How much should you pad the next estimate?
Twenty-four finished tasks, each with what it was estimated at and what it actually took. The mean miss is plus 17% and twenty of the twenty-four ran over, which is a bias rather than bad luck.
Work Starter Statistics free
After you install, this is the model to open.
How Wrong Are Our Estimates, Usually?
- In your spreadsheet, click the Sortia icon in the strip of icons down the right-hand edge. No strip? Click the arrow at the bottom-right to open it. You can also use Extensions, then Sortia, then Open Sortia.
- Click Start from a template and put that name in the search box.
- Pick the card with that name and click Load this template. It arrives on a new tab with real numbers already in it.
The answer
The size of a typical miss, and its direction:
- Mean miss
- +17% median +16%
- Ran over
- 20 of 24 tasks
- Spread
- 15.01 percentage points
- True average overrun
- 11% to 23% at 95%
A bias is the one estimating error you can fix without getting any better at estimating: multiply the next estimate by 1.17 and most of it is gone, and the interval says 17% is the middle of the evidence rather than the cautious end of it. The last column is worth reading before you leave. The four tasks that came in under estimate are the four sized by the person who then did the work, and the twenty that ran over were sized in a planning session. What none of this tells you is which tasks will be the bad ones, because the spread stays 15 points wide after the bias is corrected.
The model
Estimate and actual side by side with a percentage miss column, so a two-day task and a twenty-five-day task can sit in the same summary, plus a column recording who sized each one.
| Tasks | 24, all finished |
| Estimates | 3 to 25 days |
| Miss | calculated as a percentage |
| Confidence level | 95% |
| Mean, median, spread and interval | calculated |
Once it is in your sheet
- The model arrives with real numbers in it and runs as it stands, so you can press the button first and understand it second.
- Change the numbers to yours. The sheet marks which cells are inputs and which hold formulas, and most labels carry a note explaining the row.
- Press the run button at the bottom of the panel. It is labeled for the tool you are in, and the result lands on its own tab, with a written reading of it beside the figures.
Never used Google Sheets? Start here goes the whole way, in seven steps, and assumes nothing.
Next question
- What three-point estimate does your history give?What Shape Are Our Task Durations?
- How many people does the rota actually need?Cover Every Hour With the Fewest People
- What is a point of utilization actually worth?Will the Team Bill Enough Hours?
- Twelve recurring tasks timed before and after an automation rollout. The paired test makes the 23-minute average saving certain, the sheet turns it into a 13-week payback, and the two tasks that got worse are the real finding.Did the Automation Deliver?
- Which big clients are quietly the expensive ones?Which Kinds of Client Do We Actually Have?
- Is the bigger fee the better client?Which of Two Clients Should Get the Slot?
Every model like this one, and the method behind them: Statistics in Google Sheets.