# Supply Run and Sourcing Sheet **Version:** 1.0 **Date:** 2026-08-22 **Type:** Instrument — a blank working sheet, filled at the shelf --- ## What this is for A par level tells you how much of something the property should hold. It does not tell you when to go and get it, where from, or what it cost when you did. This sheet is the other half: **one page per supply run.** It is filled in by whoever does the run. It needs no login, no spreadsheet and no training. If you can carry a clipboard to a shelf, you can complete it. It pairs with the par-level template, which supplies the first two columns. --- ## 1. The sheet One row per item. The first three columns are known before you leave; the last four are written at the shop and at the shelf. | Item | Par | On hand | Gap | Source | Unit price | Line total | |---|---|---|---|---|---|---| | | | | | | | | | | | | | | | | | | | | | | | | - **Par** — from the par-level template. Copy it; do not re-estimate it here. - **On hand** — counted, on the day, at the shelf. Not remembered, not carried over from the last sheet. - **Gap** — par minus on hand. Written down even when it is zero, because a zero you wrote is different from a line you skipped. - **Source** — where it was actually bought, not where it is usually bought. - **Unit price** — what one unit cost on this run. - **Line total** — units bought x unit price. **Run total: ____________ Receipt total: ____________** --- ## 2. Scheduling the run The gap between runs is not a habit and not a calendar convenience. It is the **cycle length** already used to build the par: ``` Par = (units served x consumption per unit x cycle length) + safety stock ``` If runs happen every ten days, par must be built on ten days. If someone shortens the run cycle to a week and nobody rebuilds par, the property now carries three days of stock it is paying for and storing without reason. If the cycle stretches and par does not, the property runs out on the last two days of every cycle and nobody records why. **So: change the cycle, rebuild the par. They are one number in two places.** Write the intended cycle at the top of the sheet, and the date of the previous run beside it. Those two figures together are the only early warning you get that the schedule has quietly drifted. **Cycle length: ______ days Previous run: ____________** --- ## 3. Coupons, local sources, and the rule that keeps them honest Local sourcing and coupons are worth real money, and they are also where supply records most often stop being true. The rule is one line: > **A discount is not a saving until it appears on the receipt.** A coupon that expired, that did not apply to the size actually bought, that needed a loyalty card nobody had, or that the till simply did not take, is not a saving. It is a saving that was *expected*. Expected savings recorded as real ones make the next budget wrong, and nobody finds out for a season. So the sheet records both, side by side: | Discount claimed | Where from | Expected | On receipt? | Actual | |---|---|---|---|---| | | | | Y / N | | If **Expected** and **Actual** disagree, that is not an error to tidy up. It is the most useful line on the page — it is the only place the property learns which sources actually pay out and which only look like they do. Over a few runs this column answers a question no invoice can: **is the cheaper source actually cheaper, once the failed discounts and the extra drive are counted?** --- ## 4. The closing check The sheet checks itself before it leaves the room. Two arithmetic closes: 1. **Every line closes.** Units bought x unit price = line total. 2. **The page closes.** All line totals + tax − actual discounts = receipt total. If the page does not close, something on it is wrong, and it is wrong *now*, while the receipt is still in your hand — not in three weeks when an invoice arrives and nobody remembers the run. A sheet that closes is not proof the buying was good. It is proof the record is true, which is the only thing a record can promise. --- ## 5. Questions to ask before the first run Answer these on your own property, in writing, before filing the first sheet. Each one has caught a wrong assumption somewhere: 1. **Is the par you are copying built on this property's own room count?** A count taken from a status report is a record of work, not a description of the building. Check it against the room list. 2. **Who counts on-hand — the person who orders, or someone else?** If they are the same person, the count and the order are one judgement, not two. 3. **Which items are actually shared between departments?** An item counted in two closets and pars set in both is bought twice. 4. **What happened on the last run that nobody wrote down?** A substitution, a store that was out, a size change. Those are the events that quietly move consumption per unit, and they never appear in an invoice. --- ## What this instrument does not do It does not price, forecast, or recommend a vendor. It records what was needed, what was bought, where from, and what it truly cost — accurately enough that those questions can be asked later from evidence rather than from memory. The sheet publishes blank. Every property's filled sheets are its own.