Consider a report on a plant upgrade: 62% complete, tracking to the approved date, with no changes. The schedule says something else: the driving path has lost most of its float, two milestones are held by entered dates, and the controlling sequence has changed.
Flyvbjerg and Gardner, in How Big Things Get Done, report 91.5% of 16,000 projects running over budget, over schedule, or both. Flyvbjerg’s database puts the share delivering on cost, time, and benefits at 0.5%. The schedule carries the evidence of slippage.
Percentage Complete Hides the Sequence
Percentage complete is an input. It says little about sequence. A furnace reline can be 62% complete by installed tonnage while still waiting on the tie-in that controls everything downstream.
Float and logic are calculated from the network and move whether reported or not. That makes them useful risk indicators.
Four Signals That Appear Before the Date Moves
Float erosion on the driving path. Total float is the days an activity can slip before delaying completion. Read the trend, not just the value. For example, a driving path with 15 days of float in March and two in June has spent 13 days of protection, even though the completion date held all quarter. That project is one bad week from a public date change.
Hard constraints. A constraint is a date entered into the schedule that overrides the logic around it. If a milestone holds its date while predecessor activities slip, a constraint may be doing the work. The report date then reflects an instruction rather than the calculation.
Missing logic. Activities without predecessors or successors sit outside the calculation. They cannot drive completion or demonstrate delay. These gaps often appear around commissioning and vendor-supply work.
Critical path churn. The driving sequence should change as work changes. If it changes every update without explanation, the schedule may be getting rebuilt instead of updated, making the original delay harder to trace.
The Defence Contract Management Agency’s 14-Point Assessment provides thresholds:
| Check | DCMA threshold | What a breach conceals |
| Missing logic | 5% of activities | Work that cannot show delay |
| Hard constraints | 5% of activities | Dates held by instruction |
| Negative float | 0% | An unachievable completion date |
| Finish-to-start relations | 90% or more | Overlaps that hide the true sequence |
Two Ratios Worth Tracking Every Month
The Baseline Execution Index (BEI) compares activities completed with those the baseline said would be complete. Below 1.0, the project is retiring work more slowly than planned, and the gap grows.
The Critical Path Length Index (CPLI) compares remaining critical path duration with the time remaining to the target date. DCMA looks for 0.95 or better. Below that figure, the remaining plan does not fit the time left, so each update must absorb the difference somewhere.
Both ratios are calculated from the file, not reported by the contractor. They can move earlier than the completion date, providing early warning. On an 18-month plant upgrade, a BEI drifting from 1.02 to 0.88 over two quarters is a clearer warning than a reassuring narrative.
Ask for the Native File, Not the PDF
A PDF shows dates. The native file shows how dates were produced: logic, constraints, float, calendars, and changes. A PDF review answers a question nobody is arguing about.
This matters most when a project reaches a claim. AACE International Recommended Practice 29R-03 sets out the integrity preconditions a schedule must meet before any delay analysis based on it carries weight, and how forensic schedule analysis works depends on source files from each period, not printed summaries. Projects that keep only PDFs discover the problem 18 months later, when the records they need no longer exist in usable form.
What the Standards Expect
The Society of Construction Law Delay and Disruption Protocol treats the updated programme as the primary tool for managing project time (§4.7). Paragraph 8.1 addresses float ownership, important when parties argue over who consumed the 13 days. The GAO Schedule Assessment Guide expects near-term work to be planned in more detail than work two years out, giving reviewers a practical test.
What to Do Before the Next Review
A schedule that reports on track while its float, constraints, and logic say otherwise is still telling you the truth. It’s telling you in the part of the file most monthly reviews never open. Ask for the native file, read the float trend on the driving path, and check the four structural measures before accepting the headline date. The evidence is already there.
Article received via email
