What is the difference between stock keeping and an item master?

These are two layers that teams constantly blur. The item master is the dictionary: one record per SKU holding the things that are true everywhere: name, description, unit of measure, barcode, supplier, cost, category, whether it's lot-tracked or serialized. Stock keeping records are the ledger: how many of that SKU sit in bin A-12, in van 4, at the jobsite, right now. One definition, many balances. The separation matters because the two change differently and break differently. Item master changes are governance events: rename a SKU, change its unit of measure, merge a duplicate, and every location and every historical report is affected at once. Stock records change hundreds of times a day through receipts, picks, and transfers, and their failures are count drift, not definition drift. Most spreadsheet inventories collapse the layers into one row per item with a quantity column, which works until the second location shows up. Then someone adds a column per location, then a tab per truck, and now the same item's name is spelled three ways across tabs and no report can sum it. The practical hygiene: one owner for the item master with rules for creating and merging SKUs, and unit-of-measure discipline (is the count in eaches or cases?) settled in the master, not improvised at each location.