What a daily total hides about a Tuesday morning

A municipal portal’s daily chart was honest and still useless. Splitting staff from citizens changed the rooms that needed work.

Daily totals are not lies. They count something. The trouble is that they count several kinds of presence as one: a clerk logging in to print a certificate, a parent looking for a school-place form, a person opening the same PDF three times because the page does not say the window has closed.

An editor of a city services portal brought us a rumour: nobody uses the thing after 14:00. The daily chart could not speak to the rumour. It did not distinguish a staff session from a citizen visit. After the sample was split — a morning’s work for their information desk, not for us — the afternoon filled with staff traffic, and three citizen rooms went quiet: school-place appeals, the waste calendar, and a cupboard of PDFs.

The useful sentence was not “usage falls after lunch.” It was “citizens finish the three rooms they can finish in the morning, and the PDF cupboard does not say it is closed.” That sentence can be acted on. A daily total can only be watched.

If you keep a chart of days, keep beside it a cut that the product itself understands: citizen against staff, first opening against return, booking against complaint. If you cannot cut the sample that way, the chart is a weather report. It will not settle a redesign.