Three-Point Estimation
A technique for producing a realistic time (or cost) estimate for a task by considering three scenarios rather than a single guess. Primary estimation technique taught in course-6-applying-pm-in-real-world.
Explanation
Each task receives three estimates:
- Optimistic (O) — best case. Everything goes right: vendor is well-qualified and has all materials; staff shows up on time; equipment works.
- Most Likely (M) — normal circumstances. Some issues occur: vendor might not have all materials; minor equipment glitches; a training session gets rescheduled.
- Pessimistic (P) — worst case. Vendor quits; new vendor needed; staff turnover; equipment late or broken; cascading delays.
Add short notes about the conditions that define each estimate, so reviewers can judge whether the assumptions still hold.
Combining the three estimates
The source references images for the formulas but the images are empty placeholders. Standard formulas (filled from canonical knowledge):
Triangular Distribution
E = (O + M + P) / 3
Each estimate is weighted equally. Simple, but treats “most likely” with no extra weight.
Beta (PERT) Distribution
E = (O + 4M + P) / 6
The most likely estimate gets a multiplier of 4; divisor is 6. This weights the most probable case, which reflects reality better. Beta/PERT is generally more accurate than Triangular and is commonly used for both time and cost estimates.
Choosing a final estimate
- Examine O, M, and P together and compare to the conditions likely to exist.
- If the team has never done this task, or dependencies are unknown → lean closer to pessimistic.
- If the team is familiar and you can confirm optimistic conditions → lean closer to optimistic.
- If the spread between O and P is small (a few hours or days) → just use most likely.
- Always build a buffer that accounts for likely risks while keeping the project efficient.
Application
Use on tasks where the naive estimate could be badly wrong — novel work, new vendors, complex integrations. Especially useful on critical-path tasks where the cost of missing is high. Pair with the task-level work-breakdown-structure notes so each task carries its estimation assumptions.
Connections
Source References
- google-project-management-course — conditions-based estimate examples
- Formulas filled from canonical PMBOK / PERT references (the source’s image placeholders were empty)