A portfolio is not a list of every project someone once approved. It is a deliberate set of investments chosen to fit strategy and capacity — which means the PMO's most useful product is often a well-argued no.
Select the work; do not just inventory it
Project management asks 'are we doing this right?' Portfolio management asks 'are we doing the right things?' If the office only reports status on whatever is already running, it is a news service. Selection is the job: compare proposed work against strategy, capacity, risk, and the benefits you can actually measure.
Use a short intake: problem, outcome, cost range, who will own the benefit, and what stops if we say yes. Demand that arrives as a slide titled 'strategic initiative' with no capacity ask is not a proposal. It is a wish that wants a colour on a roadmap.
Score enough to rank, not enough to fake science. A simple forced rank — strategy fit, value, risk, effort — beats a 28-criteria model nobody believes. When two scores are close, that is a leadership conversation, not a reason to add another weighting until the favourite wins.
Balance the mix against real capacity
Portfolios die of addition. A new regulatory project does not create new senior engineers. Map the scarce people and the scarce calendar — year-end freeze, audit window, peak trading — and show what gets delayed if the new work starts. Capacity is the constraint most boards prefer not to see.
Balance types of work, not just totals. All-run-the-business and no-change is stagnation. All-transformation and no-keep-the-lights-on is how the operational platform catches fire during the transformation. A healthy mix is a decision: how much change this organisation can absorb this year.
Kill or pause on purpose. A portfolio that never stops anything is not ambitious. It is unmanaged. Write the stop rule before you need it: benefit no longer stands, a higher-value item needs the same people, or a gate condition was missed. Stopping work is a portfolio success if the capacity goes somewhere better.
Run a conversation, not a parade
The portfolio review should decide: start, stop, slow, or fund a change. If the hour is a tour of green-amber-red tiles, you have built a cinema. Send the pack in advance. Spend the room on the three items that need a choice.
Make trade-offs visible in one view: what we would drop to start the new thing. Executives who 'want it all' are often not being greedy. They have never been shown the shared resource. Your job is the picture, not a lecture about prioritisation theory.
Keep a parked list. Ideas that lost this quarter should not vanish into folklore and then reappear as emergencies. A visible park — with a next review date — is how you stay honest without pretending every good idea can start on Monday.
Summary & next steps
Portfolio work is selection and balance against strategy and capacity. Rank simply, show what stops if something starts, and use the review to decide — not to narrate.
Next: maturity models — how to assess the office and the organisation without turning 'level 3' into a personality.