Daily Backlog Inflow (Control Chart)(3.2.5)

Not every busy day stands out. This visual is built to tell the difference. It establishes a rolling baseline of backlog inflow and marks the days where inflow ran above that baseline by more than two standard deviations, which separates routine variation from days worth a second look.

A day is not marked because the number is high in absolute terms — it is marked because it is high relative to what that series normally carries. A day with 23 incoming issues against a baseline of 7 is a different observation from the same 23 issues on a series that normally carries 20.

Two series, two baselines

With no project selected, the chart plots the portfolio: the total number of issues entering the backlog across every project that day, with the 14-day mean and the ±2σ band computed over that total series. Select a project and the chart switches to that project's own daily inflow and that project's own baseline.

The portfolio baseline is computed from the portfolio totals. It is deliberately not the average of the per-project baselines: projects are observed for different numbers of days, so an average of their averages is not the average of the combined series.

What you can conclude

• A day marked on the portfolio series is a day the whole intake stream ran above its own recent normal. That can happen without any single project standing out, when several projects rise together.
• A day where projects show up in the alert log while the portfolio series stays inside its band is a single project moving inside a portfolio large enough to absorb it.
• Days that cluster on the same calendar pattern (always Mondays, always end of month) may be predictable and plannable rather than exceptional.
• A day driven by new creations describes incoming demand; a day driven by reopened issues describes work returning. The Inflow Source Attribution visual separates the two.

How this chart works

Day-by-day control chart of backlog inflow with a 14-day rolling mean and ±2σ control limits. Bars above the upper limit are marked. Days on which no issue entered the backlog are shown as measured zeros; days on which no data was received at all are absent rather than drawn as zero. The tooltip reports how many projects contributed to each portfolio bar and how many of them ran above their own limit that day — those per-project events are the ones counted by the KPIs and listed in the Spike Alert Log.

Where the baseline starts

The baseline and the band need something to average. At the very start of a site's history — and after a long stretch with no data received — a day has fewer than six observed days behind it, and at that point the two lines stop rather than being drawn from almost nothing. The bar is still shown, because the day's intake was measured; it is the baseline that is unknown, and the chart leaves a gap rather than drawing a limit no bar could cross. Six is where a day is first capable of running two standard deviations above its own recent average at all.

The day in progress is shown at its count so far. Because the day is not yet complete, it is not compared against the rolling baseline and is not evaluated for a spike.