jobs, and the way one method bends to fit any scale. The rest of the book takes it stage by stage. But none of the machinery matters until the first idea is accepted: a portfolio you did not choose is one you are paying for, every day, in the work you cannot afford to start.