Major projects rarely move from healthy to failing overnight.
More often, deterioration is gradual. A dependency becomes uncertain. A decision is deferred. A supplier misses an intermediate commitment. A milestone becomes increasingly difficult to defend. Teams begin discussing workarounds. Assumptions that once seemed reasonable become less certain.
Individually, none of these developments necessarily constitutes a crisis. Collectively, and over time, they may tell a very different story.
The challenge for senior management is not simply whether this information exists. In most organizations, it does.
The more difficult question is:
Does evidence that a delivery is deteriorating reach the people who can intervene early enough to make a difference?
Research and industry experience suggest this remains a persistent problem.
Projects usually deteriorate before they fail
Research published by the Project Management Institute (PMI) on early warning signs in complex projects makes an important distinction between conventional project performance measures and early indicators of trouble.
Traditional performance measures tend to be lagging indicators. They describe consequences that have already become measurable. Early-warning approaches instead look for leading indicators — evidence that may precede a visible deterioration in formal project performance.
This distinction matters.
A schedule variance tells management that something has slipped. A cost variance confirms that additional money has already been consumed. A missed milestone demonstrates that a commitment has not been met.
These are valuable measures, but they may become visible only after the underlying conditions have been developing for some time.
PMI's research found that project professionals are often not particularly effective at identifying early warning signs and are even less effective at acting on them. As projects become more complex, formal project assessments can also become less effective at detecting emerging problems.
The implication is significant: the first evidence of delivery trouble may not appear in the formal project indicators that senior management normally reviews.
One of the most interesting findings in PMI's research is the importance of what might be called soft evidence.
Researchers found potential warning signs in areas such as stakeholder alignment, trust, unrealistic expectations, organizational behavior, leadership, competence and differences between quantitative risk assessments and people's qualitative assessment of the situation.
The methods used to uncover warning signs are equally revealing.
They included interviews, group discussions, peer reviews and observations in meetings.
This suggests an important limitation in how organizations traditionally think about project information.
A risk register records a recognized risk.
A project schedule records an agreed plan.
A status report records an assessment at a particular point in time.
But the development of a problem may begin much earlier in ordinary working conversations:
"We're still waiting for approval."
"The supplier hasn't confirmed that date."
"We may need to revisit the scope."
"That decision has been pushed to next week."
"We can probably still make the milestone, but it is getting tight."None of these statements necessarily deserves escalation on its own.
But imagine similar concerns appearing repeatedly across six or eight meetings over two months.
At some point, isolated observations become a pattern.
The management challenge is recognizing that transition.
The problem becomes more difficult as information moves upward through an organization.
A project manager can participate in many of the conversations surrounding a delivery. A program director may oversee several projects. A portfolio executive may be responsible for dozens.
The higher the management level, the less practical it becomes to maintain direct exposure to the everyday conversations in which emerging delivery problems first become visible.
PMI addressed this problem directly in Raise the Red Flags. It observed that CEOs, CIOs and other executives rarely have a firm grasp of all the day-to-day developments affecting projects across their portfolios. More importantly, it warned:
“When executives finally learn that a project is about to go off the rails, it can be too late to avert a crisis.”
The article also provides an instructive example.
One project continued to report all its formal status indicators as green despite a troubled team, missed deadlines, supplier problems and continuing scope changes.The lesson is not that formal reporting is unnecessary.
It is that formal status and underlying delivery reality are not always the same thing.
PMI therefore argues that executives need to look beyond conventional metrics and engage with the broader context of how projects are actually progressing.
This problem is not confined to traditional project-management research.
In a 2026 examination of major infrastructure delivery, McKinsey described project underperformance as typically cumulative.
Delays build as schedule risks are not identified or escalated early, interfaces are poorly managed, and cost and schedule information becomes too fragmented to support early intervention. McKinsey observes that organizations may have project controls in place without those controls operating as an effective management system.
Perhaps the most important observation is this:
“Teams may have data but lack a single, trusted view of what is moving,
what is slipping, and where leadership attention is needed next.”
That is a different problem from simply lacking data.
Modern organizations generate enormous quantities of delivery information: schedules, issue registers, project plans, emails, meeting transcripts, actions, decisions, status reports, financial information and supplier updates.
The problem increasingly becomes one of connecting the evidence.
A supplier concern may be discussed in one meeting. Its schedule consequence may emerge in another. A workaround may subsequently be proposed by a technical team. Two weeks later, a program meeting may still describe the milestone as achievable.
Each piece of information can be reasonable in isolation.
The risk becomes apparent when they are viewed together and over time.
This leads to a useful distinction between reporting and warning.
Reporting asks:
What is the status of the project?
Early warning asks:
What evidence is emerging that today's reported status may no longer be sustainable?
The difference is subtle but important.
A warning system should not attempt to predict every future project failure. Complex projects contain uncertainty, and false alarms can be almost as damaging as missed ones.
Instead, its purpose should be to identify meaningful changes in the evidence surrounding delivery.
A concern mentioned once may be noise.
A concern repeated across several meetings is more significant.
A concern repeated by different stakeholders is stronger again.
A dependency repeatedly described as uncertain while the corresponding milestone remains unchanged deserves attention.
And a growing collection of such signals across schedule, suppliers, resources, scope and decisions may warrant management attention even while the formal project status remains green.
The objective is therefore not necessarily to declare that a project will fail.
It is to recognize when confidence in the current delivery commitment should reasonably be questioned.
McKinsey's discussion of major project delivery points toward the organizational value of earlier visibility.
It describes a European transmission operator that combined information from more than 50 data sources into a digital control tower. The resulting capability provided a trusted view of performance and enabled materially earlier intervention on emerging bottlenecks.
The underlying principle applies well beyond infrastructure.
Senior management does not need every detail generated by every project.
Nor does it need another stream of meeting summaries.
What leadership needs is sufficient warning that something important is changing.
A useful early-warning mechanism therefore performs a form of organizational compression:
Thousands of individual observations → persistent signals → changing patterns → a small number of issues requiring management attention.
That is what makes the familiar red, amber and green metaphor useful.
The red light does not repair the problem.
It does not replace the project manager.
It does not make the delivery decision.
Its value is simpler:
it turns on while there is still time to act.
Project-management disciplines have spent decades improving schedules, risk registers, governance, portfolio reporting and project controls.
Those mechanisms remain essential.
Yet the research points to a persistent gap.
PMI's work suggests that important early warning signs can be informal, behavioral and difficult to capture through conventional assessments. Its executive-level analysis warns that by the time serious problems become visible to senior management, the opportunity for effective intervention may already have narrowed.
More recently, McKinsey has described the same underlying problem in contemporary delivery environments: fragmented information, insufficient early escalation and a lack of a trusted view showing leadership where attention is required.
Technology has meanwhile made organizations better than ever at capturing information.
Project systems capture structured delivery data. Collaboration platforms capture discussions and decisions. Meetings can be transcribed. Actions can be extracted. Large portfolios can produce enormous amounts of digital evidence every week.
This creates a new question for project governance.
If the warning signs already exist somewhere within that evidence, how much earlier could leadership know?
That may be one of the more important delivery-management questions to answer.
Project Management Institute, Identifying and acting on early warning signs in complex projects, 2011.
Read the PMI research
Project Management Institute, Raise the Red Flags, PM Network, 2011.
Read Raise the Red Flags
McKinsey & Company, Australia's grid opportunity: Why it's time to flick the switch, 2026.
Read the McKinsey article