Analysis for people holding stock.

A last_sold column is updated by stock counts, transfers and returns as well as by sales, so it drifts toward looking more active than the product is.

Rebuild ageing from the order lines and report how far the two disagree, because that gap is itself a data quality finding worth acting on.

Nothing is recorded on the days you had nothing to sell, so the absence has to be reconstructed from zero-stock windows in the movement history rather than inferred from an absence of sales — those look identical and mean opposite things.

Value them at the product's own rate of sale either side, then re-run at a deliberately conservative rate and report both. A single confident figure for something that did not happen is the most misleading number available.

Half the dead value in one worked example arrived on four purchase orders where the quantity was set by a supplier's minimum rather than by demand.

That is the actionable finding, and it is invisible in any report that stops at which SKUs are slow. Seasonal lines need excluding first, and the exclusion listed, so somebody can disagree with it.

A stock file with quantities and costs, plus sales history at the order line. That pair is enough to age stock properly and rank it by the cash clearing it would release.

Movement history makes stockouts measurable and purchase orders make the cause traceable. Each additional file adds a section rather than being a prerequisite, so start with what you have.