Why I stopped measuring the wrong week

Weekly metrics flatter whoever picks the week. A short argument for the rolling quarter and the long memory it forces.

Adaeze Okonkwo

Adaeze Okonkwo · · 2 min read

A printed chart on a desk beside a keyboard

Every Monday for two years I opened a dashboard and read seven days of my own work back to myself. The number was always available and almost always meaningless, and it took me an embarrassingly long time to work out why.

The week is a choice, not a unit

A week has no relationship to how work actually arrives. Research lands in a lump after a fortnight of nothing. A rewrite shows up as four flat days and then one where the whole thing resolves. Measured weekly, the fortnight of reading is a failure and the day of resolution is a miracle, and neither description is true.

Worse, I got to pick the boundary. Not consciously — but when a week looked thin I would find a reason it had been unusual, and when it looked good I would not look for one. The measurement was downstream of how I already felt.

A metric you can reframe after the fact is not a measurement. It is a mood with a number attached.

What a rolling quarter forces

I moved to a trailing ninety days, recalculated whenever I looked. Three things changed immediately, and only one of them was about the numbers.

  • A bad week stopped being an event. It moved the line by about a percent, which is roughly how much it actually mattered.
  • I could no longer tell a story about any single week, because no single week was visible. The shape was the only thing left to read.
  • The quiet, structural work — the reading, the maintenance, the half-finished drafts — finally showed up, because ninety days is long enough for it to turn into something.

The last one was the real change. Weekly measurement had been quietly teaching me that only legible output counts, and I had been arranging my work to produce it.

What it costs

A rolling quarter is slow to react. If something genuinely breaks, you will see it a month after you felt it, and you have to trust the feeling in the meantime. I think that trade is worth it. The weekly number did not catch problems early either — it just produced an opinion early, which is a different thing.

Ninety days is also long enough to be honest about. I can remember a quarter. I cannot remember a week, which is exactly why the weekly number was so easy to narrate.

Share
Adaeze Okonkwo

Written by

Adaeze Okonkwo

Writes the interviews here. Spent eight years asking people the question just after the obvious one.

Lagos, Nigeria

Responses

Members can reply, and replies thread. Disagreement is welcome; the only rule is that it is about the argument.

Three more pieces on the same thread.