PMOs spend a lot of time focused on providing “transparent portfolio visibility.” But what’s most important to make visible: what’s already happened, or what’s likely to happen next?
AI Is About to Expose Status Reports for What They Are
When an executive makes a portfolio decision based on AI analysis of status reports, and that decision turns out to be wrong…who’s accountable?
The AI? The exec who trusted it? The PM who filed a sanitized report because the culture punished honesty? The PMO that designed a template with no room for real information?
I posted about this last week, concluding with this line:
“AI might finally force us to admit our embarrassing secret: a staggeringly low number of status reports say anything.”
Political sanitization is the easy villain here, and it does deserve a big part of the blame. But there’s a more basic problem sitting underneath that one. It’s cheap to fix and yet chronically overlooked:
We’ve lost sight of what status reports are supposed to be.
At their best, status reports help leaders make better decisions about what’s coming. Influence is the point. A good status report gives someone with authority a reason to do something different, early enough that they still have options.
Somewhere along the way, status reports became “1-pagers” that describe what happened since the last report and whether the project is on track to hit its next milestone or meet its budget constraints (as if that’s a black-and-white answer).
To be clear, I’m not against 1-pagers. But I do wonder if reports that focus 90-100% on what’s already happened are a good use of the 2-5 hours project managers spend building them…and the 10 minutes executives spend reading them.
Why 90% of Status Reports Fail to Say Anything
When almost all of a status report focuses on what’s already happened, usefulness tanks. That’s because reporting on the past is only useful if:
- It helps us predict what is likely to happen next.
- It helps us prepare for what will happen next.
- It helps us influence (or prevent) what will happen next.
With a few exceptions, anything that doesn’t advance those goals is fluff, or worse, CYA behavior. But prediction, preparation, and prevention are often last on a PM’s mind.
Too busy dealing with issues to deal with risk. Too busy dealing with reporting to deal with prediction. And if we’re honest, too busy complaining about what we think is in our way to actually prove it and prevent it.
Most of the metrics PMOs track today are lagging indicators.
They tell you where you’ve been. What they should be doing is pointing you toward where you’re headed, and giving you a chance to change course before it’s too late.
And a truly superior status report would frame everything in these terms, weaving real evidence into its conclusions, warnings, and requests.
But startlingly few PMs and PMOs have the data they need to do this. And when execs talk about “the promise of AI,” what they often really mean is: “Finally, I’ll get some real intelligence.”
We Think We Know Why Projects Fail. So Why Aren’t We Watching for It?
PMI’s 2025 report on maximizing project success, like many before it, documents the top reasons projects fail—and the list has barely shifted in decades.
The usual suspects: shifting priorities, unclear requirements, disengaged sponsors, scope creep, capacity problems.
What’s always struck me about these studies is that they’re mostly based on polls. They ask PMs and PMO leaders what they think caused failure.
And those instincts are probably right more often than they’re wrong. But how many PMOs are testing those instincts with data? How many can go to their executive sponsors and say, “Here’s evidence that decision latency is our biggest performance risk right now, and here’s what we recommend doing about it”?
That’s the skill set that will separate PMOs going forward: the ability to form a hypothesis about what’s driving poor performance, design a way to test it, and present evidence that changes behavior.
In that world, status reports move from perfunctory filling out of a form to an intelligence brief about what should be done to protect the project’s odds of success.
What Leading Indicators Should a Project Manager (or PMO) Measure?
Every PMO I work with has complaints. Recurring frustrations that surface in retrospectives, in hallway conversations, in the “any other business?” slot of meetings nobody wants to attend, and in one-on-ones where the real talk starts.
The best PMO leaders recognize that those complaints are hypotheses in disguise. And hypotheses can be tested. Even better: once you know what’s causing a problem (and how to see it coming), you can intervene early, before the cost compounds.
So if your PMO believes something is driving poor performance, the question is:
How would you prove it? And what would you actually do differently if you did?
Here are six leading indicators to start with—drawn from the complaints I hear most often when I ask clients about their troubled projects. Remember, a good leading indicator should help you predict and prepare for what is likely to happen next. What is a good predictor for one industry or project may not be for a different industry or project.
Test whether these work for you.
Decision Latency
When a project slips, executives ask the PM what happened. A lot of the time, the slip came from their own decision speed. And nobody was tracking it.
Decision latency (the time between when a key decision is needed and when it actually gets made) is one of the clearest early signals of whether your schedule will hold. Track it across the portfolio (average cycle time, how often a late decision moved a baseline), and you can see slip coming before the deadline moves.
It also changes the conversation. A sponsor staring at a 14-day average decision cycle on their own project can’t easily blame the PM for the timeline. (Well, they can. But the data makes it awkward.)
Make decision latency visible, then manage it. The schedule gains are often enormous.
Team Happiness
Flash polls on how the team is actually doing. Many agile teams run them every sprint. And they get dismissed as fluffy, even though they’re one of the earliest signals you have that a project is headed for trouble.
Ask about workload, trust, effectiveness, and psychological safety. Waterfall teams can borrow the practice of rotating polls every two or three weeks. The trend line matters more than any single score. A team whose happiness numbers are sliding is a team about to produce worse work, whether or not anyone has said anything out loud yet.
Planned vs. Unplanned Work
Every PM I’ve worked with believes they’re spending too much time on unplanned work. They’re usually right. They almost never have the data to prove it.
Planned vs. unplanned is simple to track: what portion of your team’s working days went to the current milestone or planned tasks, and what portion went to assessing the effort of change requests, issue resolution, inter-project pulls, and fire drills? Watch that ratio over time, and you can see scope pressure eating execution capacity well before the baseline slips.
It also reframes the conversation with sponsors. You stop saying “we have too much to do” and start saying “62% of our working time this quarter went to unplanned work, here’s the pattern, here’s what I recommend.” Much harder to wave off.
Issue Resolution Speed
How long does it take your team to resolve a high-priority issue once someone surfaces it? Most teams don’t know. They have the data (it’s sitting in the issue log), but they’ve never turned it into a trend line.
Issue resolution speed is a leading indicator twice over. It tells you how resilient your project will be when something unexpected lands (a slow-resolving team is a brittle team). And when you segment by issue type, recurring categories, and chronically slow resolutions point straight at the systems most likely to cause your next problem.
If your project is about to hit a stretch of uncertainty (say, a go-live or a vendor transition), this is the number you want to know before you get there.
User Adoption Readiness
Most project teams obsess over when the deliverable will be ready. Far fewer track whether the users will be. That gap is where a “successful” project becomes a failed rollout.
User adoption readiness is measurable long before go-live. Stakeholder awareness, training progress, preparedness, and confidence in the new way of working, sampled over time, tell you whether your final phase is going to be a smooth ramp or a scramble of rework and budget overrun. The earlier you see a downward trend, the more options you have to correct it.
Don’t assume what “ready” means to your users. Ask them, and watch the curve.
Project Attributes
This one’s for PMO directors specifically. At the meta level, you should be measuring project attributes against project performance. Dedicated BA versus no BA. Phased delivery versus big-bang implementation. Whatever the variables are in your portfolio, the ones that actually correlate with better outcomes are probably not the ones you’d guess. And they’re almost certainly not being tracked.
I did this work for a Top 10 pharma company in the late 2010s, and I’ve never forgotten it. We added attribute metadata to the portfolio and almost immediately saw that facilities projects directly led by facilities had schedule and budget overruns at less than half the rate of projects where facilities were only a consultative stakeholder. Only 25% of projects followed that model.
That PMO went from “we should be included more often” to “we should be running these projects, and here’s the evidence for why.” That’s what the shift attribute data makes possible. You can argue for the operating model that produces the results your organization wants.
If You Only Do One Thing
Pick one complaint your PMO or your PMs keep raising. “Sponsors are too slow.” “We don’t have enough capacity.” “Users aren’t adopting what we deliver.” Whatever it is, treat it as a hypothesis.
Then ask:
- How would we prove this is true?
- What would we measure?
- And if the data confirms it, how could we use it to predict and influence future performance?
If you can answer those questions, you’ve found your first leading indicator. And you’ve started the shift from reporting what happened to influencing what happens next.
What leading indicators are you tracking (or wish you were)? I’d love to hear what’s working and where PMOs are experimenting → send me a DM on LinkedIn or drop a comment here.