The purpose of a portfolio model is to improve judgement, not to pretend judgement is no longer required.
A project scores 92.4 on the portfolio model. Another scores 89.7. The first is therefore ranked higher, approved and funded.
The numbers look precise. The decision may not be.
What if the benefit estimate behind the 92.4 score depends on a market assumption that is already weakening? What if the second initiative unlocks a capability needed by three other programs? What if the scoring model rewards short-term financial return but underweights regulatory exposure, customer trust or scarce engineering capacity? What if the executive team quietly agrees that one project is politically untouchable?
Portfolio management needs formal methods. Without them, power, enthusiasm and habit can dominate allocation. But formalisation creates another danger: it can make uncertain and negotiated decisions appear objectively solved.
They are not.
The Strategic Context
Traditional portfolio thinking often starts with a rational sequence: define criteria, gather data, assess projects, compare alternatives, select the best mix and allocate resources. This approach has obvious value. It creates common language, makes assumptions more visible and disciplines decisions that might otherwise be driven by lobbying.
Martinsuo's 2013 review, however, shows why this is incomplete. Empirical research on project portfolio management increasingly depicts portfolio activity as negotiation and bargaining, situational judgement and structural reconfiguration as well as rational analysis. Project selection takes place with imperfect knowledge. Information continues to evolve. Organisational context influences which practices work. Risks, uncertainty and changing conditions affect how the portfolio is managed.
Guitarte's case study takes the argument further by framing portfolio prioritisation as a wicked problem: multiple stakeholders hold different objectives, problem definitions are unstable, consequences can be difficult to test in advance and every major allocation decision changes the situation that future decisions must address.
The implication is not that portfolio models are useless. It is that they are representations of the decision, not the decision itself.
What Leaders Commonly Misread
One common assumption is that a sufficiently sophisticated scoring model removes politics. It usually does not. It often moves politics upstream into the selection of criteria, weightings, thresholds, benefit assumptions and data definitions.
Another is that there is one objective function. Executives are asked to maximise value, protect compliance, maintain customer commitments, build future capability, control risk, preserve liquidity and stay within workforce capacity simultaneously. These objectives can conflict.
A third is that better data will eventually make the decision obvious. More complete information is valuable, but portfolio choices remain forward-looking. Benefits, technology readiness, market responses and operating impacts are partly uncertain by nature.
The fourth is that “optimal” means durable. A portfolio can be the best available choice today and be wrong three months later because its assumptions changed.
The fifth is that negotiation is evidence of weak governance. Some negotiation is unavoidable because different functions legitimately see different dimensions of value. The governance failure is not disagreement. It is allowing disagreement to remain hidden, personalised or unresolved.
Reframing the Issue
Portfolio management is better understood as structured judgement under uncertainty.
The objective is to improve the quality, traceability and adaptability of enterprise choices even when no mathematically final answer exists.
This reframing produces a useful distinction:
- Analysis should narrow the field, expose trade-offs and challenge unsupported claims.
- Judgement should resolve what the model cannot determine.
- Governance should make that judgement accountable and revisable.
A model that pretends to eliminate judgement can be more dangerous than no model at all because it creates false confidence.
Related article: From Framework Knowledge to Executive Judgement: Diagnose Before You Recommend
Why Portfolio Problems Become Wicked
Guitarte's practitioner case highlights several characteristics that make portfolio choices difficult to reduce to a linear problem.
First, stakeholder groups can define success differently. A strategy owner may ask whether an initiative advances enterprise direction. Finance and risk may ask whether it produces acceptable value within constraints. Delivery teams may ask whether it is feasible with the available architecture, resources and time. Guitarte simplifies these tensions into three useful questions: is it strategic, is it valuable and is it doable?
Second, consequences are interdependent. Funding one initiative can remove scarce resources from another. Delaying a platform program can weaken several customer projects. Cancelling a project may preserve capital but destroy an option that would have mattered under a different scenario.
Third, portfolio decisions are path dependent. Once the organisation has invested political capital, customer commitment, technical design and specialist time, future choices become harder. Sunk cost can become institutional commitment.
Fourth, the problem changes while it is being solved. By the time leaders have produced a perfect analysis, the environment may have invalidated it.
This is why “good enough” can be a disciplined concept rather than a lowering of standards. A defensible portfolio may be one that satisfies critical strategic, value and feasibility thresholds while retaining enough flexibility to adapt when evidence changes.
Rationality Still Matters
Rejecting false precision does not mean accepting intuition without discipline.
The rational portfolio model still performs three essential functions.
It creates comparability. Without a shared assessment architecture, every initiative arrives with its own vocabulary and business-case logic.
It creates transparency. Explicit criteria force leaders to reveal which dimensions they value and which assumptions drive a recommendation.
It creates challenge. A structured process makes it easier to ask why a project with weak strategic contribution remains protected or why a high-value initiative has no credible delivery path.
The mistake is expecting the model to do more than it can.
Decision Framework: The Defensible Portfolio Test
A portfolio decision is stronger when it can pass seven tests.
1. Strategic relevance
Can the initiative's contribution to a current strategic objective be explained without vague association? “Supports digital transformation” is not enough. Which capability, outcome or value driver changes?
2. Value logic
What benefit is expected, for whom, over what horizon, and what assumptions make that benefit plausible? Include non-financial value where material, but do not let non-financial language become immunity from scrutiny.
3. Doability
Is the work feasible with available capability, architecture, market capacity, time and organisational absorption? Strategic importance does not create delivery capacity.
4. Portfolio interaction
What dependencies, synergies, resource conflicts and risk concentrations appear only when the initiative is considered alongside other work?
5. Sensitivity
Which assumptions would change the decision if they moved? If a one-point change in a subjective score reverses the ranking, the apparent precision of the model should not drive the decision.
6. Reversibility
Can the organisation test, stage or defer commitment? Under uncertainty, reversible decisions can create information before irreversible capital is committed.
7. Decision expiry
When should the decision be reconsidered? A portfolio approval should have a logical review condition, not an assumption of permanent validity.
Related article: The Portfolio You Govern Is Not Always the Portfolio You Have
From Strategy to Execution
Immediate action is to inspect the current portfolio model for hidden claims of certainty. Look for criteria whose weightings are treated as facts, benefit forecasts presented without ranges, and thresholds that imply precision the underlying evidence cannot support.
Medium-term capability building means separating the analytical process from the decision process. Analysts should be free to show uncertainty, alternative scenarios and sensitivity. Decision-makers should then record where they exercised judgement and why. This preserves accountability without forcing every choice into a numerical formula.
Long-term strategic positioning requires building a portfolio governance culture where changing a prior decision is not automatically treated as failure. If evidence changes, a high-quality governance system should be able to reverse, redesign or stop investment without needing to prove that the earlier decision was negligent.
That capability is difficult because organisations often reward consistency more than learning. Yet adaptive enterprises distinguish between indecision and legitimate revision.
Signals to Monitor
Be concerned when portfolio scores routinely cluster within narrow ranges but lead to materially different outcomes. That can indicate false precision.
Also watch for strategic initiatives that always win regardless of score, repeated overrides that are not documented, business cases that improve suspiciously when challenged, models whose weightings have not changed despite strategy changes, and projects approved despite no credible answer to the “doable” question.
Another warning signal is the absence of dissent. A complex portfolio with no disagreement may have exceptional alignment. It may also have suppressed challenge.
References
- Guitarte, A. 2015, 'A case study on taming the wicked problem of portfolio management', paper presented at PMI Global Congress 2015—North America, Orlando, FL.
- Martinsuo, M. 2013, 'Project portfolio management in practice and in context', International Journal of Project Management, vol. 31, no. 6, pp. 794–803.
Questions for the Leadership Team
- Which assumptions in our portfolio model are treated as objective even though they are judgement calls?
- Where do strategic value, economic value and deliverability currently conflict?
- Which initiatives would still be funded if their sponsors were removed from the discussion?
- What decision could we make reversibly instead of committing fully now?
- Which portfolio choices have been overridden outside the formal model, and were the reasons documented?
- When does each major portfolio decision expire and require re-examination?
Closing Perspective
Portfolio management becomes dangerous when analytical discipline is confused with mathematical certainty. Leaders need models that expose trade-offs, not models that hide judgement behind decimals. The goal is not to discover a permanently optimal portfolio. It is to make the best defensible choice with the evidence available, remain explicit about uncertainty and retain the governance discipline to change the choice when reality changes.
About EraNorth Insights
EraNorth Insights publishes practical analysis on strategy, projects, operations, transformation and decision intelligence for professional and organisational use. About EraNorth.