Rebuilding the Recovery Portfolio: Stop Ranking Promises
In an uncertain recovery, the portfolio is not a ranking of promises; it is a sequence of choices that preserves the right to change direction.
The green portfolio that could not be funded
The portfolio committee met with forty-seven initiatives and an apparent abundance of discipline. Each proposal had a business case, a five-year return, a risk score and a red-amber-green assessment. Thirty-one sat within a narrow band between 75 and 82 points. On paper, almost everything was attractive. In practice, the organisation had £82 million of requests competing for £38 million of available investment, and the scoring model could not say what should stop.
The revealing detail was elsewhere. Only nine initiatives had a named business owner who could state the operational measure that would move within six months. Seventeen depended on demand assumptions set before the downturn. Eleven required the same small group of systems analysts during the first quarter. Six were continuations whose original rationale had weakened, but whose prior expenditure made cancellation feel like waste.
This is the post-recovery portfolio problem in miniature. As demand begins to return, the pipeline fills faster than confidence. Projects deferred during the contraction reappear beside new opportunities. The temptation is to rebuild by restoring the old queue and scoring it more carefully. That is precisely the wrong response.
Portfolio rationalisation after a shock should not attempt to identify the perfect long-term ranking. It should create a disciplined sequence of commitments, each proportionate to what is actually known.
Why the old scorecard stopped discriminating
Traditional scoring models work best when proposals can be compared on reasonably stable assumptions. Weight strategic alignment, financial return, risk and regulatory need; calculate a total; fund from the top until the capital runs out. The method looks objective because unlike judgements have been converted into common numbers.
But the recovery period makes the inputs unstable in different ways. Forecast volumes may be recovering in one market and flat in another. Supplier prices can be held for only a few months. Customer behaviour is moving between branch, telephone and online channels. A project with a modest return but an early test of demand may be more valuable than a larger project whose attractive economics depend on three uncertain years.
The scorecard conceals this difference because it asks every proposal to perform the same act: turn uncertainty into a single estimate. Once that happens, arithmetic amplifies confidence rather than evidence. A carefully modelled fiction can outrank a modest proposal designed to learn.
| Question | Old portfolio logic | Recovery portfolio logic |
|---|---|---|
| What is being ranked? | Complete business cases | Commitments of different size and reversibility |
| What counts as value? | Forecast return | Return plus the value of resolving uncertainty |
| How is risk treated? | A deduction from the score | A reason to shorten the next commitment |
| What earns further funding? | Approval at the outset | Evidence produced at the next gate |
The mechanism matters. When a proposal receives full approval, teams, suppliers and executives organise around its completion. Cancellation then becomes an admission of failure. By contrast, when the first decision purchases a bounded piece of evidence—a prototype, a process trial, a negotiated option or a three-month release—the next decision can be made without defending the entire original forecast.
In an uncertain recovery, the portfolio is not a ranking of promises; it is a sequence of choices that preserves the right to change direction.
Rationalisation is not a purge
The word rationalisation often produces the wrong behaviour. Leaders announce a percentage reduction, every sponsor trims the stated cost, and weak initiatives survive in smaller form. The portfolio becomes cheaper without becoming clearer. The organisation still carries too many starts, too many dependencies and too many claims on scarce specialists.
A serious rationalisation begins by separating three kinds of commitment:
- Obligations that must be met to maintain service, safety, legal compliance or essential infrastructure.
- Options that can test a strategic opportunity through a bounded commitment and a near-term learning objective.
- Convictions that justify substantial, difficult-to-reverse investment because the evidence and strategic need are already strong.
The categories are not labels of prestige. They determine the funding logic. Obligations require challenge on scope and timing, not invented financial returns. Options require a hypothesis, a spending limit and an explicit next decision. Convictions require visible ownership, capacity and benefits measures strong enough to survive executive scrutiny.
Return to the forty-seven-item pipeline. Twelve initiatives were genuine obligations, consuming £14 million. Eight were convictions, but resource analysis showed that only four could start without overloading the systems-analysis team; these consumed £17 million. The committee did not divide the remaining £7 million among the other twenty-seven proposals. It funded seven options at no more than £1 million each, each with a decision within ninety days. The rest were stopped or held without active teams.
That decision reduced concurrent starts from forty-seven to twenty-three, and the critical analyst group moved from 164 per cent planned utilisation to 92 per cent. More importantly, the organisation bought seven pieces of evidence rather than seven new promises. Two options later justified larger investment, three were redesigned, and two stopped after exposing weak demand. The stopped work was not failure. It was the return purchased by small, reversible commitments.
The purpose of rationalisation is not to make every surviving proposal look stronger. It is to make the next portfolio decision cheaper, earlier and more honest.
The strongest case for rigorous scoring
The defence of the conventional scorecard deserves to be taken seriously. After a period of disruption, organisations need consistency. Without common criteria, influential sponsors can dominate, strategic fashions can displace essential investment, and every decision can become a negotiation. A quantified model imposes discipline and leaves an audit trail.
That argument is correct about the need and incomplete about the instrument. Comparability is valuable; false precision is not. The answer is not to abandon strategic criteria or financial analysis, but to apply them to comparable decisions.
An obligation should be compared with other ways of meeting the obligation. An option should be compared by the importance of its uncertainty, the cost of resolving it and the time to the next decision. A conviction should be compared with other major commitments on strategic contribution, return, execution capacity and downside. Combining all three into one total score creates consistency only in appearance.
There remains a place for net present value, payback and weighted criteria. They are useful tests of a proposition. They should not be allowed to disguise whether the organisation is deciding to comply, to learn or to commit.
Rebuild the pipeline around evidence
The annual planning cycle will not disappear, nor should it. Capital limits, operating plans and board oversight require a coherent view of the year. But an annual allocation should cease to mean an annual promise to complete everything approved in one sitting.
The practical shift is straightforward:
- Fund the next defensible commitment. Release only enough money and capacity to reach a meaningful evidence point.
- Name the decision that follows. Every option should state in advance what evidence will cause expansion, redesign or cessation.
- Constrain starts by scarce capacity. Financial affordability is not delivery capacity. Rank the bottleneck roles and refuse a start that depends on capacity already committed.
- Review continuations against current evidence. Prior expenditure is not a reason to continue. The relevant question is whether the next pound and the next month are still justified.
- Keep a visible reserve. A fully allocated portfolio cannot respond to evidence; it can only defend its plan. Holding back a modest portion of capital and key capacity makes reprioritisation real.
This approach can look less decisive than approving a complete programme of work. In fact, it is more demanding. Sponsors must specify what will be learned, finance must distinguish approval from staged release, and the portfolio office must track decisions and dependencies rather than merely consolidate status reports.
The pattern that recurs in rebuilding investment pipelines is simple: organisations get into difficulty not because they lack proposals, but because they confuse approval with knowledge. The recovery does not reward the portfolio that predicts most confidently. It rewards the portfolio that commits with discipline, learns before inertia takes hold, and can still change its mind.