The dashboard is green. The program isn’t.

By the time a program is visibly in trouble, the status report has usually been green for months. That is rarely a reporting failure. It is an accurate reflection of what people have learned is safe to say.

I have walked into programs in county and state government, inside a Fortune 500 enterprise, and across a hospital system. Different sectors, different constraints, entirely different politics. The first finding is almost always the same.

The reporting was not dishonest. It was rational. Every person filling in a status field was answering a question they had learned the hard way: not where is this program, but what can I say about this program without it costing me something.

Those are not the same question, and an organization that cannot tell them apart will keep being surprised by failures it was told about, in writing, months in advance.

Red is treated as a verdict, so red disappears

Ask most delivery leaders what happens when they report a project red, and the honest answer is: meetings. Escalation. A request for a recovery plan by Friday. Sometimes a quiet conversation about whether they are the right person for the role.

What almost never happens is the thing that would actually help. Nobody arrives with another two developers. Nobody moves the dependency that is blocking the work. Nobody removes a scope item.

So the person running the program does the arithmetic. Reporting red generates scrutiny and no resources. Reporting amber buys another two weeks to fix it themselves, on their own time. They report amber. They work the weekend. And they are not being evasive. They are responding correctly to the incentives the organization built.

The consequence compounds. Small problems stay hidden while they are still small and cheap. They surface as large problems, which are expensive, which confirms leadership’s belief that red is catastrophic, which raises the cost of reporting it. The cycle tightens every quarter.

Red is a request, not a confession

The single most useful change I have made to a struggling portfolio costs nothing and takes about six weeks to hold. It is a change in what the colour means.

Red is not an admission that someone failed. Red means: this program needs something it does not currently have. The correct executive response to red is not “how did this happen.” It is “what does it need, and can we provide it.”

That sounds like a soft change. It is not. It requires leadership to actually answer the question, which means having something to give: budget, people, scope relief, a decision that has been sitting unmade for a month. If red is raised and nothing is offered, everyone in the room learns that the new policy is the old policy wearing better language, and the reporting goes quiet again within a cycle.

Where it holds, the behaviour changes fast. Red starts arriving early, when it is small. Early red is a conversation. Late red is a recovery engagement.

What this looks like in government

In county and state programs the pressure runs in the same direction, harder. Status can become part of the public record. A red project can become a news item, a council question, an issue in somebody’s election. Procurement cycles are long and rigid, so the flexibility to respond to a problem is genuinely constrained. Timelines are frequently set by a legislative or fiscal calendar rather than by the work.

None of that is an argument for optimistic reporting. It is an argument for separating two conversations that most public-sector programs run together: the program health conversation and the public accountability conversation.

Delivery teams need a channel where an honest assessment is not immediately a political event. That is not secrecy, and it does not survive contact with a genuine records request, nor should it. It is sequencing. Leadership needs to hear a problem, decide on a response, and communicate both together. When the only available channel is the public one, the incentive is to say nothing until you have to, and by then the options are gone.

The programs that go well in this environment are the ones where an executive sponsor has made it survivable to bring bad news early, and has demonstrated it at least once in a way people could see.

“Underachieving employees” is usually the wrong diagnosis

I was brought into a large enterprise where very few projects were landing on time. Resources were limited. The stated problem, offered to me on the first day, was that the staff were underperforming.

What I found was an absence of structure. Ownership was ambiguous. Decision rights were undefined, so decisions escalated by default and waited. There was no reliable cadence, so problems surfaced whenever someone happened to raise them. In that environment, work does not get distributed by design. It flows to whoever is most reliable.

Which is why the same four or five names were on every critical path. The strongest people had absorbed the gap the structure left, and they were exhausted. Meanwhile a significant number of people had been quietly written off as low performers when they were simply mis-assigned, under-briefed, or waiting on a decision nobody had made.

Productivity improved when we put structure in first: clear ownership, a real cadence, decision rights written down and honoured. Ambiguity is the most expensive thing in a delivery organization, and it costs the most from the people who care the most. Then we rebuilt relationships across the functional boundaries, because most of the delay was sitting in handoffs where nobody had a working relationship with the person on the other side.

We did not replace the team. We stopped making it impossible for them to do their jobs.

Knowing what your people can actually carry

Taking over a hospital PMO, I evaluated every individual. Not to rank them, and not to build a case for cuts. To find out what each person could genuinely deliver, and under what conditions.

Capacity is not headcount. A title tells you what somebody is called, not what they can carry. Two people with identical job descriptions and comparable experience will deliver very differently depending on the kind of program, the amount of ambiguity, the political weight, and whether they are working with people they trust.

Some people are exceptional at stabilizing chaos and lose interest once a program is running cleanly. Others are wasted on a rescue and superb at optimizing something already stable. Putting the second kind on a failing program is not a performance problem when it goes badly. It is an assignment problem, and the person who made the assignment owns it.

So we matched individuals to the challenges that fit them, and we made collaboration the default rather than a personal virtue. When project managers stopped having to defend their status in front of their peers, they started asking each other for help. The reporting improved as a side effect of people no longer needing to protect themselves inside it.

Accurate reporting is the mechanism, not the paperwork

Governance gets dismissed as overhead, and a lot of it earns that. But accurate status is not administration. It is the only instrument that makes delivery capacity visible.

Capacity you cannot see, you cannot deploy. If four programs are quietly amber and one is honestly red, leadership can move a specialist, delay a non-critical scope item, or bring in outside capability for eight weeks. If all five report green, there is nothing to act on, and the organization will keep discovering its capacity limits by exceeding them.

This is the shift that has to happen before delivery improves in any durable way. Not a new tool. Not a better template. A mindset in which the colour on the report describes the program rather than the reporter, and in which the organization’s stated purpose is getting programs out of red rather than never seeing red at all.

A green dashboard is not good news. An accurate one is.