Doudlah's inventory tracking suffers from a structural data gap: Gilbert's master spreadsheet conflicts with both Seller Central and Sellerize, creating a false picture of stock levels. The most dangerous manifestation was 5lb Yellow Cornmeal — the top-selling SKU — appearing as zero available in the spreadsheet while Sellerize reported 349 units and AWD held an additional 972. Without understanding why these numbers diverge, the team risks either unnecessary emergency shipments or actual stockouts.
Three distinct root causes were identified, each contributing independently to the discrepancy.
Sellerize tracks only inventory stored in or in transit to Amazon FBA fulfillment centers. Inventory held at Amazon Warehousing & Distribution (AWD) is completely invisible to Sellerize and does not appear in its exports.
During the holiday season, Amazon reduced Doudlah's FBA capacity, forcing a portion of shipments to route through AWD instead. This created a situation where significant stock (e.g., 972 units of 5lb Yellow Cornmeal) existed in AWD but registered as zero in the primary tracking tool.
Implication: Any inventory view based solely on Sellerize or FBA reports will systematically undercount total stock for any SKU with AWD holdings.
Amazon does not mark FBA shipments as "inbound" at the time of pickup or even carrier receipt. The inventory only becomes visible in Seller Central and Sellerize once it is physically received and processed at an Amazon fulfillment center.
This lag is compounded by LTL (less-than-truckload) shipping logistics: a carrier may pick up pallets from Doudlah's facility, hold them at a regional hub until a full load is assembled, and only then deliver to Amazon. This intermediate period — potentially two to three weeks — creates a window where stock is neither available for sale nor visible as inbound.
Example: A large shipment picked up December 29 was not yet reflected in any inventory view at the time of this review.
Implication: The team may believe they are nearly out of stock when four pallets are actually en route. Without a manual shipment log, there is no way to account for this gap.
5lb Yellow Cornmeal has two active SKUs — one being phased out (YCM5LB) and one designated as the primary. Inventory is split across both, but Gilbert's spreadsheet only tracked the primary SKU. The legacy SKU held approximately 481 units that were effectively invisible in the tracking sheet.
Additionally, because both SKUs share the same ASIN, Sellerize may aggregate or split them inconsistently, making reconciliation difficult.
Implication: Total available inventory for this SKU was significantly higher than any single view showed, but required cross-referencing multiple data sources to confirm.
Gilbert's spreadsheet uses a macro to pull data from Amazon reports rather than a live API connection. Each step in the export → import pipeline introduces an opportunity for data to be stale, mismatched, or missing. AWD data is not included in the standard FBA inventory report that the macro pulls from, so it was never captured regardless of the macro's accuracy.
A three-layer inventory view using Sellerize as the base, supplemented manually:
| Layer | Source | Method |
|---|---|---|
| FBA available + inbound | Sellerize export | Primary data source |
| AWD stock | Amazon AWD console | Manual column added to export |
| In-transit (pre-inbound) | Internal shipment log | Manual column, reconciled against log |
A new tab in the Doudlah inventory spreadsheet will log every outbound shipment with:
- Pickup/ship date
- Destination (FBA or AWD)
- ASINs and unit counts
- Amazon shipment number
- Status (in transit / received / reconciled)
This log enables the team to see stock that exists but hasn't yet appeared in any Amazon system.
Ship a large batch of 5lb Yellow Cornmeal to FBA to prevent a stockout, given that FBA-available units were near zero regardless of AWD holdings (AWD replenishment to FBA is on Amazon's schedule, not Doudlah's).