A project schedule can look finished — all activities listed, dates populated, a reasonable end date that satisfies the owner — and still be fundamentally unreliable as a management tool. The problem is usually not visible without knowing what to look for. The schedule runs. It produces dates. Those dates just don't correspond to how the work will actually unfold.
A schedule health check looks past the surface and evaluates the structural integrity of the network: the logic, the float, the critical path, the data quality. These five signs indicate a schedule that needs attention before it creates serious management problems.
Sign 1: Float Is Unusually High Across Most Activities
Float — the amount of time an activity can slip without delaying the project end date — should vary across a schedule based on work sequence and criticality. A schedule where a large percentage of activities show 30, 60, or 90 days of total float is a schedule with missing logic.
When logic relationships are omitted, activities float freely because nothing connects them to the controlling path of work. The schedule appears to show lots of schedule room. In reality, that room doesn't exist — the activities will constrain each other in practice, but the schedule isn't showing it.
DCMA's 14-point schedule assessment benchmarks total float above 44 working days as a concern. If more than 5% of your activities exceed this threshold, the logic network needs review.
Sign 2: There Are Open Ends in the Network
An "open end" is an activity with no predecessor (other than the project start) or no successor (other than the project end). Every activity in a complete CPM schedule should trace back to the start through a chain of logic and forward to the finish through another chain. Open ends mean the network has disconnected segments.
Acceptable open-end percentages vary by standard, but DCMA benchmarks activities with missing predecessors or successors at less than 5% of total activities. A schedule with 15–20% open ends has gaps that make critical path analysis unreliable.
Sign 3: The Critical Path Doesn't Trace Through Controlling Work
The critical path should connect the major phases of work in the order they actually need to happen. On a construction project, it should run through site preparation, structural work, enclosure, systems installation, and finishing in a recognizable sequence. If the calculated critical path runs through minor administrative tasks or work that clearly doesn't control the project's end date, the schedule logic is misleading you.
This happens most often when network logic is constructed from templates without being adapted to the specific project. The critical path calculation is mathematically correct — it's the longest path through the network — but the longest path isn't meaningfully connected to how the work flows.
Sign 4: Milestones Keep Moving Without Documented Reason
When milestones shift from one update to the next without a documented change reason, the schedule is being used to track what already happened rather than to forecast what will happen next. This is the schedule as a timesheet, not as a planning tool.
A healthy update process captures: what slipped, why it slipped, and what the downstream impact is. If milestone movements are happening silently — dates changing without commentary in the schedule narrative — audit the last four or five updates and look at what moved and by how much. The pattern often reveals systemic update quality issues.
Sign 5: Activities Have Been Stretched Rather Than Updated
One of the most common schedule maintenance problems: an activity was supposed to finish on day 30 but is still showing as in-progress on day 50. Instead of documenting that the activity is running long and adjusting the remaining duration, the remaining duration has been manually extended to align with a desired completion date. The schedule looks like it's tracking, but the data doesn't reflect what's actually happening.
Signs of this pattern include: in-progress activities with oddly round remaining durations (30 days, 60 days) that don't match realistic work remaining, activities that have shown the same percent complete for multiple update cycles, and actual finishes that consistently match revised planned finishes rather than original planned finishes.
What a Health Check Actually Involves
A schedule health check is a structured review of the schedule file against documented quality standards. The review covers:
- Logic completeness — open ends, missing predecessors/successors
- Float distribution — activities with unusually high or negative float
- Critical path validation — does it trace through controlling scope?
- Constraint usage — hard constraints that override logic without documented reasons
- Duration assessment — activities with very long or unrealistically short durations
- Update quality — are actuals being entered accurately or being manufactured?
- Baseline comparison — is baseline vs. actual variance being tracked and explained?
The output of a health check is a prioritized list of issues and recommended corrections, along with an assessment of how reliably the current schedule can be used for decision-making. Some issues are minor and cosmetic; others represent structural problems that need to be corrected before the schedule is used for any serious analysis.
We review schedule files in Primavera P6 and Microsoft Project against documented quality standards and provide a clear report of what needs to be corrected.