A reorder is already late when the warehouse team realizes a fast-selling SKU is about to run out. The useful question is when should brands reorder before availability becomes a problem, without turning every demand spike into excess inventory. The answer is not a fixed reorder point copied from last quarter. It is a per-SKU decision based on expected demand during the time it will take to receive usable inventory.
For a product business selling across ecommerce, marketplaces, wholesale, and more than one warehouse, that decision needs to account for sales velocity, seasonality, incoming inventory, supplier lead time, and the stock policy you have chosen. Reordering at the right time protects revenue. Reordering the right amount protects working capital. You need both.
When should brands reorder inventory?
Brands should reorder when their projected inventory position is expected to fall below the stock they need to cover demand through the next replenishment cycle. In plain terms: place the order before your available and incoming inventory can no longer cover forecast demand during supplier lead time, plus an appropriate buffer.
A practical starting calculation is:
`Reorder point = forecast demand during lead time + safety stock`
That calculation is useful, but only if each input reflects operating reality. A 30-day supplier lead time is not a complete lead time if it excludes order approval, production queue time, freight booking, port delays, receiving, and the time needed to make inventory available for sale. Likewise, average daily sales are not a reliable demand input for a seasonal SKU, a recently promoted item, or a product with a growing wholesale account.
The reorder point tells you when to act. The order quantity is a separate decision. It should take account of demand through the next coverage period, incoming purchase orders, supplier minimums, case packs, warehouse capacity, and the cash you are prepared to commit.
Start with projected days of supply, not current on-hand
On-hand quantity is easy to see and easy to misuse. A SKU can look well stocked at 500 units while being only ten days from a stockout. Another can look low at 100 units but have enough supply for two months because demand is slow and an inbound shipment is arriving next week.
Days of supply puts inventory in demand context. At a basic level, divide sellable inventory by forecast daily demand. More useful still is projected days of supply: begin with on-hand stock, subtract expected sales over time, then add incoming quantities on their expected receipt dates.
Suppose a SKU has 1,200 sellable units and a per-SKU forecast of 40 units per day. It appears to have 30 days of supply. But 300 units are reserved for a wholesale order, demand is expected to rise to 55 units per day in three weeks, and the supplier needs 35 days to deliver. The 30-day answer is misleading. The buyer should be looking at the projected inventory curve, not a static count.
This is why a buying plan must show both stock risk and timing. A reorder recommendation without a stated stockout date, expected demand, and incoming inventory date leaves the operator to rebuild the reasoning in a spreadsheet.
Use usable inventory, not theoretical inventory
Only include inventory that can realistically cover demand. Exclude damaged units, stock allocated to committed orders, inventory held in the wrong warehouse if transfers are slow or costly, and inbound quantities with uncertain arrival dates. If a container is technically shipped but likely to miss its planned arrival, treating it as available stock can delay a necessary reorder.
For multi-warehouse businesses, location matters as much as total quantity. One warehouse may have weeks of coverage while another is about to stock out. A transfer can be the right move when it arrives faster or costs less than a new purchase order, but it still needs to be planned as a lead-time event.
Forecast through the full replenishment window
A reorder decision needs demand forecasted from today until new inventory is received and available to ship. That window is longer than many teams assume. It can include internal approval time, supplier confirmation, production, transit, customs where relevant, receiving, quality checks, and putaway.
Demand ESP forecasts every SKU from sales across all channels and accounts for seasonality. That matters because the demand expected over the next 45 days may be materially different from the last 45 days. Historical reporting can tell you what sold. A forecast helps determine what is likely to sell before the next replenishment arrives.
Forecasts should not be treated as promises. They are decision inputs. If a promotion is scheduled, a retailer is expanding an assortment, a channel has gone quiet, or a product is being discontinued, the operator should adjust the plan. The discipline is to document the assumption and see its effect on stock risk and purchasing cash before approving an order.
Set safety stock to match the cost of being wrong
Safety stock is the buffer between normal variation and a stockout. Too little buffer makes the business vulnerable to delayed receipts or stronger-than-expected demand. Too much turns uncertainty into excess inventory and ties up cash that could fund faster-moving products.
There is no universal safety-stock percentage. A replenishable hero SKU with volatile demand and a long overseas lead time deserves more protection than a slow seller available from a domestic supplier in a week. Margin, substitution options, supplier reliability, and the operational damage of a stockout all affect the decision.
A useful policy separates items by their role. High-velocity, high-consequence SKUs can carry a larger days-of-supply target. Predictable products with reliable suppliers can run leaner. Long-tail SKUs may be purchased less frequently or only to a supplier minimum. The point is not to apply one blanket rule to every item. It is to make the policy explicit so the reorder recommendation can apply it consistently.
Treat incoming inventory as a schedule, not a yes-or-no field
Many ordering mistakes happen because a team knows inventory is inbound but does not model when it will actually be usable. An incoming purchase order due in 20 days helps only if it is received in full, at the right location, and available before projected stock runs out.
For every inbound order, track expected quantity, receipt date, warehouse destination, and confidence in the date. When a supplier changes an ETA, rerun the buying plan immediately. A late inbound shipment can turn a previously safe SKU into an urgent order. Conversely, a confirmed early receipt may allow you to defer a purchase and preserve working capital.
The same principle applies to partial deliveries. If 2,000 units are ordered but only 500 will arrive first, the buying plan should use the 500-unit receipt as the supply event it actually is, not treat the full purchase order as a single future solution.
Turn the analysis into a purchasing decision
The operator needs a clear output: what to reorder, when to place the order, how much to buy, and what risk remains if the order is delayed. That is where inventory planning moves beyond reporting.
Spark combines sales history with on-hand and incoming inventory, supplier lead times, purchasing data, warehouse context, and inventory policies to build an actionable buying plan. It identifies stock and excess-inventory risk, then prepares draft purchase orders from the recommended actions. The operator reviews quantities, changes dates or suppliers where needed, and approves before anything is sent or committed.
That approval step matters. A system can calculate faster than a person can update a spreadsheet, but it cannot know that a supplier has warned of a production constraint or that a retailer has postponed a launch unless the operator brings that information into the decision.
For a quick check away from a desk, Spark Solo users can text Sparki to ask what needs reordering. A reply of YES creates a draft purchase order, not a committed purchase. It is a practical way to keep an urgent stock-risk review moving while preserving operator approval.
Keep reorder decisions current as conditions change
A monthly planning cycle can work for stable, slower-moving catalogs. It is often too slow for fast sellers, variable demand, or long supply chains. The right cadence depends on the SKU. Review high-velocity and at-risk items more often, while letting stable products follow a lighter rhythm.
The objective is not to create more planning meetings. It is to maintain a current view of the exceptions that require action. A live buying plan should surface a delayed inbound order, a change in projected days of supply, or a SKU that has shifted from excess risk to stockout risk before the consequence reaches the customer.
For teams that want inventory decisions inside their existing AI workflow, Spark MCP connects through OAuth to Claude, ChatGPT, and other MCP clients. It can map and validate operational data during onboarding, with nothing imported until the operator approves. That is a more current working model than legacy inventory software built around static exports and manual report interpretation.
The best reorder process is not the one with the most complicated formula. It is the one your team can trust enough to use every week: forecast demand per SKU, model real lead times and incoming supply, set a deliberate buffer, and make each purchase order a conscious trade between availability and cash.