Updated September 2026: in February 2020, after two quiet days, my manager told me new work had arrived that looked complicated and would take two weeks. From experience, his estimates tended to be the reverse of reality. I have rewritten it into advice on calibrating the estimates you are given for new work.
After two hard weeks of leadership reviews and a difficult account team, I had two days free. Then my manager messaged that new work had arrived: complicated, about two weeks of effort. Two worries followed. When he called work easy, it usually turned out hard, and the reverse; and when he said two days, it tended to take the whole week.
Why are handed-down estimates often wrong?
Because the person giving them is usually not the one doing the work, and has heard only a summary. Some managers overstate effort to secure capacity; others understate it to get a quick yes. Neither is dishonest, but both mislead.
How do you calibrate someone's estimates?
Keep a simple record: what was estimated, what it actually took. After a few requests, you will see a pattern, such as a manager who is consistently optimistic by a factor of two. Knowing the pattern is more useful than guessing each time.
What questions give you a better estimate?
- What exactly is the deliverable, and in what format?
- What already exists that can be reused?
- Who else is involved, and what do they owe?
- When is the first review, and when is the final deadline?
Answers to these let you form your own estimate within a day.
How do you share your estimate?
Share it with your manager with the reasoning: "Given two review rounds and the pricing dependency, I think this is about eight days." A reasoned estimate is hard to dismiss, and it gives your manager something to plan with.
Should you enjoy the quiet days before the work starts?
Yes, but use part of them to prepare. Gather the reusable material, read the client's background and set up the files. When the work turns out larger than promised, those two days of preparation are what keep it from spilling into evenings.
What if the estimate is wrong anyway?
Update it as soon as you know. Say what changed and the new expected date. Early revisions are routine; late surprises damage trust.