Source: https://www.svs-digital.com/guides/supplier-stock-feed-sellable-quantity-rules Language: en Content updated: 2026-11-01 Description: Turn supplier or ERP stock into a controlled sellable quantity by defining product identity, reservations, safety stock, update freshness and failure rules. Inventory and supplier integrations # Supplier stock feed: how to define the quantity your store can sell A supplier file that says “12 in stock” does not automatically mean your online store should offer 12 units. Some may already be reserved, shared with another sales channel, held as a buffer or tied to a different size or warehouse. Before building a stock connection, define what the number means, when an order changes it and what the store does when the source is late. By SVS Digital · Published 01.11.2026 ## Start here For each product variant and location, define the authoritative source or an explicit aggregation rule, map identifiers unambiguously and calculate sellable quantity only from states whose meaning and freshness you can verify. Subtract only reservations and safety stock that the source quantity does not already exclude. Then set a maximum acceptable feed age and a fallback for unavailable or incomplete data. Test orders, cancellations, duplicate updates and outages before the connection can change the full catalogue. ## 1. Match the exact item, variant and location Start with identity, not quantity. Build a mapping table containing your store variant ID, internal SKU, supplier code and, where available, a stable trade-item identifier. Treat size, colour, pack size and condition as separate variants when they can carry different availability. A similar title is not a safe match, and an unknown code should enter an exception queue rather than update the nearest-looking product. GS1 defines GTIN as a way to uniquely identify trade items that may be priced, ordered or invoiced. A GTIN can support matching, but it does not replace your mapping when suppliers use their own codes or package levels. Shopify likewise separates the customer-facing product variant from the inventory item and records inventory levels per location. Your data contract should preserve those distinctions. - Store variant ID and internal SKU. - Supplier or ERP product code. - Size, colour, pack and other stock-relevant attributes. - Location or warehouse represented by the row. - Owner and resolution path for an unknown or duplicate code. [GS1: Global Trade Item Number](https://www.gs1.org/standards/id-keys/gtin)[Shopify: Manage inventory quantities and states](https://shopify.dev/docs/apps/build/orders-fulfillment/inventory-management-apps/manage-quantities-states) ## 2. Separate physical stock from sellable stock Write down the meaning of every input field. On-hand stock, available stock, incoming stock and the quantity a store may promise are different values. Shopify's inventory model, for example, separates available, committed, reserved, damaged, safety-stock and quality-control states. Do not subtract a reservation twice if the supplier's “available” field already excludes it. Use a formula your team can explain, such as sellable quantity = source quantity minus reservations not already reflected minus safety stock not already reflected. If the source provides only physical stock, identify which confirmed reservations or open orders have not yet reduced it and subtract only those. If you do not allow backorders, stop the result at zero; define a separate rule if you do. Record units of measure and rounding rules. Never mix pieces, packs and cartons without an explicit conversion. - Exact definition of the source quantity. - Reservations already included or excluded. - Safety-stock rule by product or category. - Units of measure and pack conversion. - Whether negative and unusually large values are blocked for review. [Shopify: Manage inventory quantities and states](https://shopify.dev/docs/apps/build/orders-fulfillment/inventory-management-apps/manage-quantities-states) ## Need to turn supplier stock into reliable store availability? Bring a sample feed, a few product variants and your current order flow. SVS Digital can review the identifiers, stock rules, reservations and failure paths, then scope a bounded inventory-integration pilot. [Discuss an inventory integration](https://cal.com/svsdigital/free-consultation) ## 3. Choose when orders reserve stock across every channel If the webshop, sales team and marketplace all take orders, a periodic stock copy alone cannot see demand that has not reached the source system yet. Choose the event that reserves stock. A cart reservation can hold stock unnecessarily unless it expires promptly, while reserving only at fulfilment can leave a long window for two channels to sell the same unit. Confirmation, payment approval or another business event may be appropriate, but every channel must follow a compatible rule. Microsoft documents soft reservations as a way for parallel order-taking systems to share an available-for-reservation quantity: each channel requests a reservation, the service reduces that quantity, and the soft reservation is later offset when the matching physical consumption or hard reservation is recorded. Odoo exposes several reservation timings, which reinforces that timing is a business decision. Define cancellation, expiry and failed-payment release paths as carefully as the reservation itself. - Event that creates a reservation. - System that validates availability. - Identifier linking a reservation to an order line. - Events that consume, cancel or expire it. - Policy when a channel cannot reach the reservation service. [Microsoft Learn: Inventory Visibility reservations](https://learn.microsoft.com/en-us/dynamics365/supply-chain/inventory/inventory-visibility-reservations)[Odoo: Reservation methods](https://odoo.com/documentation/19.0/applications/inventory_and_mrp/inventory/shipping_receiving/reservation_methods.html) ## 4. Set a freshness limit and an outage rule Define how old a successful update may be before the quantity is considered stale. The limit should reflect how quickly stock changes, the normal feed interval and the consequence of a wrong promise. Store both the source timestamp, when available, and the time your connection processed the record. A file received today may still contain yesterday's snapshot. Decide the fallback in advance. Depending on the product, you might keep the last verified quantity for a short period, reduce it, pause sales, show a longer lead time or require a manual check. A missing row is not automatically confirmed zero, and an incomplete file must not erase an entire catalogue. Alert a named owner with the last success, affected scope and reason. - Normal update interval and maximum accepted age. - Source timestamp and processing timestamp. - Minimum expected file size or product count. - Fallback by product risk and sales velocity. - Named recipient and escalation time for stale data. ## 5. Make changes traceable and safe to retry Log the source, product, location, old value, new value, calculation inputs, processing time and outcome for each material change. Shopify's documentation supports a reference document URI for tracing an adjustment back to the system or business document that initiated it. The exact mechanism varies by platform, but the operational principle is portable: an employee should be able to explain why availability changed. Give each import or event a stable identifier so retrying it does not apply the same change twice. Absolute snapshots and quantity deltas need different handling. Reject malformed rows before writing, isolate failed items instead of hiding them in a successful batch, and provide a reconciliation view that compares the store with the authoritative source. - Unique import, event or document reference. - Before-and-after quantity with calculation inputs. - Idempotent retry behavior. - Separate counts for updated, unchanged, rejected and missing records. - Reconciliation report for unresolved differences. [Shopify: Manage inventory quantities and states](https://shopify.dev/docs/apps/build/orders-fulfillment/inventory-management-apps/manage-quantities-states) ## 6. Pilot the failure paths, not only the happy path Start with a small group containing simple products, variants, low stock and at least one awkward mapping. Test a normal sale, two near-simultaneous orders, cancellation, return, unknown SKU, duplicate file, late file, partial file, source outage and retry. Compare every result with the source and ask the person who resolves exceptions to perform the recovery steps. No periodic integration can guarantee that an item is always available: sales can occur between updates, supplier figures can be wrong and systems can be unavailable. Faster synchronization may reduce delay but does not fix unclear ownership or inconsistent reservations. Release only after the business owner accepts the remaining risk, monitoring and manual fallback. - Accepted sellable-quantity formula for the pilot items. - All reservation and release events tested. - Stale, partial and duplicate inputs tested. - Reconciliation completed with no unexplained difference. - Named owner for alerts and manual corrections. ## Need to turn supplier stock into reliable store availability? Bring a sample feed, a few product variants and your current order flow. SVS Digital can review the identifiers, stock rules, reservations and failure paths, then scope a bounded inventory-integration pilot. [Discuss an inventory integration](https://cal.com/svsdigital/free-consultation) ## Sources and further reading - [Shopify: Manage inventory quantities and states](https://shopify.dev/docs/apps/build/orders-fulfillment/inventory-management-apps/manage-quantities-states) - [Microsoft Learn: Inventory Visibility reservations](https://learn.microsoft.com/en-us/dynamics365/supply-chain/inventory/inventory-visibility-reservations) - [GS1: Global Trade Item Number](https://www.gs1.org/standards/id-keys/gtin) - [Odoo: Reservation methods](https://odoo.com/documentation/19.0/applications/inventory_and_mrp/inventory/shipping_receiving/reservation_methods.html) ## More guides AI consulting and process reviews ## [How to map process handoffs before adding AI automation](https://www.svs-digital.com/guides/map-process-handoffs-before-ai-automation) Map owners, queues, missing information and exception paths in one real workflow before deciding where AI or ordinary automation belongs. 31.10.2026 AI adoption and team training ## [Employee rules for generative AI: what company data can you use?](https://www.svs-digital.com/guides/employee-rules-generative-ai-company-data) Create practical employee rules for approved AI tools, company data, human review and mistakes without blocking every useful experiment. 30.10.2026 AI voice agents ## [AI voice agent handoff: rules for transfer, callback and escalation](https://www.svs-digital.com/guides/human-handoff-rules-ai-voice-agent) Define when an AI voice agent should transfer a caller, create a callback task or finish the call, with practical fallbacks and tests for your team. 29.10.2026 --- SVS Digital · Dunajska cesta 5, 1000 Ljubljana, Slovenia Contact: [uros@svs-digital.com](mailto:uros@svs-digital.com) [All resources](https://www.svs-digital.com/ai)