A Field Checklist for Three-Point Estimation You Can Use This Week
Ask someone how long a task will take and you usually get one number — the number they hope for on a good day. Single-point estimates feel decisive, but they hide everything that matters: how much the work could swing, and which way the surprises tend to run. Three-point estimation fixes this by asking for three numbers instead of one, then combining them into an expected value and a sense of spread.
The method has been quietly useful for decades, and it is especially handy now, with hybrid teams handing work back and forth across time zones and supplier lead times still behaving unpredictably. You do not need software. You need three honest numbers per item and a little discipline. Here is a checklist you can run this week.
Before you estimate
Break the work down until each item is something one person can picture finishing — large items hide the variability that makes estimates useful.
Ask the people who will actually do the work, not just the lead. They know where the soft ground is.
Agree on what 'done' means for each item, so the three numbers describe the same finish line.
Decide your unit up front — hours, days, or story-sized effort — and stay consistent.
Getting the three numbers right
Optimistic (O). The time if everything cooperates — no rework, no waiting on others, no surprises. This is a realistic best case, not a fantasy.
Most likely (M). The duration you would bet on if you had to commit today, given a normal run of luck. This usually anchors the estimate.
Pessimistic (P). The time if the plausible bad things happen — a key person is out, a dependency slips, the integration fights back. Not the apocalypse, just a bad-but-believable week.
Expected value (E). Combine them. The simple triangular average is (O + M + P) / 3. The PERT weighting, which trusts the most-likely figure more, is (O + 4M + P) / 6. Pick one and use it consistently.
Spread. The PERT standard deviation, (P − O) / 6, tells you how uncertain the item is. Two tasks with the same expected value but very different spreads are not the same risk — and the wide one deserves your attention.
Turning estimates into a plan
Sum the expected values to get the project's expected duration, but do not stop there. The reason three-point estimation beats a single guess is that it surfaces uncertainty you can act on. Combine the spreads across items to set a contingency buffer that reflects real risk rather than a flat ten-percent fudge. Hold the buffer at the project level, not hidden inside each task, so it does not quietly get spent.
A few habits keep the method honest. Watch for anchoring — once someone says a number, the other two drift toward it; ask for the pessimistic case first to break the pull. Resist padding individual estimates 'to be safe,' because that is exactly what the explicit buffer is for. And after the work is done, compare actuals to your three points: over a few cycles your team's estimates get noticeably sharper, which is the real payoff.
If you want estimates your sponsors can trust and a buffer that holds up to scrutiny, XNM's program & project delivery advisory can help you build the discipline into how your teams plan.