The Critical Path Method: Scheduling Projects with Dependencies
The Critical Path Method (CPM) is a project scheduling technique that identifies the longest sequence of dependent tasks, which determines the shortest possible project duration. In a simple network with fixed durations and no resource conflicts, delaying a critical activity delays the calculated finish unless another part of the plan changes.
How CPM Works
CPM involves four steps: list all tasks, identify dependencies, estimate durations, and calculate the critical path.
Step 1: List All Activities
Break the project into individual activities with defined start and end points. Each activity should be small enough to estimate accurately at a level that the team can estimate and verify. For a well-scoped project, this list comes directly from the work breakdown structure.
Step 2: Identify Dependencies
For each activity, determine which other activities must be completed before it can start. Dependencies fall into four types:
- Finish-to-Start (FS): Task B cannot start until Task A finishes. This is the most common dependency type.
- Start-to-Start (SS): Task B cannot start until Task A starts. Both tasks can run in parallel after that point.
- Finish-to-Finish (FF): Task B cannot finish until Task A finishes.
- Start-to-Finish (SF): Task B cannot finish until Task A starts. This is rare in practice.
Step 3: Estimate Durations
Assign a duration estimate to each activity. Use historical data, expert judgment, or estimation techniques. For uncertain estimates, the Program Evaluation and Review Technique (PERT) uses three estimates — optimistic, most likely, and pessimistic — to calculate an expected duration.
Step 4: Calculate the Critical Path
Build a network diagram showing all activities and their dependencies. Perform a forward pass to calculate the earliest start and finish times for each activity. Then perform a backward pass to calculate the latest start and finish times. The difference between earliest and latest times is the float (or slack) for each activity.
In the unconstrained example below, the critical path has zero total float. Imposed deadlines, calendars, and other constraints can change the float reported by scheduling software.
Reading the Network Diagram
Each activity node in a CPM network contains:
| Field | Description |
|---|---|
| ES (Early Start) | Earliest the activity can begin |
| EF (Early Finish) | ES + Duration |
| LS (Late Start) | Latest the activity can begin without delaying the project |
| LF (Late Finish) | Latest the activity can end without delaying the project |
| Float | LS - ES (or LF - EF) |
Activities with zero float are on the critical path. Activities with positive float have scheduling flexibility — they can be delayed by up to their float without affecting the project end date.
Practical Applications
Resource Leveling
CPM reveals which activities have float, allowing project managers to shift non-critical activities to balance resource utilization. If two activities need the same resource but one has float, the one with float can be rescheduled.
Schedule Compression
When the calculated project duration exceeds the deadline, CPM helps identify where to apply schedule compression techniques:
Crashing. Add resources to critical path activities to reduce their duration. It can increase cost and shorten the schedule when the work can use the additional capacity effectively.
Fast-tracking. Run critical path activities in parallel instead of sequentially. Fast-tracking increases risk because dependent activities overlap.
Both techniques focus on the critical path because compressing non-critical activities does not shorten the project.
Risk Identification
Critical and near-critical activities deserve close schedule attention, alongside risks with large potential impacts elsewhere in the network. Risk management strategies should prioritize these activities. A delay in any critical path activity is a project delay.
CPM in Modern Project Management
CPM is useful when activities, durations, and dependencies can be modeled explicitly. Modern tools like Microsoft Project, Primavera, and Smartsheet automate CPM calculations and display critical paths on Gantt charts.
In agile environments, CPM is less commonly used at the task level because sprints operate on a fixed cadence rather than a dependency-driven schedule. However, CPM remains valuable for release planning, multi-team coordination, and projects with hardware or infrastructure dependencies.
CPM for Release Planning
Even agile teams face dependencies at the release level. A mobile app release might depend on API changes, database migrations, and app store review processes. Mapping these dependencies with CPM helps identify the critical path to release and highlights which work streams need to complete first.
Common Mistakes
Forgetting to update the schedule. CPM is not a one-time exercise. As activities complete (faster or slower than estimated) and new dependencies emerge, the critical path may shift. Regular updates keep the schedule accurate.
Ignoring near-critical paths. A path with only one day of float is nearly as risky as the critical path. Monitor near-critical paths because small delays can push them onto the critical path.
Over-relying on software. Tools calculate the critical path automatically, but the output is only as good as the input. Inaccurate duration estimates and missed dependencies produce misleading critical paths. Validate the tool’s output with the team.
Not communicating the critical path. Everyone on the project should know which activities are on the critical path and understand the implications of delays. Make the critical path visible in project dashboards and status reports.
Worked network: two branches and a final review
Use working days and start counting at day 0. Activity A takes 2 days. After A, B takes 5 days and C takes 3 days; B and C can run in parallel with different people. Activity D takes 2 days and requires both B and C to finish.
| Activity | Predecessor | Duration | Early start | Early finish | Late start | Late finish | Total float |
|---|---|---|---|---|---|---|---|
| A | None | 2 | 0 | 2 | 0 | 2 | 0 |
| B | A | 5 | 2 | 7 | 2 | 7 | 0 |
| C | A | 3 | 2 | 5 | 4 | 7 | 2 |
| D | B and C | 2 | 7 | 9 | 7 | 9 | 0 |
The A-B-D path takes 9 working days; A-C-D takes 7. D starts at the later predecessor finish, day 7. Working backward from day 9 gives D a latest start of day 7. B must therefore start by day 2, while C can start as late as day 4. C has 2 days of total float.
If C grows from 3 days to 6, its path becomes 10 days and controls the finish. If B and C need the same person and cannot overlap, the original parallel assumption fails: after A, their combined 8 days of work must be scheduled before D. The finish becomes day 12 under that sequential arrangement. Resource availability belongs in the model before a date is promised.
Check the model at every update
Record actual starts and finishes, then estimate the remaining work from the current position. Review open-ended activities with no successor, unexplained date constraints, and relationships added only to force a desired finish. Compare the new controlling path with the previous one and explain any change in the status report.