A PERT chart emphasizes task relationships and uncertain duration estimates in a network; a Gantt chart emphasizes tasks across calendar time. Use PERT to reason about uncertainty and Gantt to run the schedule.
What is the difference between a PERT chart and a Gantt chart?
A PERT chart is a network diagram that emphasizes activity relationships and can use optimistic, most likely, and pessimistic estimates to model uncertain durations. A Gantt chart is a time-scaled bar chart that emphasizes planned start and finish dates, duration, ownership, progress, milestones, and the calendar forecast.
Use PERT when the hard question is “How is the work logically connected, and how uncertain are our estimates?” Use a Gantt chart when the hard question is “When will the work happen, and are we still on track?”
Both can represent the same activity network and identify a critical path. Modern scheduling software often combines network calculations with a Gantt interface, so you may use PERT thinking without maintaining a separate PERT diagram.
PERT vs Gantt at a glance
| Question | PERT chart | Gantt chart |
|---|---|---|
| Visual form | Nodes and connecting arrows | Rows and horizontal time bars |
| Main emphasis | Logic and uncertainty | Calendar and execution |
| Time scale | Often not drawn to scale | Drawn against dates |
| Duration estimate | Often three-point | Usually one current duration |
| Dependencies | Central | Central in a robust schedule |
| Resource assignments | Not the focus | Commonly shown |
| Progress tracking | Awkward | Natural |
| Best stage | Analysis and early planning | Planning, communication, and control |
What does PERT mean?
PERT stands for Program Evaluation and Review Technique. It was developed for work where activity durations were uncertain. Instead of relying on one estimate, the classic method asks for three:
- Optimistic (O): a favorable but plausible duration;
- Most likely (M): the duration expected under normal conditions;
- Pessimistic (P): an unfavorable but plausible duration.
The traditional weighted expected duration is:
Expected duration = (O + 4M + P) / 6
If a task has O = 4 days, M = 7 days, and P = 16 days, its PERT expected duration is:
(4 + 4×7 + 16) / 6 = 8 days
That is not a guarantee. It is a structured estimate that gives the most-likely case four times the weight of each outer case. The spread between optimistic and pessimistic values also exposes where uncertainty deserves attention.
What does the network reveal?
Consider a small approval project:
| ID | Activity | Expected duration | Predecessor |
|---|---|---|---|
| A | Define requirements | 3 days | — |
| B | Design option | 5 days | A |
| C | Security review | 4 days | A |
| D | Build prototype | 6 days | B |
| E | Approval test | 3 days | C, D |
| F | Final decision | 0 days | E |
The network branches after A: design/prototype and security review can run in parallel. E cannot start until both branches finish. A forward pass calculates earliest dates; a backward pass calculates latest allowable dates. The longest path through the network controls the earliest finish and is the critical path.
Read What Is the Critical Path Method? for the calculation and practical use of float.
How would the same project look in a Gantt chart?
The Gantt chart places A through F on actual working dates. It can show that the security reviewer is away on Friday, the prototype owner has another assignment, and the approval meeting only happens on Tuesdays. It also gives the team a surface for recording progress and revising remaining duration.
This calendar context matters. A mathematically valid network can still produce an unrealistic forecast if resources, holidays, constraints, and update practices are ignored.
When should you use PERT?
PERT-style analysis is useful when:
- the work is novel and historical estimates are weak;
- research, engineering, or approval durations vary widely;
- several parallel paths converge on one milestone;
- stakeholders need to understand schedule uncertainty;
- the team is deciding where to buy contingency or reduce risk;
- a single deterministic finish date would be misleading.
Do not turn every routine task into an elaborate three-point exercise. Focus it on high-uncertainty or critical activities where the range can change a decision.
When should you use a Gantt chart?
Use a Gantt chart for the operating schedule: assign owners, set working calendars, show milestones, update actual progress, and communicate forecast dates. It is more legible for weekly project control because bars reveal overlaps and slippage across a calendar.
Use the seven-step Gantt method to build the operating plan, review float and slack, and test the result with the schedule quality checklist.
Can you combine PERT and Gantt?
Yes. A practical combined workflow is:
- Build the activity network without dates.
- Identify uncertain activities and collect O, M, and P estimates.
- Calculate expected durations and challenge extreme ranges.
- Identify the initial critical and near-critical paths.
- Move the network into a calendar-aware Gantt schedule.
- Assign owners and check resource availability.
- Save an approved baseline in a tool that supports formal baselines, or preserve an approved snapshot.
- Update actuals and remaining durations; reassess risk as evidence improves.
For a major commitment, a Monte Carlo schedule-risk analysis is stronger than treating the PERT expected value as a promised finish date. It samples duration distributions repeatedly and reports probabilities such as “an 80% chance of finishing by October 20.”
Common PERT mistakes
- Treating optimistic as “best imaginable” and pessimistic as “disaster.” Both should be plausible cases.
- Letting a manager choose all three estimates without the people who understand the work.
- Adding contingency to every input and again to the final schedule.
- Reporting the weighted mean as a confidence level.
- Ignoring correlation between tasks affected by the same risk.
- Calculating logic once, then managing execution in a disconnected task list.
How can GanttFather support a PERT-based plan?
Calculate or workshop the optimistic, most likely, and pessimistic estimates first, then enter the chosen planning duration for each activity in GanttFather. Connect the network with FS, SS, FF, or SF relationships and lag, apply working dates, assign owners, and use critical path to see which chain currently controls the finish.
GanttFather does not perform probabilistic PERT or Monte Carlo simulation. It is the calendar-aware operating schedule after the uncertainty analysis, so retain the three-point estimates and risk assumptions outside the single duration entered in the chart.
Create a free GanttFather project for the calendar schedule once the activity network and planning durations are ready.
Frequently asked questions
Is PERT the same as critical path method?
No. Both use activity networks. CPM traditionally uses deterministic durations to identify the longest path, while PERT was designed around probabilistic or three-point duration estimates. In practice, project teams often combine their ideas.
Is a PERT chart better than a Gantt chart?
Neither is universally better. PERT is clearer for network logic and uncertainty; Gantt is clearer for dates, overlap, ownership, and progress. Choose the artifact that supports the decision.
Can Excel calculate a PERT estimate?
Yes. If optimistic, most likely, and pessimistic estimates are in cells B2, C2, and D2, use =(B2+4*C2+D2)/6. A spreadsheet still requires a separate network or schedule model to calculate dependent dates.
Does PERT include resources?
Classic PERT focuses on activities, logic, and duration uncertainty. Resource constraints must be analyzed separately and can change the feasible critical path.
Sources
