Primavera P6 has been the scheduler's tool of choice on major construction projects since the 1990s. It's not elegant software, but it handles complex logic networks, multi-calendar projects and earned-value computation better than anything else at scale. This guide is for planners who are new to P6 or returning to it after a gap — not a manual, but a practical map of what matters.
The three concepts that drive everything
Before opening P6, it helps to understand the three ideas that underpin every calculation it makes.
Setting up a new project: five steps
Set the correct calendar
Every activity duration is calculated against a calendar. Set the project default calendar before adding activities — changing it later recalculates every duration. For GCC projects, configure Friday–Saturday weekends and local public holidays as exceptions.
Build the WBS first
The WBS is the skeleton everything else hangs on. Build it before adding activities. A four-level hierarchy works for most projects: Project → Phase → Discipline → Deliverable. Changing the WBS after activities are assigned is painful.
Add activities with durations, then logic
Don't add logic while creating activities. Build the full activity list first, set durations, then connect the relationships in a second pass. This avoids accidentally creating circular logic and keeps the network readable.
Set a baseline before the project starts
Once logic is checked and approved, save the baseline — Project → Maintain Baselines → Add. This is the frozen reference against which all future progress is measured. Never work directly in the baseline; always work in the current schedule and compare.
Assign resources if you need earned value
Float and critical path work without resources. Earned value — SPI, CPI, forecast finish — requires resource assignments with budgeted costs. If EV is a contract requirement, assign resources before baselining. Adding them after makes the baseline cost unreliable.
Updating progress: the weekly discipline
A P6 schedule is only as useful as the progress data feeding it. The update cycle should be:
- Set the data date to today's date (or the update cutoff).
- Enter actual start and finish dates for activities that have started or completed.
- Update % complete or remaining duration on in-progress activities.
- Run the scheduler (F9) to recompute dates, float and critical path.
- Check for new open ends or constraint violations before saving.
The biggest mistake is skipping the data date update — leaving it at last week's date makes every float calculation wrong by the number of days elapsed.
Field assignees report progress on WhatsApp — OPTEAM reads the reply, updates the activity and syncs back to P6. The data date is always current because the data is always current.
The five-check schedule quality test
Before sharing or relying on any P6 schedule, run these five checks. They take ten minutes and catch 90% of the problems that make a schedule misleading:
- Data date: Is it today? If not, float is already wrong.
- Open ends: Any activity without a predecessor or successor (except start/finish milestones)? Find and fix them.
- Hard constraints: More than 5% of activities? Investigate — constraints mask missing logic.
- Leads (negative lags): Any? Query each one — a few are legitimate, many indicate shortcuts in the logic.
- Critical path sanity: Does the critical path make physical sense? If a finishing activity is critical while the structure below it isn't, the logic is broken.
- Set the calendar and WBS before adding activities — changing them later cascades problems.
- Build the activity list first, logic second — it keeps the network clean.
- Save a baseline before work starts; never modify it after.
- Resources are required for earned value — assign them before baselining if EV is a requirement.
- The data date is the most important field in the schedule — keep it current.
Connect your P6 file to a live control room.
OPTEAM reads your XER file, runs the quality checks and keeps the schedule current with field updates — automatically.