The Supply Chain Numbers That Actually Matter
Most supply chain advice starts with frameworks. This starts with the arithmetic those frameworks are quietly built on.
Figures and typical ranges described here reflect US manufacturing, retail and distribution practice.
If you searched "supply chain numbers that matter," here is the direct answer: reorder point, safety stock, carrying cost, lead time variability and fill rate are the five figures that most inventory and procurement decisions ultimately reduce to. Everything else — supplier scorecards, software selection, resilience strategy — is built on top of these.
Demand during lead time is the number underneath everything else
Before any formula makes sense, you need one number: how much you'll sell (or use) between placing an order and receiving it. If your supplier takes 14 days and you sell 20 units a day on average, demand during lead time is 280 units. This single figure is the foundation of your reorder point, and it's worth calculating deliberately rather than eyeballing it — most inventory shortfalls trace back to someone's rough guess about lead-time demand being wrong in one direction.
Reorder point: when to place the next order
Reorder point is demand during lead time plus safety stock. It answers one question: at what inventory level do I need to place my next order so I don't run out before it arrives? Get this number right and stockouts become rare without inventory piling up. Get it wrong in one direction and you're expediting emergency shipments; get it wrong in the other and cash sits on a shelf. The reorder point calculator on this site runs the full calculation on your own demand and lead-time figures.
Safety stock: the buffer for the bad case, not the average case
Safety stock exists because demand and lead times don't run at their average every single time. The simplest usable method sizes it for the worst case you've actually observed: peak demand times peak lead time, minus average demand times average lead time. It's less statistically rigorous than a service-level formula using standard deviations, but it's transparent, and transparency matters when you have to explain the number to someone else.
Carrying cost: what holding inventory actually costs you
Carrying cost is the annual cost of holding inventory, expressed as a percentage of its average value — typically the sum of capital cost, storage and handling, and risk (shrinkage, obsolescence, damage, insurance). Combined, it usually lands between 20% and 30% a year, though categories with fast-moving or perishable goods run higher. This is the number that makes "just order a year's supply to get the bulk discount" a much worse idea than it looks on the purchase order alone.
Lead time variability: the risk that reorder point alone doesn't capture
A supplier that takes 14 days on average but sometimes takes 25 is a different risk than one that takes a reliable 16 days every time, even though the average might be similar. Reorder point calculations that only use the average lead time understate the risk from a supplier whose lead time swings widely. If a supplier's reliability is inconsistent, it's often worth tracking their actual delivery dates for a few months before trusting a single average number in your planning.
Fill rate: the number your customers actually experience
Fill rate — the percentage of demand met from stock on hand, without a backorder or a stockout — is the metric that translates your inventory decisions into customer experience. A 95% fill rate sounds good until you calculate that it means roughly one in twenty orders is short, which compounds badly across a large customer base. Most of the KPI guide on this site comes back to fill rate because it's the one number a non-supply-chain manager immediately understands.
Where these numbers stop being enough
These five figures cover single-item, relatively stable-demand situations well. They get harder to apply cleanly with highly seasonal demand, new products with no sales history, or components shared across many finished goods. In those cases, the formulas are still the starting point — you just need to segment your inventory (by value, by criticality, by demand pattern) before applying them, rather than using one blanket set of assumptions across everything you carry.
How this connects to the rest of the site
Each of the other guides on this site builds on one of these numbers: the reorder point guide goes deeper into the worked example, the carrying cost guide unpacks what the percentage actually includes, and the KPI guide explains which of these numbers to put in front of a manager versus which to keep for your own planning. If you only read one guide on this site, this is the one worth returning to.
A quick sanity check before you trust any of these numbers
Before building a reorder point or safety stock figure into your ordering system, sanity-check the inputs against reality: does your "average daily demand" actually reflect the last quarter, or is it an outdated figure from a spreadsheet nobody's updated in a year? Stale inputs are a more common cause of bad inventory outcomes than the formula being wrong — the math is simple; keeping the inputs current is the actual discipline.
How seasonality complicates the average-demand figure
A single annual average-demand number hides seasonal swings that matter enormously for reorder timing. A product that sells at double its annual average rate for eight weeks around a seasonal peak needs a reorder point calculated on that peak-season demand rate during the relevant window, not on the smoothed annual average, which would leave you badly understocked exactly when demand is highest.
Documenting your assumptions where others can see them
Writing down the demand and lead-time assumptions behind each figure — not just the resulting numbers — makes them auditable when someone questions a reorder decision months later, and makes it far easier to spot which specific assumption went stale when actual results diverge from expectations. A number without its underlying assumption attached is much harder to debug when it turns out to be wrong.
General information for supply chain and procurement decisions, not consulting advice — your industry, scale and specific contracts may change what applies.