Inventory Quantities

Inventory quantity data returned from the /inventory endpoint is sourced from a cached sync of our warehouse inventory system and is not real-time. Quantities may be 30 minutes or more out of date relative to the live state of the warehouse.

This applies to all quantity fields within the warehouseCount object in each inventory response item.

Understanding warehouseCount

The warehouseCount object reflects the most recently synchronized snapshot of warehouse inventory. Because this data is cached and synced on a periodic schedule, it should be treated as an approximate view of availability rather than a live count.

Example response (partial):

{
  "warehouseCount": {
    "incoming": 20,
    "available": 5,
    "updatedAt": "2024-04-20T14:32:00Z"
  }
}

The updatedAt Field

Each warehouseCount object includes an updatedAt timestamp that indicates when the cached quantity was last changed by the sync process.

This field does not represent the last time a sync ran — it represents the last time a sync detected a difference and wrote a new value.

The updatedAt field is always represented in UTC.

When updatedAt changes

updatedAt is only updated when the sync process detects a delta (a difference) between the live warehouse inventory system and the cached data.

Warehouse system	Cached value	Delta detected?	updatedAt updated?
10 units	10 units	No	No
20 units	0 units	Yes	Yes

In plain terms: if the numbers already match, the timestamp stays the same. If the sync finds a discrepancy and corrects the cached value, the timestamp reflects when that correction was made.

What this means in practice

• A recent updatedAt means the sync detected and reconciled a change at that time.
• An older updatedAt means the cached quantity has been stable and consistent with the warehouse system since that timestamp — not that the data is stale.
• You should not use updatedAt alone to determine data freshness. Assume all warehouseCount values may lag up to 30 minutes or more behind real-time warehouse state.

Recommendations

Currently, it is not recommended to use this data as if it were considered real time. If your integration is sensitive to inventory accuracy (e.g., order logic, availability checks), we recommend:

• Building in a buffer when making availability decisions based on warehouseCount
• Treating quantity data as a soft signal rather than a hard guarantee