How to Configure Shopify Low-Stock Alerts

Start with tracked inventory and a named replenishment owner. Shopify's inventory tools show and adjust product quantities, but one global low-stock number rarely fits every catalog. Group variants by supplier lead time, order cycle, demand volatility, minimum order quantity, margin, shelf life, and service requirement. Store the approved threshold and effective date outside the workflow so an operator can explain why one SKU alerts at twenty units while another alerts at five.

Choose the quantity and location scope deliberately. A total quantity can hide that the location serving online orders is empty, while a location-only threshold can ignore transferable stock elsewhere. Define how available, committed, incoming, unavailable, reserved, and safety-stock quantities affect the decision. Exclude preorder items, intentionally discontinued variants, display stock, and products that continue selling below zero unless their owner has approved a separate rule. Keep bundle components tied to their governing component inventory rather than the merchandising bundle title alone.

Shopify's Flow examples use the inventory quantity changed trigger and compare both the current and prior product-variant inventory quantity. Build the condition so it fires when inventory moves from above the approved threshold to at or below it. This crossing pattern prevents the workflow from emailing the team again after every order while the item remains low. Record the exact comparison operators because a less-than rule and a less-than-or-equal rule fire at different moments.

Decide how the workflow should behave after restock. The variant must move back above the threshold before a later decline can create a new crossing alert. Test partial receipts, returns, cancellations, manual adjustments, transfers, and fulfillment changes that can move quantities in either direction. If several locations change during one order or transfer, ensure the workflow identifies the relevant location and does not collapse separate operating exceptions into one ambiguous message. Add a workflow version and change note to every release.

Send the message only to a monitored internal inbox, channel, or task queue. Include store, product, variant, SKU, location, current quantity, prior quantity, threshold, workflow run, event time, owner, and links to the inventory record and replenishment evidence. Do not include customer data. Name the expected next step, such as review the open purchase order, confirm supplier lead time, evaluate a transfer, or document that the item is intentionally winding down.

Separate urgency from noise. A threshold crossing for a fast-moving hero SKU with no incoming purchase order may require immediate review, while a slow-moving variant with a confirmed receipt may only need monitoring. Use distinct severity rules supported by catalog and supply evidence rather than putting every event in the same urgent channel. Deduplicate downstream tasks by variant, location, and open exception so one inventory change does not create multiple owners or conflicting orders.

Use a safe test variant or a controlled development store. Exercise quantity above the threshold, exactly at it, below it, repeated reductions while still low, a restock above it, and a second crossing. Verify the intended message appears once with correct product and location data. Review Flow run history for failed conditions and actions. Confirm that a malformed email destination or unavailable connector produces a visible workflow error instead of silently losing the exception.

Review thresholds on a fixed cadence alongside days of inventory remaining, supplier lead time, open purchase orders, transfers, backorders, and actual stockouts. Track alert delivery, acknowledgement, duplicate rate, stale unresolved alerts, and threshold changes. Do not measure success only by how many messages were sent. A good system surfaces the right crossing once, links it to usable evidence, records the decision, and closes when the condition is resolved or intentionally accepted.

An alert should open a review, not place an order automatically. Join the variant and location to open purchase orders, expected receipts, transfer options, supplier minimums, lead time, demand plan, promotions, safety stock, and current sell-through evidence. Name the source and timestamp of every input. When the data conflicts, keep the exception open and assign one owner instead of letting separate teams create a purchase order and transfer for the same shortfall.

Close the alert with a structured outcome: purchase order created, receipt already confirmed, transfer approved, threshold revised, product winding down, stock corrected, or no action with reason. Link the decision back to the Flow run and preserve the quantity at decision time. Review whether the alert arrived early enough to support the chosen action. That feedback improves the threshold and workflow without pretending the notification alone prevented a stockout or caused future availability.

When a threshold changes, require an effective date and reason and test the crossing logic again. Clear or migrate stale open alerts deliberately so the new threshold does not leave two unresolved exceptions for one variant. Report threshold quality by replenishment class, not as one catalog-wide average that hides important supplier and location differences.