Humans are terrible at estimating time, and it is usually because we omit the smaller steps involved in completing a task. The fix is a question. Say the task out loud, ask "and then?", answer honestly, and keep going until the task runs out of places to hide. Every answer becomes its own task with its own estimate, and suddenly the number on your sheet is real.
Where the question comes from
There is a scene in the 2000 cult classic Dude, Where's My Car? where Ashton Kutcher's character is placing an order at a fast food drive-through. After every item he lists, the attendant asks "and then?" He repeats the order. "And then?" He repeats it again. His frustration builds until finally, asked "and then?" one more time, he snaps back with "and then you can put it in a brown paper bag."
That is it. That is the technique. That level of granular, slightly maddening detail is exactly what you are after.
I quoted a stoner comedy in a business book and I would do it again, because nobody forgets it.
What it looks like on a real task
Here is one of mine. I could easily put this on my sheet as "blog article, 1.5 hours." And I would believe it, because I am a decent writer and 90 minutes for an article sounds generous. I wrote a book, for God's sake.
Then I play the game.
I need to write an article.
And then? I probably need to spend some time finding images to put in the article.
And then? I should ask someone to proofread it.
And theen? I need to get it ready to post online.
And theeen? I'll want to get it posted on our social media networks and in our newsletter.
And theeeen?
Now I have five tasks instead of one. My 90 minutes was never 90 minutes. It was 90 minutes of writing plus three or four hours of everything that has to happen for the writing to matter.
That difference, multiplied across a six-week plan, is why the plan collapses in week two.
The side effect nobody notices
There is a subtle second benefit in that example, and it is easy to miss.
When I listed "I should ask someone to proofread it," I was planning someone else's time as well as my own.
That means when the task goes onto my overview sheet, I can also ask that person to block time in their week for it. They are not ambushed on Thursday with an urgent request. It was on their sheet on Monday.
When you start planning your time in advance, you encourage the people around you to do the same. It spreads without you having to sell anyone on a methodology.
How far to take it
Stop when the next answer is something you would do automatically without thinking, or when every remaining piece is under 90 minutes and you can actually picture yourself doing it.
That second test is the useful one. If you cannot picture doing the task, there is at least one more "and then" hiding in it. Picturing it is not a metaphor here, it is the mechanism: playing the work forward in your head is what surfaces the roadblock before you hit it, which is also the reason this is not a job to hand to a model.
You do not need to reach the brown paper bag. But you should get closer to it than feels necessary.
Where it fits in the cycle
The game runs during planning, right after the brain dump and before anything lands on a sheet. The dump tells you what exists. The game tells you what each item actually costs. Only then can you arrange days so none of them holds more than six hours of work.
Skip it and you will plan a beautiful six weeks made entirely of tasks you have not thought about yet.
The whole sequence is in the guide to running your work in six-week cycles.