Esc

↑↓ move↵ openIndex · Pagefind
Project management

Out-of-Sequence Progress

Definition

Recorded work that started or progressed before the schedule relationships would normally allow it.

Topic: Scheduling

Updated

Out-of-sequence progress means the recorded work no longer follows part of the planned network logic. It can reflect a changed field sequence, partial access, an inaccurate relationship, or incorrect progress data. The finding identifies a modeling question; it does not explain the cause on its own.

How it happens

The most common form is an activity with an actual start while a finish-to-start predecessor has no actual finish. Others include an activity that finished before its predecessor finished, or one that started before a start-to-start predecessor started.

The scheduling engine finds these at each update, after progress is entered against the new data date. Primavera P6 lists them in its scheduling log. The underlying causes usually fall into four groups:

Cause What it looks like Typical response
Field resequenced the work Crews worked around a delayed predecessor Revise logic with a recorded reason
Activity too coarse One “Level 2 drywall” activity covers rooms released at different times Split by area so logic can follow access
Logic was never right A hard link where the dependency is only partial Correct the relationship type or lag
Data error Progress posted to the wrong activity or date Correct the progress record

How settings treat the remaining work

Once an out-of-sequence activity is confirmed, the engine needs a rule for its remaining duration. Retained logic, the P6 default, makes it wait for the open predecessor. Progress override ignores that relationship and schedules the remaining work from the data date. The same progress can give two different forecast finishes, so the setting must be stated and kept consistent between updates.

Worked example

Illustrative: the schedule links “Complete all overhead rough-in” to “Start ceiling grid” with a finish-to-start relationship. The field team begins ceiling grid in an available room while rough-in continues elsewhere. The actual start conflicts with the modeled relationship even though the field sequence may be practical.

At the October 31 update, rough-in has 8 working days remaining. Ceiling grid has started and has 12 days remaining.

  • Under retained logic, the remaining grid waits for rough-in: it finishes 8 + 12 = 20 working days after the data date.
  • Under progress override, it finishes 12 working days after the data date.

The better fix is to change the model rather than choose between them. Split rough-in and ceiling grid into zones, link each zone’s grid to that zone’s rough-in, and the network describes the sequence actually being built. Then the forecast doesn’t depend on the setting.

Review the activity breakdown and location-specific records. The schedule may need more detailed activities and justified logic changes. Do not alter recorded actual dates merely to make the warning disappear.

  • Out-of-sequence progress is a conflict between actual dates and logic.
  • Invalid dates are actuals after the data date or forecasts before it, a separate check in the DCMA 14-point assessment.
  • Open ends are activities missing a predecessor or successor. They are a logic gap rather than a progress conflict.

Why it matters

A schedule with many out-of-sequence activities has a critical path that depends on a setting rather than on the plan. That weakens forecasts, recovery plans and any time impact analysis that uses the update as its starting point. Counting out-of-sequence activities each update, and tracking how each was resolved, is a simple measure of whether the schedule still matches the job.

What to check each update

  1. Run the schedule and pull the list of out-of-sequence activities from the scheduling log or a filter.
  2. Confirm each against field records, such as daily logs and foreman reports. Fix data errors first.
  3. Decide for each confirmed case: keep the logic, revise it, or split the activity. Record the reason.
  4. Note the progress setting used in the update and keep it the same as last month unless there’s a documented reason to change.
  5. Compare the count with the last update. A rising count usually means the plan and the field are drifting apart.

Common mistakes

  • Editing actual dates so the warning disappears.
  • Carrying the same out-of-sequence activities for several updates without a decision.
  • Deleting relationships instead of correcting or splitting them.
  • Not stating the progress setting in the update narrative.
  • Changing logic without keeping the original update for comparison.

Why scheduling settings matter

Different treatments of out-of-sequence work can produce different forecasts for the same reported progress. Record which setting was used, preserve the original update, and explain any revised logic in the narrative.

See P6 schedule settings and the audit checklist. For the full sequence from data review to a documented decision, start at the Schedule Forensics hub.

FAQ

What is out-of-sequence progress in a schedule?

It is recorded progress that conflicts with the schedule's logic, most often an activity with an actual start while a finish-to-start predecessor has not finished. The work happened in a different order from the one the network describes, so the schedule has to decide how to treat the remaining work.

What causes out-of-sequence progress?

Common causes are a field sequence that changed for good reasons, activities too large to show partial access, relationships that were never right, and progress data entered against the wrong activity or the wrong date. The finding points to a modelling question; it doesn't say which cause applies.

How do you fix out-of-sequence activities?

Confirm the progress data first. If it's right, decide whether the logic is wrong, the activity needs splitting by area, or the dependency still holds for the remaining work. Change relationships only with a recorded reason, keep the original update, and explain the change in the narrative. Don't edit actual dates to clear the warning.

How does P6 handle out-of-sequence progress?

Primavera P6 applies the project's scheduling option. Retained logic, the default, makes the remaining work wait for the open predecessor. Progress override schedules the remaining work from the data date. A third option, actual dates, also exists. The choice can change the forecast finish for the same progress.

Is out-of-sequence progress a problem in a schedule review?

A few instances are normal on a live job. Many, or the same ones carried across several updates, suggest the logic no longer reflects how the work is being built, which weakens the critical path and any delay analysis that relies on it. Reviewers usually count them and ask how each was resolved.

Sources

Keep exploring

Related across os.construction

Open file