Fulfillment by Amazon announced this month a new low-inventory-level fee and, separately, an AI tool to help merchants avoid that fee.
The announcements offer a glimpse of what’s coming for retail inventory management.
Low-inventory Fee
FBA’s low-inventory-level fee takes effect on April 1, 2024.
The fee affects “standard-sized products with consistently low inventory levels relative to customer demand.”
Amazon noted that these products affect FBA’s distribution capabilities and shipping costs, adding that the fee will “only apply when a product’s inventory levels relative to historical demand (known as historical days of supply) is below 28 days” for both short-term (last 30 days) and long-term (90 days) metrics.
The fees vary based on the product size and historical days of supply: 0-14 days, 14-21 days, and 21-28 days.
New sellers and new products receive exemptions for six months or a year.
To minimize or avoid the low-inventory-level fee, sellers can send additional units, ensuring the short-term historical days of supply exceed 28 days.
Minimum Inventory
FBA also announced that beginning on April 1, 2024, it would provide a new metric to help sellers maintain the proper level of inventory to both maximize sales and, seemingly, avoid the fee.
The new minimum inventory metric uses, unsurprisingly, machine learning and artificial intelligence models to forecast demand and replenishment.
Thus the metric is a recommendation to sellers for the minimum number of units they should keep in Amazon’s fulfillment centers.
Maintaining this recommended level will presumably help meet customer demand and offer faster delivery time since FBA will warehouse the items in distribution centers closer to the sellers’ likely customers.
I have seen firsthand how predictive inventory management can help a company. I worked a few years ago for a regional, omnichannel farm and ranch retailer that implemented a machine learning model for its purchasing and inventory management. The model resulted in lowering inventory by roughly $2 million. It also boosted sales because it had the right products at the right time.
The minimum-inventory-level metric will join Amazon’s other inventory monitoring indicators, including:
- Historical days of supply,
- Inventory Performance Index,
- Capacity limits,
- Restock recommendations.
Inventory AI
Taken together, Amazon’s low-inventory-level fee and minimum-inventory-level metric offer a preview of what could be coming to other ecommerce platforms and marketplaces.
Let’s consider three implications.
First, Amazon is confirming what most large merchants already know: Inventory levels impact shipping costs. SMBs can apply Amazon’s methods to consider shipping costs when they decide what to buy.
Second, the new minimum-inventory-level metric provides FBA sellers with a target that optimizes sales. Many excellent software tools do this now, but the cost can be relatively expensive for smaller brands. The new metric is evidence of available, cost-effective computing power. If Amazon can offer AI-powered sales forecasting, presumably so can Shopify and many other ecommerce platforms.
Third, look for even small sellers to improve purchasing and supply chain management, as Shopify and others roll out their own AI-driven predictive inventory tools.
These improvements likely produce higher profits from lower shipping and inventory-carry costs since businesses would know with more certainty which products could sell in the next 90 days.
In short, AI will impact ecommerce and retailing. Amazon’s model for optimal inventory is likely among the first of many.