Bryan Gilpin
President
Bryan has spent his career building and growing great organizations to deliver technology that improves lives around the world
I have sat through more quarterly portfolio reviews than I can count, on the manufacturer side and now with clients. The format is remarkably consistent. Every program gets a slide. Every slide has multiple red, yellow, and green status indicators. The meeting runs long, everyone leaves with updates and various action items, and nothing on the list is dramatically different from when they walked in.
Then one quarter a program goes from green to red in a single step, and everyone in the room is surprised except the engineers who have been working on it.
If your company runs a disciplined stage-gate process, you may be thinking that your major programs do get stopped and that it happens at a gate review rather than in a portfolio meeting. That is fair. When stage-gate works, it works. But it usually governs only your major development programs. The sustaining work, the line extensions, the projects that never cleared the threshold for formal gate governance, and the programs you inherited through an acquisition all sit outside it. In most companies I have seen, that is the larger share of the portfolio by headcount, and nothing on the calendar forces a decision about any of it.
The harder part is what happens inside the gates you do run. Even at a real gate review, the question in the room is usually “How do we fix this so it can pass?” rather than whether it should continue at all. Fixing is the more comfortable conversation. A gate review will absorb it without complaint, and the program moves forward with a remediation plan instead of a decision.
Last month I wrote about the three months after budget approval, when finishing this year and starting next year collide for the same senior people.
The capacity the 2027 plan assumed usually sits somewhere specific: inside programs that should have been finished, combined, or stopped but were not. Most of those are not the programs on your gate calendar. This piece is about how to decide the rest.
Why reviews stop deciding
Reporting feels safer than deciding. A status update cannot be wrong in the way a stop decision can. Stopping a program the board has already seen feels like admitting a mistake, so it stays on the list at reduced effort, which is the most expensive way to run a program. Isabelle Royer studied this pattern in Harvard Business Review and found that the problem usually starts with a champion’s enthusiasm, which spreads until the organization treats warning signs as temporary setbacks.¹
Color codes can make it worse. Red, yellow, and green without strict trigger definitions reflect how the team feels about the program, filtered through what they think leadership wants to hear. Robert Cooper, who has studied product portfolio management for decades, is direct about this in Research-Technology Management: a status signal “should not be the main go/no go criterion” at a project review or gate meeting. The question that should decide is “Does the project’s value going forward justify the resources that must be invested to complete the project?”² That is a different question from “Is it on track?” and almost nobody asks it.
The last reason is simpler. Most companies have no agreed criteria for when a program is done or should stop. Without them, every stop decision is personal, and personal decisions get deferred.
What it costs
Rose Hollister and Michael Watkins, writing in Harvard Business Review, found that most organizations “struggle to kill initiatives, even those that no longer support their strategy,” and that senior leaders “pile on more and more, expecting teams to absorb it all.”³ In a development organization, every program that should have closed is still holding engineers, test capacity, and budget. Added up across the portfolio, that is often the exact capacity the 2027 plan was built on.
My colleague and VP of Strategic Programs at Suntra, Steve Bowman, who ran portfolio governance at J&J, Takeda, and Fresenius before joining us, puts it this way:
“I have never seen a portfolio review that was short on information. I have seen plenty that were short on decisions. The slide count goes up every quarter, and the number of programs that actually change status goes down.”
Criteria that trigger a decision
The fix is not a better dashboard. It is agreeing in advance what evidence forces a call, so that the review has to produce one. The criteria I have seen work are objective enough that a program manager cannot argue with them and simple enough that a board can follow them.
Continue when milestones are being met on evidence, the major technical risks are retiring on schedule, and the people assigned are actually working on it. All three, not two of three.
Combine when two programs are solving the same platform, regulatory, or supply problem separately, or when one program’s output is another’s input and they are being managed as if they were independent.
Pause when a critical input is missing and has a date: a clinical readout, a supplier qualification, a decision from another part of the business. Pausing is honest. Running at ten percent is not.
Stop when the value going forward no longer justifies what it will take to finish, when the market or regulatory assumption it was approved on has changed, or when the program has missed the same milestone twice with no change in approach. Hollister and Watkins recommend adding a sunset clause to every initiative and requiring each to reapply for funding in annual planning.⁴ In my experience that one mechanism changes the tone of the entire review, because the burden of proof moves from the person proposing a stop to the program itself.
What good governance looks like
A real intake bar: no program gets on the list without an initial business case, a defined outcome, an owner, and named people who are available, not people who will be available once something else finishes.
Health judged on evidence, ideally by someone outside the program team. Milestones met, risks retired, resources in place. Not a color.
Defined exit points. Every program has a definition of done and set points where the team must recommend continue, combine, pause, or stop.
Escalation that works. A risk reaches an executive decision when it appears, not at the next quarterly review.
And every review ends with decisions and owners, not updates.
One more category deserves a mention. After an acquisition, inherited programs tend to run on autopilot because nobody who approved them is still in the room. They should go through the same intake bar as if they were new.
Before January
Hold one portfolio review before the 2027 programs kick off that is only about decisions. What finishes, what stops, what gets combined, and who is freed up as a result. Then three questions for your leadership team:
How many of your active programs would you approve today if they came in new?
When did your portfolio review last stop a program?
Who in your organization is allowed to say a program is done?
If those answers are harder than they should be, we are glad to talk them through at SuntraMedTech.com/contact.
Sources
1 Isabelle Royer, “Why Bad Projects Are So Hard to Kill,” Harvard Business Review, February 2003.
2 Robert G. Cooper and Anita F. Sommer, “Dynamic Portfolio Management for New Product Development,” Research-Technology Management 66, no. 3 (2023): 19–31.
3 Rose Hollister and Michael D. Watkins, “Too Many Projects,” Harvard Business Review, September–October 2018.
4 Hollister and Watkins, “Too Many Projects.”
Bryan has spent his career building and growing great organizations to deliver technology that improves lives around the world