2.4PERT network & critical path
The essentials in 30 seconds
- Critical path = the longest sequence: it sets the total duration — a delay on it delays everything.
- The PERT network models tasks and dependencies as a graph and computes early/late dates and floats.
- Reflex: vigilance on critical tasks (zero float), slack on others — and to shorten, act on the critical path.
Running mini-case — Bilal learns where to look first
Bilal worried about the site mockup, three days late — until his PERT network showed it has a week of float, unlike the tool development, with zero float. He shifts vigilance: daily tracking on the tool (critical), a simple weekly check on the mockup. When the sponsor asks to save time, he already knows what to act on.
Sequel to cases 2.1-2.3 — Bilal spots the tasks really setting his deadline.
1. Not all tasks are equal
Some tasks can slip with no consequence, others shift the whole project end. The critical path is the longest task sequence, the one determining the project's total duration.
A delay on a critical task delays everything; a non-critical task has float instead. Bilal stops stressing over the mockup (with float) and watches the tool (float-free).
2. The PERT network
The PERT network models tasks and their dependencies as a graph, computing early/late dates, floats, and thus the critical path.
This analysis shows where to focus vigilance (critical tasks) and where you have slack (float tasks) — including to smooth resources or shorten the project.
🔎 Concrete example — 🔎 Concrete example — Shortening the project
Act on the right lever.
Crashing: add resources to critical tasks (extra cost).
Fast tracking: parallelize sequential tasks (increased risk).
→ speeding a NON-critical task changes nothing to total duration.Outcome: vigilance focused where it counts.
Reflex: recalculate the critical path along the way.
Section vocabulary
- Critical path
- The longest sequence of dependent tasks: it sets the project's minimum duration. Any delay on it delays the project.
Simple example: At Bilal's: develop then test the tool — float-free.
What it is not: the priciest or riskiest task.
Reflex: to save time, only act on this sequence.
- PERT network
- Graph of tasks and dependencies, used to compute early/late dates, floats and critical path.
Simple example: The graph of Bilal's packages, linked by dependencies.
What it is not: a Gantt chart (bars on a calendar).
Reflex: model as a network before computing dates and floats.
- Float
- Delay a task can take without shifting the project (total float) or the next task (free float).
Simple example: Bilal's mockup can slip a few days with no consequence.
What it is not: a consequence-free delay on a critical task (it has none).
Reflex: use floats to smooth resources.
- Early / late date
- A task's extreme start/end dates without delaying the project. Their gap gives float.
Simple example: Computed forward (early) then backward (late) on the network.
What it is not: dates imposed by the sponsor.
Reflex: the early/late gap reveals each task's slack.
- Critical task
- Float-free task, on the critical path: watch first.
Simple example: The tool development, tracked daily by Bilal.
What it is not: any late task (if it has float, it isn't critical).
Reflex: keep the short critical list posted and closely tracked.
In practice — In practice — Identifying the critical path
- Model your tasks and dependencies as a network (each task with its duration, linked to predecessors).
- Compute early dates (forward pass through the network), then late dates (backward pass).
- Derive each task's float (late − early): zero-float tasks form the critical path.
- Focus vigilance on critical tasks; use other tasks' floats to smooth resources.
Ready-to-say script — explaining the critical path to the sponsor (1 min)
"Here are the tasks setting our end date: they have zero float, I track them daily. Others have slack. If you want to save time, we can only act on critical ones — by adding resources or parallelizing, at a cost or a risk."
Say when a shorter deadline is requested: show where effort pays.
Copy-paste template — finding my critical path
NETWORK (tasks + durations + dependencies): _______________
EARLY DATES (forward pass): _______________
LATE DATES (backward pass): _______________
FLOATS PER TASK: _______________
CRITICAL PATH (zero float): _______________
REINFORCED WATCH ON: _______________Small business / tight budget
Even by hand on paper: list sequences, spot the longest — a spreadsheet is enough for small networks.
Algerian SME
Bilal recalculates after each contingency (vendor delay, outage): the critical path can change — vigilance follows the new critical.
Remote team
Publish the critical path in the shared space: the whole team knows which tasks must never slip.
Big project / software
Let MS Project or a tool compute dates and floats — but understanding the logic stays essential to arbitrate.
✓ Best practices
- Focus vigilance on critical tasks (zero float).
- Use non-critical tasks' floats to smooth resources.
- To shorten, only act on the critical path.
✗ Common pitfalls
- Speeding a non-critical task (no effect on total duration).
- Treating the critical path as frozen for the whole project.
- Forgetting to recalculate after a major delay.
Key takeaways
- The critical path sets the project's total duration.
- A delay on a critical task delays the whole project.
- Non-critical tasks have float.
- PERT computes dates, floats and critical path: where to be vigilant.
48h Challenge — apply tomorrow
On a 5–6 task mini-network of your project, compute early/late dates and floats (template above) and name the critical path. Post it for the team.
- I can define critical path and float without hesitating.
- I know why we only speed critical tasks.
- I've identified a real case's critical path.
Frequently asked questions
How to shorten a project whose critical path is too long?
Two classic techniques: "crashing" (adding resources to critical tasks, at extra cost) and "fast tracking" (parallelizing normally sequential tasks, at increased risk). Always act on the critical path — speeding a non-critical task changes nothing to total duration. It is the most frequent mistake: putting energy into tasks that already have float.
Can the critical path change during the project?
Yes: if a non-critical task is very late and exhausts its float, it can become critical and alter the critical path. That is why you recalculate regularly during tracking. The critical path is a snapshot at a given time, not data frozen for the whole project duration. A project manager therefore watches not only critical tasks, but also those dangerously close to exhausting their float.
Bilal's sponsor wants to save a week and proposes adding resources on the mockup, which has float. What to answer?