Consumption rate measures how much of an item is used per unit of time. For purchasing, it is often expressed as units per day. Define whether the rate includes sales, recipe usage, internal use, transfers or losses before using it to plan replenishment.
A sales rate, observed stock depletion and a forecast of future usage are related but different numbers. Keeping those meanings explicit helps avoid counting spoilage twice or treating missing stock as customer demand.
Basic consumption-rate formula
Observed usage rate = usage quantity during the period / length of the period
Assume a fictional retailer records 240 units sold over 30 calendar days. Its observed sales rate is eight units per calendar day. Genuine zero-sales days belong in that denominator. Missing data, closed days and periods when the item was unavailable need separate interpretation.
If the business instead divides by 20 open days, the result is 12 units per open day. Neither number is interchangeable with the other. A lead-time or coverage calculation must use the corresponding calendar.
Ingredient consumption from recipes
Multiply each menu item's sales rate by the ingredient quantity used per sold item, then sum the contributions. Account for recipe units, batch yield and any explicitly modeled preparation loss.
Suppose a sandwich uses 100 grams of bread and sells 30 times per day:
Bread usage = 30 sandwiches/day × 100 grams/sandwich
= 3,000 grams/day = 3 kilograms/day
If another recipe consumes one kilogram daily, combined recipe usage is four kilograms per day. That describes theoretical usage from those recipe sales. Staff meals, measured waste, portion differences and direct bread sales need their own treatment.
“Recipe yield” can mean portions produced per batch, while ingredient quantity describes how much goes into that batch. Confirm the conversion instead of multiplying two fields whose units do not match. If the recipe already includes a preparation-loss allowance, do not add that same loss again.
Sales, counts and losses answer different questions
| Evidence | What it can establish | What to investigate |
|---|---|---|
| POS sales | Recorded customer sales | Returns, unavailable stock and unmapped products |
| Recipe mapping | Expected ingredient use from sales | Portions, units, yields and recipe changes |
| Physical counts | Actual stock at a point in time | Count timing, unit errors and missing receipts |
| Waste or internal-use record | A documented non-sales depletion | Whether already included elsewhere |
| Transfer record | Goods leaving one location for another | Source demand versus consumption at destination |
A simplified stock reconciliation is opening stock plus receipts and net transfers in, minus closing stock. The result measures depletion that still needs explanation; it does not prove that every missing unit was sold or consumed normally.
For example, start with 100 units, receive 50 and count 60 at the end. With no transfers or other adjustments, depletion is 90 units. If recorded sales are 80, ten units remain to reconcile. They might be waste, internal use, missing records or a counting error. Do not automatically label them recurring demand.
Choosing a history window
A 30-day window is one practical starting point, not a guarantee of a stable forecast. Short windows react faster but can be noisy. Longer windows include more observations but can hide recent changes. Weekly seasonality, promotions, a new item and infrequent demand each need context.
For intermittent sales, a calendar-day average including zero days remains a valid description of the observed period. SBA is another forecasting candidate that separately smooths positive demand sizes and their intervals. Compare forecasts on data not used to fit them and on the horizon relevant to buying.
How the rate affects purchasing
With constant future usage and no other depletion, 40 usable units at eight units per day give five days of cover. That is a simple projection, not a promised stockout date. Incoming quantities, reservations, variable demand and losses can change the result.
A reorder point uses demand over lead time plus the chosen buffer under its assumptions. An order-up-to target determines the desired position under a separate policy. The purchase quantity then considers usable stock, relevant incoming orders, outstanding commitments and supplier constraints. There is no universal “PAR minus on-hand plus lead-time demand” formula that fits every policy.
Use the PAR level calculator to explore its stated assumptions, then check the MOQ and pack constraints on the resulting purchase.
Consumption inputs in LineNow
LineNow's planning calculations can use mapped sales, configured manual rates and supported inventory history. The demand-statistics calculation commonly uses a 30-day window and routes sufficiently observed intermittent or lumpy histories to SBA. Counts and other configured inputs can also affect the final recommendation.
For items such as cleaning supplies that have no useful POS signal, a manual usage input can be appropriate. Demonstrate how that input interacts with mapped sales, transfers and count calibration in your setup. Check the resulting quantity instead of assuming every screen uses one unmodified daily number.
Explore LineNow's inventory replenishment workflow with an item whose current recommendation seems wrong. Trace its units, mappings, receipts, count history and open orders through to a reviewed purchase. The living purchase order then keeps supplier changes and receipts available for the next decision.