Decision friction is a portfolio risk

What the paper develops

Portfolios track work, dates, spend, and risk. The decisions that control the work often remain scattered across meetings, inboxes, and personal notes. When no one sees the queue, no one manages it.

Consider an invented late program, offered as an illustration and not as a reported case. Its post-mortem starts with estimation, complexity, or vendor performance. Then someone walks through the calendar and finds five weeks spent waiting for a policy decision, a funding answer, or an owner who was never named. The question was raised early. It moved through several forums, gathered opinions, and produced no decision.

Decision friction is the time and effort required to turn a recognized question into an answer the work can use. It becomes a portfolio risk when teams wait, restate the question, or build against assumptions while the answer is pending. The research cited is general, and the link to portfolio risk is this paper's reading.

Use one visible inventory for decisions that matter. Each entry should state the choice, name the person who can make it, identify the evidence that could change it, set a latest useful date, show the work at risk, and record the outcome.

Write down one answerable choice. "Architecture concerns" is a theme. "Extend the billing platform or adopt the vendor module for the mid-market by March" gives a forum something it can decide.

Assign the decision to the person with authority over the area in dispute. A committee may advise; it is the owner only when its charter gives it authority to decide. A distribution list is a notification channel, never an owner.

Before commissioning more analysis, ask whether a plausible result could change the choice. If it cannot, the analysis only extends the queue. When uncertainty is real, record it and make the tradeoff between more information and more waiting explicit.

Work backward from the first commitment that depends on the answer. The latest useful date is the last point at which the decision can still change the work. "This choice loses its cheapest option on April 12" gives the forum a concrete consequence of delay.

Show the investments, constrained roles, and dependencies waiting on the answer. The same policy question blocking three programs is not three separate project issues. It is one portfolio bottleneck.

Record the answer, the reasoning, the caveats, and the evidence that would justify reopening it. Then confirm that affected owners have updated their plans. A decision is complete when the work changes, not when the meeting ends.

The inventory turns a vague delay into a specific management problem. An aging queue can reveal that a particular forum is too slow, a decision right is unclear, or a recurring question should become standing policy. It also protects teams from the opposite mistake: treating every request for speed as a reason to skip needed evidence or review.

Ask each program lead for the three questions holding up work today. Start with the ones that cross team lines, tie up constrained capacity, set precedent, or are hard to reverse. Review aging decisions and approaching dates inside the existing portfolio forum. A separate decision board often creates another queue.

The inventory cannot make leaders choose between competing priorities. It can make the cost of waiting visible: this decision, this owner, this date, and this work held up. That is enough to make an avoidable delay governable.