Project Scheduling: Dependencies, Critical Path Method and Gantt Charts
Once the scope is broken into work packages, the next question is when. A schedule turns a list of work into a time-based plan, showing which tasks come first, which can overlap, and which determine the finish date. The most important technique for this is the Critical Path Method (CPM), and the most familiar way to display the result is the Gantt chart.
The Scheduling Steps
- Define activities: break work packages from the WBS into tasks.
- Sequence activities: determine dependencies between them.
- Estimate durations: see Estimating IT Projects.
- Develop the schedule: calculate start and finish dates, find the critical path and resolve resource conflicts.
- Baseline and control: approve the schedule, then monitor and manage variance.
Dependencies
Four relationship types link activities:
| Type | Meaning | Example |
|---|---|---|
| Finish-to-Start (FS) | B cannot start until A finishes | Testing starts after build is complete |
| Start-to-Start (SS) | B cannot start until A starts | Documentation starts once coding starts |
| Finish-to-Finish (FF) | B cannot finish until A finishes | Testing finishes after final code delivery |
| Start-to-Finish (SF) | B cannot finish until A starts | Old system shuts down after new one starts (rare) |
Finish-to-start is by far the most common. Dependencies are also classed as mandatory (inherent to the work, such as you cannot test code not yet written), discretionary (preferred practice) and external (outside the project, such as a vendor delivery). Add lead (overlap) and lag (a required wait) where needed, for example a three-day lag after a server order for delivery.
The Critical Path Method: Worked Example
The critical path is the longest sequence of dependent activities, and it determines the shortest possible project duration. Any delay to a critical-path activity delays the project. Consider this small project, with durations in days:
| Activity | Description | Duration | Predecessors |
|---|---|---|---|
| A | Requirements | 3 | — |
| B | Build application | 5 | A |
| C | Set up test environment | 2 | A |
| D | System testing | 4 | B, C |
| E | Prepare training material | 3 | C |
| F | Deploy | 2 | D, E |
Step 1: list the paths and add durations.
- A → B → D → F = 3 + 5 + 4 + 2 = 14 days
- A → C → D → F = 3 + 2 + 4 + 2 = 11 days
- A → C → E → F = 3 + 2 + 3 + 2 = 10 days
The longest is A-B-D-F, so the critical path is 14 days.
Step 2: forward and backward pass. The forward pass gives earliest start (ES) and earliest finish (EF); the backward pass gives latest start (LS) and latest finish (LF) without delaying the project.
| Activity | ES | EF | LS | LF | Float |
|---|---|---|---|---|---|
| A | 0 | 3 | 0 | 3 | 0 |
| B | 3 | 8 | 3 | 8 | 0 |
| C | 3 | 5 | 6 | 8 | 3 |
| D | 8 | 12 | 8 | 12 | 0 |
| E | 5 | 8 | 9 | 12 | 4 |
| F | 12 | 14 | 12 | 14 | 0 |
Activities with zero float (A, B, D, F) are critical. Float (or slack) is how long an activity can slip without delaying the project: C can slip three days, E four. That information helps you decide where to apply management attention and where resources can be moved if needed.
Gantt Charts
A Gantt chart displays activities as horizontal bars along a timeline, with dependencies drawn as arrows and milestones as diamonds. Its strengths are clarity and familiarity: sponsors can see at a glance what happens when. Its weakness is that it can hide uncertainty by implying a precision the estimates do not have, and large charts become unreadable. Use summary bars for executives and detailed views for the team. Tools are compared in Project Management Tools Compared.
Milestones
A milestone is a zero-duration marker for a significant point, such as “Requirements signed off” or “UAT complete”. Milestones give stakeholders measurable checkpoints and help communicate progress without drowning in detail. Choose a handful that mean something to the business.
Compressing the Schedule
When the calculated finish is later than required, two classic techniques help:
- Crashing: add resources to critical-path activities to shorten them. It increases cost and works only where extra people genuinely speed up the task.
- Fast tracking: perform activities in parallel that were planned in sequence. It adds rework risk. For example, starting build before design is fully approved.
Only compressing the critical path shortens the project, and once you shorten it, a different path may become critical. Beware also of the tendency that adding people to a late software project can make it later because of onboarding and communication overhead.
Resource Constraints and Buffers
A schedule that ignores who is available is fiction. After sequencing, check resource loading and level conflicts; see Resource Management and Team Building. Critical chain thinking goes further by removing padding from individual tasks and adding a shared project buffer, which protects the date without hiding safety margin in every task.
Schedule Control
- Update progress regularly using remaining effort, not only percent complete.
- Compare against the baseline, and track the trend of the critical path and float.
- Run what-if scenarios before agreeing to a change.
- Use schedule performance index from earned value; see Earned Value Management Explained.
Common Mistakes
- Building a Gantt chart before defining scope and dependencies.
- Assigning 100 per cent availability to people with other responsibilities.
- Padding every task invisibly, then still missing the date.
- Ignoring external dependencies such as vendor deliveries and approvals.
- Never rebaselining after approved change, leaving variance meaningless.
- Scheduling sequences activities, estimates durations and finds the critical path.
- The critical path is the longest dependent chain; zero float means critical.
- Crashing and fast tracking compress schedules, at a cost in money or risk.
- Check resource availability and keep the schedule updated against a baseline.