E-commerce inventory and SKUs: organize stock before it becomes a problem

Map the physical stock before choosing settings

Inventory settings should describe what actually happens to products. Start by listing where stock is held, who counts it and which systems can change the quantity. A shop shelf, a storeroom and a supplier warehouse may have different owners and update routines. If those differences are ignored, the website can show a confident stock figure that nobody can verify.

Document the order path from sale to dispatch. Identify when an item is reserved, when it is picked and when a cancellation releases it. Also record sales made outside the website, such as in-store purchases or manually confirmed orders. The store cannot maintain an accurate picture if those transactions never reach the inventory process.

Keep the first map simple enough for staff to understand. A list of locations, responsible people and stock-changing events is often enough to expose gaps. Do not buy an inventory application before identifying the problem it needs to solve. A tool may automate updates, but it cannot decide which conflicting spreadsheet reflects reality unless the business establishes a reliable source.

Distinguish products, variants, SKUs and barcodes

A product is the customer-facing item or family. A variant is a specific sellable combination, such as a shirt in a particular size and color. A SKU is an internal identifier that helps the business track that combination. A barcode is another identifier used for scanning or external product identification, depending on the system and context.

These concepts should not be collapsed into one field. Two color variants may share the same product description while requiring separate stock counts. The same variant stored in two locations is still the same sellable item, with quantities assigned to different places. Shopify's SKU guidance recommends unique SKUs for variants within the store and consistent identifiers for the same variant across locations.

Create a short naming rule the team can follow. A hypothetical code such as TEE-BLK-M might identify a medium black T-shirt, but your format should fit the catalog and existing systems. Avoid encoding a temporary price or shelf position into the identifier. Check whether suppliers, accounting tools or fulfillment partners already depend on established SKUs before changing them during a store project.

Create one approved product and stock source

Decide which system is authoritative for each kind of information. The product catalog may hold descriptions and dimensions, while an inventory system controls quantities. What matters is knowing where a change starts and how it reaches the other tools. Two systems can be connected without both being allowed to overwrite the same value unpredictably.

Prepare an opening stock count for launch or migration. Count physical items, identify damaged or reserved stock and record the date. Reconcile differences before importing quantities into the new store. Copying an old spreadsheet without checking it can reproduce an existing problem in a more polished interface. Keep the approved opening file so later discrepancies can be investigated.

Assign permissions based on responsibilities. Staff who update descriptions may not need to adjust stock counts, and a supplier may not need access to customer orders. Document who can approve a correction and how the reason is recorded. The catalog preparation guide explains how the product information and image references can remain organized alongside this stock process.

Decide how availability appears to customers

Available, out of stock, made to order and backorder are different promises. Use them deliberately. If a product is produced after purchase, a zero physical count does not necessarily mean it should be described like a ready-stock item. The customer needs to understand the production and dispatch process before placing the order.

Avoid allowing sales beyond stock by accident. If the business accepts backorders, define which items qualify, how the expected timeline is checked and who handles delays. If it does not accept them, test that unavailable variants cannot be ordered through an unexpected route. A setting should reflect a business decision rather than remain enabled because it was convenient during testing.

Keep the message consistent across the product page, cart and confirmation. If an item needs a longer preparation time, do not let a general delivery banner imply immediate dispatch. Customer service should be able to explain the same promise the store makes. Review the delivery information guide when connecting stock status to the timeline customers see.

Handle bundles without double-counting stock

Bundles create an extra inventory decision. The business may pack a bundle in advance and treat it as its own stock item, or assemble it from components after an order. Those approaches require different stock logic. Decide which one reflects the operation before selecting an app or creating product records.

Consider a hypothetical gift set containing one notebook and one pen. If the set draws from the same notebook and pen stock sold individually, selling a set needs to affect the available components. If the store only reduces a separate bundle quantity while leaving components unchanged, it may promise more notebooks than the business owns. The exact implementation depends on the platform and chosen tools, so test it rather than assume.

Record substitutions and partial availability rules. Can staff replace a component, or must the customer approve a change? What happens when one component is unavailable? Keep the page content aligned with those decisions. A bundle photograph showing fixed contents should not quietly become a promise of whatever is left in the storeroom. Accurate inventory and accurate merchandising need to work together.

Reconcile multiple channels and manual orders

List every place a customer can buy. Website, physical shop, marketplace and manually confirmed WhatsApp orders may all draw from the same stock. A single channel can appear accurate while the combined operation oversells. Decide how each sale is recorded and how quickly the shared availability is updated.

If synchronization is manual, make the process explicit. Identify the person, update frequency and exception route when an item is nearly sold out. A simple reliable manual process can be better than an unattended integration, but its limits should be understood. Do not promise real-time availability when updates happen only at the end of the day.

For connected systems, test a sale, cancellation, refund and stock adjustment. Check where each event is created and whether the receiving system handles it correctly. Refunds do not always mean an item has physically returned to sellable stock. Keep that distinction clear. Ask the integration provider how errors are reported so a failed update does not remain invisible until a customer orders an unavailable item.

For example, if a shop assistant confirms a manual order, record it in the agreed system before another channel promises the last unit. A message thread alone should not become a second unofficial stock ledger. Give staff one clearly defined place to check availability and one route for reporting uncertainty.

Use stock checks and exception records

Choose a counting routine that fits the range and sales pattern. Frequently sold or easily confused items may need closer attention than slow-moving products. The goal is to detect differences early enough to investigate them. A single annual count may reveal a discrepancy without showing when or why it happened.

Record adjustments with a reason. Damage, loss, supplier shortage, counting error and returned stock are different events. A note that simply says fixed quantity does not help the team improve the process. Keep evidence proportionate to the business, but make the record understandable to someone who was not present when the correction happened.

Review recurring exceptions. If one product is repeatedly wrong, inspect its variant setup, bundle relationships, packaging and picking labels. If an entire location is unreliable, the issue may be an update process rather than individual product data. Avoid changing quantities repeatedly without addressing the source. Clear exception records turn an inventory problem into a practical investigation rather than a cycle of unexplained corrections.

Test and hand over the operating routine

Before launch, run a small set of representative orders. Include a simple item, two variants of the same product, a bundle and an item close to selling out. Verify the customer-facing availability and the stock records after each event. Then test a cancellation and a return decision so staff understand what changes automatically and what needs approval.

Train the people who will maintain the process using real tasks. Show how to find an SKU, inspect a variant, record a correction and identify an integration problem. Give them a short procedure for an uncertain stock situation. Customers should receive a clear answer while staff investigate, rather than a promise based on an unverified screen.

Keep the process updated when the business adds a location or sales channel. An inventory model that worked for one shop may need revision when wholesale or a marketplace is introduced. BrandNerds can help scope catalog and store operations around those changes. The best inventory setup is one the team can explain, check and maintain, with a clear connection between physical stock and what customers are allowed to order.