One of the things that probably passed under the radar in my last post, Once upon a time…, is the following statement:
Set a schedule, sit down and write. Set a schedule, a time slot, sit down and write.
(emphasis is mine)
My to-do lists are words in the wind
It’s a strong statement, I know. Let me elaborate on that.
I run my day with to-do lists, religiously, for more than 2 years. The routine was something like:
- Define what you want to do, today
- Check all the done items
- At the end of the day:
- run a short retrospective
- copy leftovers to the day after
After a year or so, I added a sort of prioritization:
- At the beginning of the day select the 4 most important things and work on those first.
I failed :-)
It never worked out very well, some days were literally awesome, but the majority were still a nightmare. Here is why:
I’m a procrastinator
I’m very good at convincing myself that something can be done later and that there are more important things to do now. You know, things like watching TOP 10 Awesome Lego Creations on YouTube. Being a procrastinator makes it very hard to stick to a plan that is not really a plan, but more a suggestion.
Estimates are the root of many evils
“Define what you want to do, today” combined “At the beginning of the day select the 4 most important things and work on those first.” starts from the assumption that it’s reasonable to predict that those 4 things can be done in a day. If most of the times that turns to be wrong, your to-do list becomes a sort of constantly forward sliding window.
The way we are set up in Particular Software makes it a little bit more complex to predict in advance how your day will be, making a relaxed, and to-do lists based, unplanned day a kind of wounded territory for disruptions. Many things might happen that drag you away from your original to-do list for the day. And this in combination with being a procrastinator is a recipe for a productivity disaster.
All this can only lead to poor satisfaction: you reach the end of the day, look back at your to-do list to find out that:
- the original plan, mostly, failed
- there are many more things on the list than at the beginning of the day
- most of them are not checked, so they’ll simply clutter your tomorrow
Short term planning changed my day
I’m now experimenting, with success so far, a different way of organizing my week. Yes, the whole week. It’s a combination of careful planning, relaxed planning and to-do lists. Plus a little bit of refactoring.
I turned my to-do list into a simple backlog, I’ll add all the things I want or I need to do. This is it, no priorities, no estimates, nothing. Just a flat list of things that sooner or later needs to be done.
Every Friday morning I spend ~45 minutes planning next week:
- I’ll go through my backlog, and based on various factors, I decided what I’ll work on next week
- the lucky backlog items will find a spot on my calendar
Finding a spot on a calendar implies a rough estimate of the work required to get the thing done. My rule of thumb is: make it so that it’s not longer than 1 hour. If it is: split the work into multiple backlog items and re-plan.
The end goal is to fill the first 2 and a half days of the week, from Monday to Wednesday at lunch time. All the scheduled slots are set as
It doesn’t end here. On Fridays I also plan the rest of the week, from Wednesday afternoon to end of Friday. The difference is that all the scheduled slots are set as
free. This means that whoever tries to schedule something with me can hardly find a spot till Wednesday morning, but will see the rest of the week as free.
On Wednesdays morning I finally consolidate the rest of the week, turning what were
free slots to
busy, and adjusting or changing what needs intervention (refactoring).
Planning for the unplanned
Everyday there is one slot, sometime during the day, defined as
Inbox triage & TODOs. It basically serves two purposes:
- Groom my backlog by looking and new things that are coming through my inbox, whatever inbox means (we’ll talk about that in a future post)
- Act like a buffer to absorb wrong estimates.
Buffer consumption is what’s happening right now with this post. It was planned for this morning, last slot before lunch time. Due to a meeting that went late in the first slot in the morning, the whole morning has been pushed back by 30 minutes. The
Inbox triage & TODOs is now giving away for “free” 30 minutes to allow me to:
- finish the post
- make so that morning delay doesn’t affect the afternoon
What about emergencies then?
They don’t exist, I’m sorry. Your emergencies are not allowed to disrupt my day. In Particular we have a clearly defined process to handle them. This is the On Call schedule. And this the way to handle things when the roof is on fire. Everything else can wait 2 days. If something that needs attention happens on Monday afternoon, it can easily wait till Wednesday afternoon. Otherwise it means that we have a much bigger problem than the emergency itself. This is another story, though, that deserves it’s own post.