Stockout: find the shortage before changing the reorder rule

Separate a physical shortage from reserved stock, a late receipt, and a stale channel balance before deciding what to buy.

Kartik Deshpande··7 min read

A customer order needs to ship today, but the warehouse cannot supply the quantity promised. A stockout occurs when usable inventory cannot cover demand at the location and time it is needed. You can have stock elsewhere, stock reserved for another order, or a delivery on the way and still be short for this order.

Before raising the reorder quantity, identify which balance failed. Confirm the physical count, existing commitments, quality holds, and the next usable receipt. Expedite or reallocate when the shortage is real; correct the inventory record or channel setting when the goods are present but the selling system does not reflect them. Keep your current planning tool if it can show that evidence and support the correction.

Start with the unfilled order and its promised date

Choose one affected order line. Record its item identifier, unit, fulfillment location, requested quantity, promised shipment time, and remaining unfilled quantity. Include the time you extracted each balance. A carton count from yesterday and a piece count from this morning cannot explain the same event until you align their units and cutoff.

Ask the warehouse to distinguish missing goods from goods that cannot be released. A damaged batch, quality hold, or allocation to an earlier order needs a different response from an empty bin. Keep those quantities separate in the investigation, with the person authorized to release or reassign them. An analyst should not remove a hold to make a report balance.

Shopify’s inventory-state documentation distinguishes on-hand, available, committed, unavailable, and incoming quantities. On-hand includes available, committed, and unavailable units; incoming goods are not available to sell until received and moved into the available state. Use the equivalent definitions in your own system instead of treating every column labeled stock as interchangeable.

Reconstruct one shortage without subtracting orders twice

Consider this hypothetical fixture, not a customer result or a software test. A warehouse holds one hundred pieces of item A. Sixty are committed to existing orders, and fifteen are on quality hold. The new order needs forty pieces today. Another fifty pieces are due tomorrow, after today’s shipping cutoff, and there is no approved alternative warehouse.

Input at today’s cutoffPiecesTreatment in this example
Physical on-hand100Starting balance
Existing commitments60Unavailable to the new order
Quality hold15Requires authorized release
Available to the new order25100 - 60 - 15
New order due today40Compare with 25 available
Short today1540 - 25
Incoming tomorrow50Does not cover today’s promise

The shortage is fifteen pieces. Do not add tomorrow’s fifty to today’s available balance, and do not subtract the sixty committed pieces again if the exported available field already excludes them. If the new forty-piece order has already been included in commitments, return to the balance immediately before that allocation or reconstruct all demand lines from the same cutoff. Mixing those approaches counts the new order twice.

Now change one input at a time. If the quality team releases all fifteen held pieces before cutoff, forty become usable and the order can be covered under the stated assumptions. If the release happens tomorrow, today’s promise still fails. If the physical count finds five fewer pieces, available stock falls to twenty and the shortage rises to twenty. These outcomes are arithmetic checks, not authorization to change an order or a hold.

Evidence behind a shortage decision
Usable balance
Count, holds, commitments
Dated requirement
Order and shipping cutoff
Approved response
Owner and source evidence
Keep the cutoff attached to every layer so a later receipt cannot erase the earlier service failure.

Assign a cause that the records support

Follow the transaction trail before choosing a root cause. If the physical count differs from the recorded balance, inspect receipts, picks, returns, adjustments, and transfers around the cutoff. The inventory reconciliation process covers that investigation. Record an unresolved difference as unresolved rather than assigning it to warehouse error without evidence.

If the balance agrees with the count, inspect the replenishment timeline. Compare the purchase-order release, supplier confirmation, arrival, receiving, and quality-release times. The supplier can deliver on the agreed arrival date while stock becomes usable too late for the customer promise. A lead-time calculation should preserve those separate events so the corrective action goes to the right owner.

Next, compare actual requirements with the demand used for planning. A promotion, a new wholesale order, or a changed production schedule may explain the gap. Preserve the original plan as well as the later revision. Updating the forecast after the shortage and then comparing actual demand with that revised forecast hides what the planner knew when the purchase decision was made.

A channel mismatch needs its own evidence. Capture the warehouse balance and the channel’s sellable quantity with timestamps and location mappings. A stale feed, wrong item mapping, or fulfillment-location setting can stop sales even while suitable stock exists. Replacing the planning system will not establish which of those happened; reproduce the mismatch and check the connection responsible for that record.

Choose a response with an owner and a deadline

For a confirmed shortage, compare the earliest usable receipt with the order’s remaining deadline. An expedite must include transit, receiving, and any inspection delay. A transfer needs available stock at the sending location and a decision about the orders that stock was expected to cover. Moving the shortage to another customer is an allocation choice, not extra supply.

If substitution is permitted, confirm the substitute’s specification and the customer’s acceptance before changing the fulfillment plan. If partial shipment is permitted, record the shipped quantity and the remaining promise separately. Customer service needs an approved message and date; a planner’s tentative supplier estimate should remain tentative until somebody confirms it.

Keep a short exception record containing the affected order line, quantity short, usable receipt estimate, chosen response, approving owner, and next review time. Attach the count and supplier confirmation that support it. Close the exception only when the agreed outcome is recorded, whether that is fulfillment, an accepted substitution, or cancellation. A purchase order issued today does not by itself resolve an unfilled customer order.

Change the replenishment rule after the cause is known

Review the reorder point formula when the evidence shows that the trigger, usable lead time, or demand assumption no longer fits. Check supplier pack sizes and the review interval as well. A correct threshold can still produce a shortage if nobody reviews it until after the remaining stock has been consumed.

Microsoft’s Dynamics 365 safety-stock documentation describes planning against a minimum and creating planned replenishment when projected stock falls below it. That is a planning behavior, not a guarantee against stockouts. The buffer still needs a chosen service objective, credible inputs, and receipts that become usable when expected.

Track the same definition over time. For example, count order lines with some quantity unfilled at their promised shipping cutoff, divided by all order lines due in the period. Label that measure as an order-line shortage rate; it is different from the percentage of hours an item was unavailable on a storefront. Keep canceled demand visible under a stated rule, and do not infer lost revenue from unobserved demand.

Where Permute fits in a shortage review

Permute sells connected business data and custom applications. We can scope an exception review that brings permitted order, stock, receipt, and supplier records together. The first task is agreeing on item identity, units, fulfillment location, and the source of each status so two exports do not become two versions of the same requirement.

Those definitions can support a repeatable view for an operations manager: which orders lack usable stock by their promised dates, what evidence supports each gap, and who owns the next decision. Source timestamps and unresolved mappings should remain visible. Confirm the available connections, refresh schedule, and access controls for the proposed implementation before relying on it.

Explore a shortage review on permitted data

Start with a small sample of open orders and the stock records supporting their availability.

Keep allocation and purchasing decisions with their owners

Permute is not a forecasting engine and does not choose your safety-stock target, release quality holds, or approve substitutions. A report is not permission to place an order or change a customer promise. Any external action requires a supported workflow, explicit approval, and the relevant permissions. Assess those requirements separately from the reporting scope.

Investigate a stockout at the item, location, and deadline where demand went unfilled. Reconstruct usable stock before changing the replenishment rule, and keep commitments from being counted twice. Assign a corrective action that addresses the documented cause and preserves the effect on other orders. Measure the outcome with a stable definition so the next review can tell whether that action helped.

Review one shortage across your systems

Bring an affected order and its inventory and receipt records to identify the missing definitions or evidence.

Questions about investigating stockouts

Does negative inventory prove that a stockout occurred?

It proves that the recorded balance needs investigation, but it does not establish the physical situation at the shipping cutoff. Delayed receipts or incorrectly timed transactions can produce a negative record. Compare the transaction history with the count and the unfilled order before assigning a cause.

Can sales history measure all demand lost during a stockout?

Sales history records what customers bought, not every purchase they would have made if stock were available. Preserve backorders, cancellations, and unavailable periods where those records exist. Treat any estimate of unobserved demand as an estimate with stated assumptions.

Should I increase safety stock after every shortage?

Review the cause first. A larger buffer does not correct a wrong unit conversion, an unposted receipt, or a stale channel balance. Change the buffer when the demand and supply evidence supports a different planning policy, with the policy owner’s approval.

Is a spreadsheet enough for the first investigation?

Yes, if the sample is manageable and each row retains its source, unit, location, and cutoff. Preserve the original extracts and document corrections separately. Consider a recurring application when the same joins and exception checks must be repeated across more orders.