Gantt charts are not project managers.

The most polished CRO project plans I have ever seen — color-coded Gantt charts with critical-path dependencies, RACI matrices nested three levels deep, milestone checkpoints with auto-generated burndown — were on studies that ran late. The least polished project plans I have seen — sometimes a shared spreadsheet, sometimes a Monday morning standup — were on studies that hit dates. The correlation is not perfect, but it is strong enough to be uncomfortable, and it points at something the CRO industry has gotten backwards about what project management actually is.

A Gantt chart is a representation of the work. It is not the work. The work, on a clinical trial, is finding the missing dry ice that should have arrived at the Phoenix site yesterday and didn't. The work is calling the central lab at 7 a.m. to confirm the courier window before the site coordinator gets in. The work is talking the PI through why the protocol amendment doesn't actually require him to redo all his already-enrolled subjects' consents. The work is the sponsor's clinical operations lead waking up at 6 a.m. her time on a Tuesday to confirm a same-day kit substitution. None of this work appears on a Gantt chart, because none of it is on the critical path until the moment it is, and by then the critical path has shifted and a new chart needs to be drawn.

This is the practical observation that has shaped how we hire and deploy project managers. The skills that produce a clean Gantt chart — methodical decomposition, clear visual layout, structured dependency mapping — are real skills, and they are useful skills, but they are not project management. They are project documentation. The two are routinely conflated. A CRO that hires for project management on the basis of who can produce the cleanest plan is hiring a documenter, and documenters do not find dry ice at 6 a.m.

What we hire for instead is operational instinct under pressure. The PMs who hold our studies together, on the engagements I have lived through, share three traits. They follow up without being asked. They escalate fast and de-escalate faster. They are willing to do the unglamorous work themselves rather than hand it off — calling the courier, calling the lab, calling the site coordinator at home if it has to be home — because they know the work has to happen and they are not above it. The Gantt chart, where one exists, is a tool they use; it is not a tool they are.

The deeper problem with Gantt-chart project management is not that the chart is unhelpful. It is that the chart confers a feeling of control. A sponsor who reviews a clean Gantt chart at the kickoff feels, reasonably, that the study is in hand. The PM who produced it feels the same way. Both feelings are mostly false. The chart represents the planned trajectory of the study. The study's actual trajectory will diverge from the plan within the first month, and the question is whether the PM is the kind of person who notices the divergence and acts on it before the chart needs to be redrawn.

The PMs who notice are the ones who are talking to sites every week. They have the coordinator's cell phone number. They know which freezer at which site has been on a slow leak since November and is therefore the next operational risk on the program. They have a relationship with the central lab's receiving manager that survives a courier mix-up. None of this knowledge lives in the chart. It lives in the PM, and it has to be cultivated by the PM doing the unglamorous follow-up work that the chart cannot do for them.

What this means, practically, for sponsors evaluating CROs: the project plan is a poor signal of project management capability. It signals that someone on the CRO's team has invested an afternoon in formatting a Gantt chart, which most CRO teams have. The signal that matters is harder to see. It is whether the PM you will be working with has the operational instinct to call you before something becomes a problem, the willingness to do the unglamorous work, and the relationship density at the sites your study will activate.

Asking the right questions in the kickoff is more useful than reviewing the project plan. Who is the PM, and how many studies have they personally project-managed to completion? When was the last time they called a site at 6 a.m.? When was the last time they spent an afternoon coordinating a courier reroute? If the answers are abstract, the project management is abstract. The Gantt chart will be impressive. The study will run late.

None of this is an argument against planning. We use Gantt charts. We use RACI matrices when they earn their place. The argument is about what those tools are for. They are aids to memory and aids to communication. They are not substitutes for the human who is doing the actual project management work — and a CRO that mistakes the artifact for the practice will produce beautiful documents on top of stalled studies.