The Difference Between a To-Do List and a Calendar Is an Estimate

A to-do list can hold thirty items and remain, in some sense, plausible. A calendar cannot, because a calendar has a bottom. When you move the same thirty items onto a day, the eleventh one lands at seven in the evening and the rest do not fit at all, and this is not a flaw in the exercise. It is the exercise.

I spent three years as a consultant working from lists and being reliably wrong about what a week held; how few focused hours a week actually contains is the other half of that error. The correction was not a better list. It was being made to write down, for each item, how long it would take — which a list never asks and a calendar always does. My estimates in the first month were off by a factor I found genuinely embarrassing: tasks I called thirty minutes averaged just over an hour.

That factor is the reason most people abandon this in week one. You block your day, you fall behind by ten in the morning, and the natural conclusion is that the method does not work. The accurate conclusion is that your estimates do not work, and you have just been given the first honest measurement of that in your career. It stings, and it is the most valuable output of the whole practice.

So the first month is not for planning. It is for calibration. I would suggest doing it in the least ambitious way possible: keep your existing list, and each morning move only the three things you actually intend to do onto the calendar with a time estimate — fewer blocks work better anyway. At the end of the day, write the real duration next to each. Do not adjust anything else. After twenty days you will have sixty data points and a personal multiplier, and mine turned out to be about 1.8.

Applying the multiplier is where it becomes useful, and it is worth being mechanical about it rather than trying to estimate better. I did not get better at estimating; I got a number and I multiply by it. Anything I feel is an hour goes in as an hour and three quarters. This has been more accurate than four years of trying to be more careful, because the error was never carelessness, it was the planning fallacy, a systematic bias, and biases do not respond to effort.

Two specific categories break the multiplier, and both are worth blocking differently. Anything involving another person’s response should be blocked as the work you will do in the gap, not the task itself — “send the brief” is four minutes, and “wait for the brief to be approved” is not a block at all, it is a dependency and belongs somewhere else entirely. And anything you have never done before should not be estimated; block a fixed exploratory hour and re-estimate afterwards, because a first estimate of an unfamiliar task is not a bad estimate, it is not an estimate.

The failure mode I would warn about most strongly is what I did in month four: treating the calendar as a commitment device and feeling obliged to finish a block because it said so. That produces worse work than the list did, because the estimate was always approximate and finishing something inside a wrong estimate means finishing it badly. The block is a prediction. When the prediction is wrong, the correct response is to update it, which takes ten seconds and is the entire skill.

Three years in, the practical effect for me is not that I do more. It is that on Wednesday I can say what will and will not happen this week, and be right about it most of the time, and that turns out to be worth considerably more to the people I work with than the extra hour I used to imagine I was going to find.