⏳PERT Estimate Calculator

About this tool

List each task with an optimistic, most likely and pessimistic estimate, in days, hours or any unit. You'll see each task's expected time and spread, the project's expected total, and the duration you can promise at 50%, 80%, 90% and 95% confidence. You'll also get a warning if an estimate isn't in order.

How PERT estimates work

Single-number estimates are almost always optimistic, because things rarely go better than planned but often go worse. PERT (Program Evaluation and Review Technique), developed for the US Navy's Polaris programme in the 1950s, asks for three numbers per task: optimistic (O), most likely (M) and pessimistic (P). The expected duration is (O + 4M + P) ÷ 6, a weighted average that lets a long pessimistic tail pull the estimate up, and the standard deviation is about (P − O) ÷ 6. For tasks done one after another, the expected durations add up. The uncertainty combines as the square root of the summed variances, so the total is more predictable than any single task. The example's four tasks have a most likely sum of 23 days, but an expected total of 24.8 days, with about a 90% chance of finishing within 28.2 days. Commit to an 80–90% figure rather than the most-likely sum, and keep the difference as contingency. Asking the people who'll do the work for their three numbers also surfaces risks early. Honest limits: the maths assumes independent, sequential tasks. Work done in parallel, shared risks (the same person off sick, the same supplier late) and scope changes aren't captured, and estimates are only as good as the inputs. The Gantt chart, the RICE score calculator and the working days calculator are related.

Frequently asked questions