My Weekly Review Went From 70 Minutes to 20 by Removing Four Things

My weekly review had eleven steps and took a little over an hour, and I had built it up one step at a time over six years, always in response to something going wrong. That is a reasonable way to build a process and it has a predictable outcome: every step is a monument to a past failure, and none of them ever get removed, because removing one feels like inviting that failure back.

I removed four this spring and the review now takes about twenty minutes. It has been eleven weeks and none of the four failures has recurred, which is not proof, but it is enough that I want to write down what went and why.

The first cut was reviewing the completed items. I had been going through everything finished that week, ostensibly to notice patterns. In six years I can identify exactly one useful observation that came from this, and it came in the first month. What it actually provided was a sense of accomplishment, which is a real benefit, and which I now get in about fifteen seconds by looking at the count rather than the list.

The second was the calendar look-back. Going through the past week’s calendar to catch anything that generated follow-up work. This was added after I forgot to write up a meeting. The reason it can go is that I fixed the underlying problem differently — I now write the follow-up in the last two minutes of the meeting itself — and once the real fix was in place, the review step was checking for a category of error that had stopped occurring. It took me two years to notice that.

The third was the full capture inbox pass. Getting to zero every Friday, processing everything into tasks or archive. This one I was most nervous about and it turned out to be the biggest single saving, about twenty-five minutes on a normal week. What replaced it is a search for two things — unanswered questions addressed to me, and anything with an attachment I have not filed — which takes three minutes and catches, as far as I can tell, everything the full pass caught that mattered. The rest of what the full pass did was archive things, and things archive themselves by being ignored.

The fourth was reviewing all my ongoing responsibilities in turn. Six areas, a couple of minutes each, checking whether anything needed attention. This felt like the most conscientious part of the review and I now think it was the least useful, because a responsibility that needs attention almost always announces itself some other way, and a responsibility that is quiet produces nothing when you look at it except a vague feeling you should do more.

What is left is four steps: the three questions from my recovery review, plus a pass over next week’s calendar to see what is coming. Twenty minutes, and it is the calendar pass that produces most of the actual decisions.

The thing I would generalise is the question I used to decide each cut, because I think it applies to any process that has accreted over years: when did this step last change what I did? Not what it might catch, not why it exists — when did it last cause a different action. Four of my eleven steps could not answer, including two I would have defended strongly if asked in the abstract.