Portfolio Triage Under Fire: Why Fair-Weather Prioritisation Failed the Crisis
In a crisis, the portfolio is not a ranking problem; it is a sequence of irreversible choices.
Executive Summary
The financial crisis has exposed a weakness that years of portfolio-management improvement did not address. Most organisations built prioritisation systems for conditions of relative stability: annual planning cycles, broadly predictable funding and enough time to compare proposals through weighted criteria. When budgets were reduced by 30 or 40 per cent within weeks, those systems produced neither speed nor a defensible stopping order.
The problem was not simply poor data. It was a category error. Weighted scoring is designed to distinguish attractive proposals before commitment. Crisis triage must decide what to protect, pause, redesign or stop after commitments, dependencies and political expectations already exist. These are different decisions.
A representative portfolio of 86 initiatives, with £240 million of remaining planned expenditure, had ten working days to remove £95 million. Its scoring model ranked 31 initiatives as “strategic”, placed 64 within a narrow 13-point range and contained benefit figures last refreshed before the market deteriorated. The model appeared rigorous but could not identify which commitments were reversible, which controls were mandatory, which programmes would release cash soon enough or which stopped projects would damage the work that remained.
This paper recommends a crisis portfolio-triage regime built around five tests: enterprise survival and mandatory obligation, near-term cash effect, dependency consequence, reversibility, and credible value under revised conditions. Decisions should be made in short waves, with explicit categories and named executive owners. Minimum evidence must be good enough for a consequential choice, not complete enough for an annual investment case.
The central conclusion is direct: fair-weather prioritisation cannot be accelerated into crisis triage merely by holding more frequent meetings. Organisations need a different governance mode, designed for constraint, uncertainty and irreversible choice.
The Week the Matrix Stopped Helping
The portfolio committee had spent two years refining its scoring model. Strategic alignment carried 30 per cent, financial return 25 per cent, risk reduction 20 per cent, customer effect 15 per cent and delivery confidence 10 per cent. Each proposal passed through finance, architecture and business review. The method was consistent, documented and regarded as a marked improvement over executive preference.
Then the funding instruction arrived: reduce the change portfolio by £95 million before the next board meeting.
The committee met with a ranked list of 86 initiatives. Thirty-one were labelled strategically essential. Nearly every regulatory or risk proposal scored above 70. Business sponsors had learned to describe benefits in the language of the matrix, so different initiatives clustered within a narrow band. The highest-ranked programme required a further £34 million before producing any saving. A lower-ranked infrastructure renewal supported nine supposedly more strategic projects. Three initiatives could be stopped immediately but would release little cash because supplier commitments had already been made.
The ranking answered the question for which it had been designed: which proposals appear most attractive against common criteria? It did not answer the question now confronting the organisation: which combination of decisions preserves viability and future capability while releasing cash within the required period?
In a crisis, the portfolio is not a ranking problem; it is a sequence of irreversible choices.
Why Fair-Weather Prioritisation Failed
Weighted scoring frameworks are not inherently weak. They counteract the loudest sponsor, create a common vocabulary and force proposals to address strategy, value and risk. Under normal conditions, they help decide which new work should enter a constrained portfolio.
The strongest defence of the matrix is that crisis is precisely when consistency matters. Without criteria, anxious executives may protect familiar programmes, sacrifice long-term capability and make politically convenient cuts. A transparent model can prevent panic from becoming policy.
That defence is correct but incomplete. Consistency of calculation does not guarantee relevance of decision. The crisis changed the meaning of the inputs and introduced factors the original model did not contain.
Strategy had moved faster than the scores
Many initiatives were aligned to a strategy formed under assumptions of growth, available credit and stable demand. A project could remain perfectly aligned to yesterday’s strategy while being unaffordable or irrelevant to the conditions of 2009.
Strategic alignment became a lagging indicator. Unless the strategy itself was restated for survival, liquidity, cost and selective investment, the highest score merely rewarded fidelity to expired assumptions.
Benefits were neither comparable nor current
Benefit estimates had been produced at different times, with different methods and different levels of sponsor optimism. Some reflected revenue growth no longer plausible in recession. Others depended on later projects that might now be stopped. Cost avoidance, cash saving and accounting benefit appeared in the same portfolio total even though only some affected immediate funding.
When the constraint is cash within the year, a five-year return cannot be treated as equivalent to a cost removed next quarter.
Dependencies were represented as annotations
Traditional portfolio records often note dependencies but do not value them. A programme may be low-ranked on its own and still provide data, infrastructure, premises or process change required by several higher-ranked initiatives. Stopping it can strand expenditure elsewhere.
In the representative portfolio, one £8 million integration programme sat in the lower half of the ranking because its direct benefits were modest. Yet seven retained programmes assumed its interfaces would exist. Cancelling it would either stop £52 million of other work or force each programme to build a local alternative. The ranking treated the programme as an item; the portfolio depended on it as a connective asset.
Sunk cost distorted judgement in both directions
Some executives argued that heavily spent programmes must continue because too much had already been invested. Others argued that past expenditure was irrelevant and only future value mattered.
Both positions can be too simple. Sunk cost should not justify bad future spending. But the current state created by past spending matters: contracts, partially implemented systems, trained staff, customer commitments and unfinished migrations. The decision is not whether to recover the past. It is whether stopping from the present state creates additional cost or risk greater than finishing to a safe point.
Governance was built to approve, not stop
Investment governance had clear gates for business-case approval and funding release. It had weaker mechanisms for pausing or terminating live work. Sponsors owned benefits but not necessarily closure costs. Programme managers were expected to deliver, not propose their own cancellation. Finance could identify budget but often not the contractual or operational consequence of withdrawal.
The portfolio could start work through a defined process. It could not stop work through an equally mature one.
The Nature of Crisis Triage
Crisis triage begins with a different objective. It is not to identify the theoretically best portfolio. It is to construct a survivable portfolio under immediate constraint while preserving the greatest reasonable capacity to recover.
That objective creates four decision categories.
| Category | Meaning | Required evidence |
|---|---|---|
| Protect | Continue with priority access to scarce funding and leadership attention | Mandatory need, survival value or exceptional near-term return |
| Pause | Halt new commitment while preserving a credible route to resume | Reversibility, holding cost, dependency treatment and review date |
| Redesign | Continue only after changing scope, sequence, commercial terms or delivery model | Minimum viable outcome, revised cost and consequence of simplification |
| Stop | Terminate and release resources; accept closure consequences | Cash release, contract exit, asset treatment, people impact and stranded dependencies |
A fifth label — “continue” — is deliberately avoided. It is too passive. In crisis conditions, continued work consumes scarce cash and attention; it should either be protected for a reason or redesigned to fit the new constraint.
Triage also changes the standard of evidence. Waiting for fully refreshed business cases across 86 initiatives can take months, by which time the organisation will have continued spending by default. But speed cannot become licence for unsupported cuts.
The right standard is minimum decision evidence: the smallest reliable set of facts needed to understand consequence, cash and reversibility.
The purpose of crisis governance is not to remove uncertainty before deciding; it is to make uncertainty visible enough that leaders can choose which consequences to accept.
Five Tests for a Defensible Decision
Each initiative should pass through five tests in sequence. The order matters because it prevents attractive long-term narratives from obscuring immediate obligations.
Survival and mandatory obligation
Does the initiative protect solvency, continuity of critical operations, legal or regulatory compliance, or a contractual commitment whose breach would create disproportionate harm?
This category must be narrow. During triage, sponsors frequently recast desirable work as mandatory. The evidence should identify the precise obligation, date, minimum scope and consequence of non-compliance. “Risk reduction” is not sufficient; the risk and required treatment must be explicit.
Near-term cash effect
How much cash will each decision release, in which month, and with what unavoidable closure cost?
Portfolio totals must distinguish:
- expenditure already committed and not recoverable;
- expenditure avoidable within the current year;
- future-year cost avoided;
- accounting benefit without immediate cash effect;
- one-off termination or write-off cost.
This prevents a portfolio from announcing a large reduction that arrives too late to meet the actual constraint. It also exposes programmes whose apparent cancellation saves little because the organisation has already crossed its major commitment points.
Dependency consequence
What retained work, operation or benefit is damaged if this initiative pauses or stops?
The analysis should follow both directions. First, what does the initiative depend upon? Second, what depends upon it? The objective is not to preserve every dependency; that would make reduction impossible. It is to see the chain and decide it consciously.
A simple dependency map often reveals that several projects should be stopped together, or that a modest enabling component should be protected while a more visible programme is redesigned.
Reversibility
Can the decision be changed later, at what cost and within what time?
A pause may preserve option value where assets remain usable, suppliers can be remobilised and knowledge can be retained cheaply. In other cases, pause is the worst of both worlds: the organisation pays holding cost, loses key staff and must repeat design work before restarting.
Reversibility should include:
- contractual break and re-entry terms;
- loss of specialist knowledge;
- expiry of licences, assets or premises options;
- customer or workforce commitments;
- deterioration of incomplete data or system migrations;
- the cost and lead time to restart.
Credible value under revised conditions
Does the initiative still create value under the demand, cost and strategic assumptions now considered plausible?
This is not a full reapproval exercise. It is a stress test. Revenue benefits should be recalculated under lower demand. Cost savings should show how they become cash. Programmes justified by growth should demonstrate relevance to consolidation or efficiency. Benefits dependent on stopped initiatives should be removed.
An initiative may still be valuable and still need to stop. Triage recognises that funding constraint can make good work unaffordable. The decision record should say so plainly rather than manipulate the score until the desired cut appears rational.
A Worked Portfolio Reduction
Consider a change portfolio with £240 million of remaining planned expenditure across 86 initiatives. The instruction is to remove £95 million while protecting mandatory obligations and the organisation’s ability to operate.
The first review identifies:
- £28 million across eleven initiatives that is legally, operationally or contractually mandatory at minimum scope;
- £47 million already committed across the portfolio and unlikely to be recovered;
- £19 million of enabling work supporting multiple initiatives;
- £36 million in initiatives whose benefits depend on demand assumptions no longer credible;
- £22 million in programmes that can be simplified rather than stopped;
- £31 million in activity that can be paused with relatively low holding cost.
The categories overlap, so they cannot simply be added. The portfolio team constructs decision packages rather than individual rankings.
One package combines stopping a customer-growth programme, cancelling two dependent channel projects and preserving a small data-quality component needed by operations. It releases £26 million over nine months, incurs £3 million in closure cost and avoids £18 million of later expenditure.
A second package redesigns a five-region operating-model programme into two regions and a common finance control. It reduces the next twelve months’ spend from £42 million to £24 million while preserving the elements needed for later expansion.
A third package protects the £8 million integration programme because it prevents duplication across seven retained initiatives. Its low standalone benefit is outweighed by £14 million of avoided local interface work.
After two decision waves, the portfolio reaches £98 million of reductions:
| Decision | Initiatives | Twelve-month cash reduction | Principal consequence |
|---|---|---|---|
| Stop | 24 | £43m | Closure costs, write-offs and lost planned benefits |
| Pause | 17 | £21m | Holding cost and restart exposure |
| Redesign | 13 | £34m | Reduced scope and later benefit |
| Protect | 32 | £0m | Concentrated funding and delivery attention |
The result is not a newly optimised portfolio. It is a defensible survival portfolio: affordable, explicit about consequence and coherent enough to govern.
Governance at Crisis Speed
The decision process must be fast without becoming arbitrary.
Establish a single constraint and decision authority
The board should state the cash reduction, timing, non-negotiable obligations and risk appetite. One executive forum must own the integrated decision. If each function protects its own work before the portfolio is assembled, the reduction will fall on politically weak initiatives rather than enterprise logic.
Use a small triage cell
A temporary portfolio cell should include finance, portfolio leadership, commercial expertise, enterprise architecture, risk and operational representation. Its task is to prepare decision evidence, map dependencies and test sponsor claims. It does not make the final choices.
The cell needs direct access to contracts, forecasts, programme schedules and key decision owners. It should work in days, not wait for normal monthly reporting.
Decide in waves
Trying to resolve all initiatives in one meeting creates superficial debate. A better sequence is:
- confirm mandatory minimums and the funding constraint;
- identify obvious stops and pauses with high cash release and low dependency consequence;
- review dependency clusters and redesign candidates;
- address the final marginal choices needed to close the gap;
- validate that the resulting portfolio remains operationally coherent.
Each wave updates the cash position and exposes the next decision frontier.
Record consequence, not just status
Every decision should state:
- category and effective date;
- twelve-month cash effect and closure cost;
- accepted operational, contractual and benefit consequence;
- dependencies to terminate, preserve or redesign;
- named executive owner;
- next review date for pauses and redesigns.
This record protects institutional memory and prevents stopped work from returning through uncoordinated local approvals once immediate pressure eases.
The Limits of Triage
Triage is not a superior form of normal portfolio management. It carries real risks.
Fast decisions can destroy capability whose value is difficult to quantify. Short-term cash can dominate long-term position. Political sponsorship can still influence which evidence is challenged. Paused work can decay unnoticed. Repeated crisis reviews can exhaust teams and make all investment appear temporary.
These risks strengthen the case for explicit consequence and reversibility; they do not justify returning to a scoring matrix that no longer answers the decision.
Nor should crisis become a permanent management style. Once funding stabilises, the organisation should rebuild a strategic portfolio, refresh benefits and re-examine paused work. The triage categories are a bridge through constraint, not an enduring substitute for investment discipline.
The strongest opposing view is that abandoning established prioritisation under pressure invites arbitrary executive cuts. That danger is serious. The prescription here does not abandon criteria; it changes them to match the decision. Survival, cash timing, dependencies, reversibility and revised value are criteria. They are simply more relevant than the pre-crisis weighting of strategic alignment and return.
Recommendation
Organisations facing severe portfolio reduction should activate a defined crisis-triage regime rather than compress the annual prioritisation process.
The regime should:
- replace fine-grained rankings with protect, pause, redesign and stop decisions;
- restate strategy around current survival, control and recovery priorities;
- require minimum decision evidence on cash, dependency, reversibility and value;
- evaluate clusters of connected initiatives, not projects in isolation;
- distinguish committed cost, avoidable cash and future-year benefit;
- place integrated authority in one executive forum;
- make decisions in waves and record accepted consequences;
- preserve review dates and knowledge for paused work.
After the immediate reduction, the organisation should retain the capability it was forced to build: current portfolio data, visible dependencies, stop-work governance and credible cash measures. These are useful in stable conditions as well. The crisis has merely made their absence impossible to ignore.
From Ranking to Choice
Weighted matrices gave organisations a sense of fairness during years when many proposals could proceed and disagreements could be resolved through incremental funding. They were useful tools for admission to the portfolio.
The financial crisis has created a different task. Budgets are not being adjusted at the margin. Whole groups of commitments must be reconsidered while markets, demand and organisational capacity remain uncertain. The cost of delay is continued expenditure, and the cost of error may be the loss of an essential capability.
Portfolio triage accepts that there is no painless ranking hidden in the data. Every serious reduction transfers cost, risk or lost opportunity somewhere. Governance becomes credible when it makes that transfer explicit and assigns authority for it.
The organisations that navigate 2009 best will not be those with the most sophisticated scoring model. They will be those able to see their commitments clearly, understand the connections between them and stop work without losing control of what remains.
Fair-weather prioritisation asks which initiatives deserve to start. Crisis triage asks which future the organisation can still afford to preserve.