Key takeaways
- Use the same stable product ID across product and local inventory data and an exact, case-sensitive store code that matches Business Profile.
- Publish price, availability, quantity, pickup method, and pickup service level from operational data with known freshness.
- Reconcile feed, landing page, structured data, checkout, and physical shelf; the strictest or most current truth should protect the customer.
- Track false in-stock and false out-of-stock experiences by item, store, source, and lag.
01
A feed can be valid while the promise is false
The file uploads, every required column exists, and Merchant Center accepts the item. A customer drives to the store because Search says it is available. The shelf is empty and staff cannot find a pickup reservation. Technical validity did not establish physical truth.
Model product ID, store code, GTIN where applicable, physical quantity, availability state, store price and currency, sale dates, pickup method, pickup service level, local delivery, source system, observed time, submitted time, processed time, and last verified shelf state. Store and product joins must be stable before freshness can help.
Swipe to compare every column
| Signal | Required relationship | Failure |
|---|---|---|
| Product ID | Same item across product and inventory data | Inventory attaches to the wrong offer |
| Store code | Exact match with the linked profile | Stock cannot attach to a location |
| Availability or quantity | Matches purchasable shelf state | Customer travels for absent stock |
| Pickup SLA | Matches page, checkout, and operations | Order cannot be ready as promised |
02
Define availability from what a customer can actually buy
Google’s local inventory specification requires accurate availability and exact store-code matching. When both quantity and availability are submitted, classification can follow the lower signal. Build conservative buffers for shrinkage, damaged stock, display-only items, holds, and orders not yet deducted from the POS.
Price must match what the customer can pay in that store, with regional tax treatment and effective dates handled correctly. Pickup and local-delivery claims need matching landing-page and checkout behavior, not just feed attributes.
03
Measure the age of truth through the pipeline
Record event time at POS, extraction, transformation, feed submission, processing, serving, and customer action. A daily file may satisfy a routine while remaining too stale for fast-moving inventory. Google recommends frequent updates and at least daily refresh in relevant verification guidance; the business may need more frequent control.
Test late files, partial store outages, duplicated IDs, case-mismatched store codes, sale expiry, returns, zero quantity, delayed orders, and a newly opened or closed location. Quarantine implausible changes rather than publishing a whole store as fully stocked after a reset.
04
Audit from the result back to the shelf
Sample customer-visible offers and trace each one through landing page, structured data, feed, POS, pickup flow, and physical stock. Also sample products that are physically available but absent online; false out-of-stock hides legitimate local demand.
Monitor feed age, unmatched stores and products, price conflicts, false in-stock, false out-of-stock, pickup failure, cancellations, substitutions, and customer service contacts. Local inventory visibility earns trust only when the trip or pickup succeeds.
Primary sources and further reading
Use the source material to validate details against your own context and current platform configuration.
- Google Merchant Center: Local inventory data specification
- Google Merchant Center: About inventory verification
- Google Merchant Center: Fix missing inventory data
This field note follows the XenGrowth editorial policy: primary sources where available, visible limitations, material review dates, and no invented first-hand experience.
Stay with the problem
