Project Portfolio Management: How Companies Choose What to Fund

A project can be well worth doing and still be the wrong one to fund, because the people it needs are already busy elsewhere. Project portfolio management is how an organization makes that call on purpose, by looking at all its projects together instead of one at a time. Proposals are collected, scored against strategy, balanced for risk and timing, approved only up to the capacity the organization has, and reviewed until they finish or are stopped. Surveys of product-development portfolios in North America and Australia found that managers rate one problem worst: having more projects than people to do them.¹

You can test your own portfolio without any software. List every project in progress or proposed with the people each one needs, then compare the total with the people free next quarter. When the list asks for more, someone is already making portfolio decisions, only by accident and in every team’s calendar instead of in one room.

What project portfolio management covers

Project portfolio management is the practice of managing a group of projects, programs and related work as one portfolio to meet an organization’s strategic goals. The Project Management Institute’s Standard for Portfolio Management (fourth edition, 2017) defines portfolio management as the centralized management of one or more portfolios to achieve strategic objectives.²

The three levels answer different questions. A project delivers a specific result, such as a new website. A program groups related projects so that together they produce a benefit, such as moving every customer service channel online. A portfolio decides which projects and programs the organization should run at all, and how to share money and people between them; PMI’s standard describes it as balancing conflicting demands among its parts.² For how a single project works from start to finish, see the guide to what project management is.

Two reference frameworks describe the practice. PMI’s standard is principle-based, written to apply whatever an organization’s project delivery approach.² PMI has a fifth edition in development that builds on the fourth.³ ISO 21504, revised in 2022, gives international guidance on the principles of project and programme portfolio management for public or private organizations of any size, to be adapted to each portfolio.⁴ Because both set out principles to adapt, an organization still has to design its own scoring rules, and the frameworks are most useful as a checklist of what those rules should cover.

What does a portfolio manager do in project management?

A portfolio manager runs the selection and review cycle for a group of projects: keeping the list of work in progress and proposed, preparing the evidence for decisions, and tracking whether approved projects still deserve their resources. PMI describes portfolio managers as overseeing a set of projects, programs and other activities grouped to meet strategic business objectives.²

In practice the portfolio manager usually prepares these decisions, and a group of senior leaders, often called a portfolio board, takes them: approve, delay or stop. PMI’s standard lists both groups among its readers: senior executives and governance boards who make decisions, and managers who make recommendations to them.² In many organizations the work sits in a project management office; Wikipedia’s article on the practice lists project managers and PMOs as the people who analyze and manage projects through it.⁵

The role differs from a steering committee, which directs one project. A steering committee asks whether its project should change scope or continue. A portfolio board asks whether that project should keep its people when another proposal would use them better.

A hypothetical mid-sized insurer shows the difference. Its claims-system project has a steering committee that approves a two-month delay. The portfolio manager then notices that the delay keeps four developers tied up through the quarter in which a regulatory reporting project needs them, and takes that clash to the portfolio board. Nobody on the claims project was wrong; only someone looking across projects could see the clash.

A portfolio manager’s most useful tool is therefore a single, honest list of who is working on what. Without it, every approval looks affordable on its own.

How companies prioritize projects: the portfolio cycle

Companies prioritize projects through a repeating five-step cycle: collect proposals in one format, score them against strategy, balance the mix, approve only what capacity allows, and review the portfolio regularly. The steps cover the four concerns that Robert Cooper, Scott Edgett and Elko Kleinschmidt set out in their 1999 study of US firms: strategic choice, resource allocation, project selection, and a balance between the number of projects and the resources available.⁶

1. 1Collectevery proposal on one form
2. 2Scoreagainst strategy, value and risk
3. 3Balancemix of risk, timing and type
4. 4Approveonly up to the capacity limit
5. 5Review and stoprevisit and free people

Then repeat from “Collect”

1. Collect proposals in one format

A portfolio can only compare projects that are described the same way. A short form asks for the problem, the expected benefit, the rough cost, the people and skills needed, the timing and the main risks. Projects already under way go on the same list, because they compete for the same people as new ideas do.

2. Score them against strategy

Scoring turns each proposal into a few comparable ratings, such as strategic fit, value and confidence of delivery. In Cooper, Edgett and Kleinschmidt’s 1999 study of new product portfolios at 205 US companies, financial methods were the most popular way to choose projects, yet strategic methods and scoring models went with the best portfolios.⁶ Their 2001 report on a large sample of North American firms reached the same conclusion: financial methods, though the most popular and rigorous, gave the worst results overall across six performance measures.⁷

The reason is the quality of early numbers. Killen and colleagues, discussing their 2008 Australian survey, note that financial data is usually not very accurate at the stage when portfolio decisions are made and may be skewed by optimism.¹ A net present value calculated from guesses looks precise without being reliable. These surveys asked managers to rate their own portfolios, so they show which methods go with better-rated portfolios, not that a method causes better results.¹ Keep the financial case, but weigh it alongside strategic fit and a frank rating of how sure the team is.

3. Balance risk, timing and type

A portfolio of only safe, small projects protects this year and starves the next; a portfolio of only bold bets can fail all at once. Balancing means checking the mix: short versus long term, low versus high risk, and spread across markets and technologies. Portfolio maps such as bubble diagrams plot each project on two of these dimensions, and in Killen’s survey of 60 Australian organizations their use was linked to better-balanced portfolios.¹

4. Approve within a capacity limit

The capacity limit is the step most organizations skip. It means counting the people and budget actually free in the next period and approving projects in score order only until they run out. Everything else waits, however good it looks.

Myth

Any project with a strong business case should be approved.

Fact

A strong case only earns a place in the queue. If its people are already committed, approving it tends to slow the projects already running.

5. Review and stop

Approved projects come back for review at set points, often when one of their stages finishes. Each review asks whether the project still fits strategy and still deserves its place in the queue, which is a different question from whether it is on schedule.

A worked example: five projects, one capacity limit

A capacity limit turns a ranked list into real trade-offs, because the second-best project may need more people than are free. The example below is illustrative: a company has 10 people available for new projects over the next six months and five candidates, each scored from 1 to 5 on strategic fit, value and delivery confidence.

The payroll project goes first because the law requires it, leaving 7 people. Following the scores, the portal takes 4 and the dashboard 2, leaving 1 person and no room for the market entry, the project with the highest value. The website redesign is declined.

Leaders could instead approve the market entry and drop the portal: payroll, market entry and dashboard use exactly 10 people. That is the real portfolio decision, a safer project with a good score against a riskier one with a bigger prize, and no formula settles it. The scores frame the choice; the balance check from step 3 and the leaders’ appetite for risk decide it.

If the list needs more people than you have, your team has met the problem the research keeps finding.

Why starting too much work slows everything down

Starting more projects than people can staff tends to make projects finish later without adding to the work done. Manager surveys on two continents rank overload as the weakest point of portfolio practice, and a study of Italian judges links juggling many cases to slower finishes.

The study

Limited evidence

Too many projects for the people available: the weakest portfolio score in Australia and North America

Catherine Killen, Robert Hunt and Elko Kleinschmidt surveyed 60 Australian organizations that manage portfolios of new product projects, using questions from Cooper’s North American surveys. Of six portfolio measures, the one managers rated lowest was having the right number of projects for the people, time and money available, matching the North American results.¹

The survey is small and relies on managers’ own ratings, but finding the same weakest point on two continents suggests it is common. Overload hides well, because each project on its own looks staffed.

The delay follows from arithmetic. Little’s law, proved by John Little in 1961, ties three long-run averages together: how many items sit in a system at once equals the rate at which they arrive times how long each one stays.⁸ If a team finishes work at a steady pace, putting more projects in progress at once lengthens the average time each one takes.

1. One at a time: the team finishes the first project early, so its benefit starts early
2. All at once: the team splits each week across all three, and none of them finishes until late
3. The cost of starting: the total work is the same, but the first finish moves from here to well past the middle

Coviello, Ichino and Persico, three economists, tested this on real workers. They studied judges in Italy who all received the same workload, because a lottery assigned the cases, and found that those who juggled more trials at once, instead of working through a few at a time, took longer to complete their portfolios of cases. The authors conclude that the juggling itself lowered the judges’ productivity.⁹ We read the study’s summary only, and judges are not project teams, but the pattern matches the survey findings.

For a portfolio, the lesson is about sequence. Three projects done one after another deliver the first result early. The same three started together deliver nothing until late, even though the total work is the same.

The hardest portfolio decision is stopping

Stopping a project, or never starting it, is the portfolio decision organizations find hardest and the one that frees the most capacity. Approving feels like progress; stopping feels like admitting a mistake, so weak projects linger and keep their people.

Researchers call the pull to keep going escalation of commitment, first described by Barry Staw in a 1976 paper: a person or group keeps investing in a course of action despite mounting evidence that it is unlikely to succeed. It is related to the sunk-cost effectsunk-cost effect: The pull to keep putting time, money or effort into something because of what has already been spent on it, rather than because of what it will bring from now on.Full entry in the glossary, in which past spending that cannot be recovered becomes the argument for spending more.¹⁰ A project that has used half its budget feels too far along to stop, though the half already spent is gone whichever way the decision goes.

A made-up retailer’s loyalty-app project shows the pattern. It is a year in, has missed two release dates, and the customer research it rested on is now out of date. Ending it frees five people for a store-systems project that the board ranked higher this year. Its sponsor argues for one more quarter “to protect the investment”, which is the sunk-cost argument in its usual form.

Three habits make stopping easier. Agree in advance what would end a project, such as the benefit falling below a set level or the cost forecast rising past a set limit, so the review has criteria to check. Review every project against new proposals, not only against its own plan. And treat a stopped project as a decision that worked: the people it frees are the capacity the next good idea needs.

The bottom line

The most valuable portfolio decision is usually a refusal: a project stopped or never started so that the approved ones have the people they need. Before the next round of approvals, agree what would end each running project, and approve new work only when the people to staff it are free. At every review, ask whether each running project would still win its place as a new proposal.