The Right to Stop
Researched by an agentic pipeline · reviewed and gated by the author
A portfolio has not stopped work until the released money, talent and attention have somewhere better to go.
The right to stop
Portfolio meetings are often full of information and short of decisions.
Every initiative has a dashboard. Each sponsor can explain the delay, defend the revised business case and point to the next milestone. Risks are coloured, benefits are restated and dependencies are logged. Yet the portfolio itself barely changes. Money remains where it was assigned. Scarce people stay attached to yesterday’s priorities. Work that no longer earns its place survives because no one in the room has a legitimate way to withdraw support.
That gap matters more than a new generation of portfolio software.
Broadcom-sponsored research published in 2026 reports that 95 per cent of organisations consider strategic portfolio management essential, but only two in ten have the data needed to prioritise work properly. [S1] Tempo’s separate survey of 667 planning and PMO leaders across 43 countries associates integrated processes, scenario planning and frequent adjustment with much stronger reported returns. Organisations with those traits said 81 per cent of projects delivered ROI, compared with 45 per cent among those without them. Frequent reviewers also cancelled more projects while reporting higher ROI. [S2] [S3]
These are vendor-linked surveys, not causal proof. They do, however, expose a useful contradiction. Strategic portfolio management is widely recognised, extensively tooled and still often unable to move resources.
The missing capability is withdrawal authority: a governed right to stop, pause, sequence or reduce work and to redeploy the resources it releases.
Visibility is not control
Portfolio visibility is necessary. Leaders cannot compare investments when cost, capacity, benefits and dependencies sit in separate systems. But a better view does not itself change an allocation.
Control requires five connected elements.
First, evidence must be comparable. That does not mean reducing regulatory compliance, customer trust, infrastructure resilience and revenue growth to one artificial financial score. It means making assumptions, baselines, time horizons, confidence and resource demands explicit enough to test alternatives.
Second, decisions need a recurring cadence. An annual planning cycle asks what should begin. A live portfolio also asks what should continue. Reviews must occur often enough to respond to changed evidence, but not so often that delivery is destabilised by executive restlessness.
Third, someone must hold decision rights. A portfolio office that can aggregate but only recommend will become a reporting service. The authority may remain with an executive committee, CFO or investment board. What matters is that the right to withdraw is explicit, not inferred during a crisis.
Fourth, resources must be movable. A project can be cancelled on paper while its budget remains trapped, its people cannot transfer and its supplier commitments continue. A stop decision creates value only when capacity can reach a stronger alternative.
Fifth, the organisation must learn. Termination evidence should improve future selection, estimates and option design. If stopping is treated as individual failure, sponsors will conceal bad news and the portfolio will repeat the same errors.
Break any link and governance becomes theatre. Data without cadence grows stale. Cadence without authority produces discussion. Authority without mobility produces symbolic decisions. Mobility without learning creates churn.
Why weak work survives
The resistance is not merely technical.
Project termination research describes escalation of commitment: decision makers receive evidence that an investment will not deliver its expected return yet continue to commit resources. Sunk cost, self-justification and optimism all contribute. In one experimental study published through the Project Management Institute, participants continued failing projects while constructing unsupported expectations of future benefits. [S4]
This is a predictable consequence of organisational design. Sponsors build status by securing investment. Delivery teams build identity around the initiative. Suppliers and internal functions develop economic interests in continuation. Near completion, the goal can quietly shift from achieving an outcome to finishing the project.
Normal reporting reinforces the problem. Status asks whether work is on plan. Portfolio control asks whether the plan still deserves resources compared with current alternatives. A project can be green and strategically obsolete; another can be red because it is solving the harder, more valuable problem.
Benefits management offers a better basis. PMI guidance explicitly recommends using benefit management to stop bad projects while warning that not all benefits can be predicted in advance. [S5] That distinction is important. Withdrawal discipline is not intolerance of uncertainty. Exploratory options need room to learn. The question is whether they are producing the evidence they were funded to discover.
A withdrawal decision
Consider an illustrative enterprise with 80 active initiatives and twelve data engineers shared across them.
One customer platform is six months late. Its sponsor argues that most of the cost has already been spent and that abandoning it now would waste the investment. A second initiative could prevent a material regulatory exposure but cannot begin because the engineers are committed.
A status review examines the recovery plan. A withdrawal review starts elsewhere.
It excludes sunk cost and compares forward choices: remaining cost, probability and timing of benefits, salvageable components, dependency effects, regulatory consequences and the value of the best alternative use of the engineers. It asks whether scope can be reduced, whether the platform can become a bounded option and what commitments make capacity genuinely unavailable.
The decision may still be to continue. Withdrawal authority is not a presumption that stopping is wise. It is a legitimate method for considering stopping without requiring a project to become an undeniable disaster.
If the platform is paused, the meeting is not finished. Leaders name where the engineers move, which contracts are unwound, what knowledge is retained and when the decision will be reviewed. The reallocation is the outcome; the cancellation is only an action.
A portfolio has not stopped work until the released money, talent and attention have somewhere better to go.
The five tests
Executives can assess whether their portfolio function has real withdrawal authority with five tests.
| Test | Evidence |
|---|---|
| Comparability | Alternatives show forward value, capacity, confidence and dependency assumptions |
| Cadence | Continuation is tested at defined points, not only after failure |
| Authority | Named leaders can pause, stop, sequence and reduce work |
| Mobility | Budgets, people and supplier commitments can be redeployed |
| Learning | Stop decisions change future selection and option design |
The tests should be applied to one bounded portfolio before a major governance redesign. Select a group of initiatives competing for the same scarce capability. Establish independent challenge, pre-agreed continuation criteria and a small set of possible decisions: continue, accelerate, reduce, pause, combine or stop.
Track more than the number cancelled. Measure time from adverse evidence to decision; capacity actually released; capacity successfully redeployed; benefits forgone and recovered; and whether later information confirms the decision. A high cancellation rate is not automatically healthy. It may indicate careless entry decisions, excessive strategic volatility or a culture that abandons difficult work.
The danger of continuous reprioritisation
The strongest case against this model deserves equal weight.
Tempo’s findings are associations in a self-reported, vendor-connected survey. Organisations that plan continuously may perform better because they already have stronger leaders, cleaner data and healthier portfolios. Cancellation may be a symptom of good governance, not its cause. The 81-to-45 per cent contrast should not be presented as the effect of buying a system or adopting a meeting cadence.
There are also portfolios where withdrawal logic must be constrained. Safety-critical, legally mandated and cyber-resilience work may have non-negotiable outcomes. Large infrastructure commitments can be economically indivisible. Early research options may appear inefficient precisely because their purpose is to preserve future choices. Frequent reviews can reward visible short-term returns and destroy patient capability building.
Scenario models introduce their own risk. They can turn uncertain benefits into precise-looking rankings and obscure the political choices embedded in weights and assumptions. A central portfolio team can accumulate authority without accountability for operating results.
The answer is not to abandon withdrawal authority but to bound it. Define protected categories. Separate mandatory outcomes from discretionary delivery choices. Use ranges and confidence rather than false point estimates. Require an independent view when the decision threatens a sponsor’s reputation. Make the portfolio board accountable for allocation quality while leaving operating leaders responsible for realised outcomes.
From annual budgets to live choices
The basic argument is not new. PMI has long defined strategic portfolio management as selecting and prioritising programmes and projects while aligning resource demand with availability. [S6] McKinsey research published a decade ago found that companies reallocating resources dynamically generated higher average shareholder returns than sluggish reallocators; it also found that many companies moved very little capital from year to year. [S7]
What is newly consequential is the 2026 combination of executive attention, proliferating technology and AI investments, and persistent inability to access prioritisation data. A current Harvard Business Review analysis makes a related point: organisations invest heavily in strategy design but less in mobilisation, the work between decision and execution. [S8]
This changes the design question. The objective is not a complete central catalogue of work. It is an operating mechanism that can make live choices across competing commitments.
Technology can help by connecting financials, workforce capacity, outcomes and dependencies. It can make scenarios faster and expose contradictions. It cannot decide whose forecast to distrust, whether a regulatory obligation outweighs a commercial return, or how much institutional patience a strategic option deserves. Those remain acts of judgement and authority.
The portfolio office therefore should not seek the power to own every outcome. It should make alternatives visible, enforce the discipline of forward-looking evidence, surface the cost of continuation and ensure that authorised decisions result in real movement. Executives retain the choice. Business leaders retain delivery accountability.
The portfolio that can change its mind
Most organisations can approve work. Fewer can reverse themselves before failure becomes public and undeniable.
That asymmetry is expensive. Entry decisions occur when optimism is high, benefits are abstract and political support is concentrated. Exit decisions arrive after reputations, teams and contracts are attached. If governance treats both moments as equivalent, continuation will win.
A mature portfolio designs withdrawal in at entry. It names review points, evidence thresholds, salvage options and alternative uses of scarce resources before commitment escalates. It rewards sponsors for surfacing disconfirming evidence. It distinguishes a valuable option from a protected zombie. It records not simply that an initiative ended, but what the enterprise did differently as a result.
The 2026 evidence does not prove that continuous planning or integrated SPM causes performance. It does support a sharper executive question than whether the organisation has a portfolio system.
Can it change its mind?
If the answer is no, greater visibility may produce a more accurate picture of strategic inertia. If the answer is yes—and money, talent and attention actually move—the portfolio becomes what it was meant to be: a mechanism for choosing the future, repeatedly, under changing evidence.
Sources
- Broadcom ValueOps / Dimensional Research — The State of Strategic Portfolio Management: 2026 Global Trends — 2026 — https://valueops.broadcom.com/the-state-of-strategic-portfolio-management-2026
- Tempo — The 2026 State of Strategic Portfolio Management report — 2026 — https://www.tempo.io/guides/2026-state-of-spm-report
- Atlassian Community — New research: The 2026 State of Strategic Portfolio Management — 24 February 2026 — https://community.atlassian.com/forums/App-Central-articles/New-research-The-2026-State-of-Strategic-Portfolio-Management/ba-p/3197046
- Project Management Institute — The psychology of project termination — 28 October 2013 — https://www.pmi.org/learning/library/psychology-project-termination-decision-maker-5914
- Project Management Institute — Beyond project success: Benefits realization opportunities and challenges — 2014 — https://www.pmi.org/learning/library/success-benefits-realization-opportunities-challenges-9272
- Project Management Institute — Strategic portfolio management: gap between strategy and results — 2006 — https://www.pmi.org/learning/library/strategic-portfolio-management-strategy-results-8111
- McKinsey & Company — How nimble resource allocation can double your company’s value — 30 August 2016 — https://www.mckinsey.com/capabilities/strategy-and-corporate-finance/our-insights/how-nimble-resource-allocation-can-double-your-companys-value
- Harvard Business Review — How to Ensure Your Company Acts on Your New Strategy — 8 July 2026 — https://hbr.org/2026/07/how-to-ensure-your-company-acts-on-your-new-strategy