Comparison
Planning fallacyvsReference-class forecasting
Planning fallacy
you estimated the happy path, which is the only version of the work you pictured.
The systematic tendency to underestimate time and cost because we plan from the best-case scenario. It survives experience: people who know they were late last time are still optimistic about this time, because each project feels uniquely well understood. The countermeasure is not more thinking but different evidence, which is what reference-class forecasting supplies.
Full entry →Reference-class forecasting
instead of estimating the migration, you looked at how long the last three migrations actually took.
Estimating by comparison to how long similar past work really took, rather than by decomposing the work in front of you. It is markedly more accurate than bottom-up estimation because it silently includes all the interruptions, rework and forgotten steps that never appear in a breakdown. Its requirement is data: a team that never records how long anything took cannot use the outside view, which is a cheap thing to start collecting.
Full entry →