The hardest thing in portfolio management isn't saying yes. It's saying no - and meaning it.
Every organisation has more demand for change than capacity to deliver it. The natural response is to approve everything and let delivery sort it out. The result is a congested, under-resourced portfolio where everything is theoretically in flight and nothing is actually moving. Real portfolio governance requires the courage to prioritise ruthlessly, sequence deliberately, and protect the organisation's delivery bandwidth. It's a leadership discipline, not a planning one.
Delivery excellence is now the entry fee. Executives want a PMO that helps decide what to deliver.
PMI's 2026 research found that 78% of senior executives want more PMO investment in strategic alignment, against 68% of PMO leaders, and 65% want benefits measured after go-live. The mandate has moved from reporting on projects to supporting decisions about the portfolio: what to start, what to stop, and where value is at risk. That means fewer status packs and more trade-offs, capacity views and benefits at risk, delivered early enough to change a decision. AI will take over much of the reporting. What remains is judgement, and a seat where strategy is set.
Source: PMI, Bridging the Gap: Positioning PMOs as Indispensable Partners in Strategy Execution, 2026
Organisations invest in change. Very few invest in knowing whether it worked.
Benefits realisation isn't a reporting exercise - it's a discipline. Most organisations track activity and confuse it with progress. A project delivered on time and on budget is not a success if the business case never materialised. I've seen benefits frameworks transform from box-ticking exercises into genuine strategic levers. The difference is accountability, not methodology. Someone has to own the outcome after the project closes.
Every initiative is approved on its own merits. The people who must absorb them are never in the business case.
Most portfolios are governed by budget and delivery capacity. Very few are governed by the capacity of the people on the receiving end. The same teams are hit by several initiatives in the same quarter, each one reasonable, and the sum of them unworkable. Leaders also tend to think the organisation is coping better than it is: in one 2024 survey, 87% of executives said their teams had a structured approach to change, against 67% of the people doing the work. The fix is practical. Measure change load by business unit, sequence initiatives by what the organisation can absorb, and ask what will stop before anything new starts.
Source: Smartsheet, 2025 Project and Portfolio Management Priorities Report
The plan is not the programme. The programme is the thousands of decisions made when the plan meets reality.
I've led programmes with 3,000+ milestones. No plan survives that intact. What separates a well-run programme from a struggling one isn't the quality of the original plan - it's the quality of the decision-making infrastructure when things change. Clear escalation paths, fast information flows, a leadership team that responds to facts rather than narratives - these are the real foundations of programme success.
The riskiest change programmes are the ones that don't know they're at risk.
Change risk is poorly understood in most organisations. The tendency is to monitor execution risk - are we on schedule, on budget? - while underweighting the strategic risks that determine whether the programme should exist at all. Is the business case still valid? Is the organisation capable of absorbing this change at this pace? Are the dependencies being managed or just documented? I've built change risk frameworks that ask the uncomfortable questions before they become expensive answers.
AI won't replace portfolio managers. But portfolio managers who use AI will replace those who don't.
The most immediate value of AI in portfolio governance isn't automation - it's insight at speed. Feeding a portfolio of status reports, risk registers, and benefits data to a well-prompted AI model surfaces patterns, dependencies, and early warnings that would take a senior analyst days to compile. The skill is knowing which questions to ask, how to frame the context, and how to interpret the output. That's a practitioner discipline, not a technology one.
Emerging practice
