Random Stow or the Disorganized Warehouse?
Random Stow is a warehouse putaway method driven by an algorithm developed by Coupang, an ecommerce monster and formidable Amazon competitor. For the Korean market, Coupang uses this counterintuitive warehouse management method, which stores items wherever there is open shelf space, rather than grouping similar SKUs together.
The warehouse might look disorganized. But it is organized chaos because the technology—AI and big data—groups products for faster picking in a way that a human can’t see or understand.
The result is faster picks because workers can pick multiple different items for an order in the most efficient path possible. When a worker gets a list of items to pick, the AI calculates a “Goldilocks” route. Because items are scattered randomly, there is a high statistical probability that one of the items on the list is already very close to the worker’s or robot’s current position.
In a traditional warehouse, if 100 people order a popular item, such as an iPhone charger, those are all stored in the electronics aisle, creating bottlenecks. The Random Stow AI distributes those 100 chargers across different locations. Because identical items are spread throughout the warehouse, multiple pickers can grab the same product from different locations simultaneously without bumping into each other. Sounds messy, yet pick rates with this method rival Amazonian pick cycles. It gets better.
Coupang’s AI analyzes massive datasets to see which items are often bought together—phone chargers and screen protectors, for example. It stows these like products near each other, even if they belong to different categories, using predictive ordering AI to minimize the travel distance for a single order.
Fast-selling items are stowed in prime locations closer to pick/pack stations, while slow-moving items are tucked away in far corners. Traditional warehouses leave open slots on shelves because they are reserved for specific categories.
Random Stow allows Coupang to fill open slots, storing more SKUs in the same warehouse footprint. But inches add up. The AI monitors every cubic inch of shelf space. When an inbound shipment arrives, the Random Stow system tells the robot or worker exactly which open bin is the perfect fit for that specific item’s dimensions. In the Korean market, real estate for ecommerce fulfillment is at a premium. No wasted space means smaller warehouses can handle faster and more picks per square foot.
Can the Random Stow approach be a lesson for the small urban final-mile fulfillment facilities in the United States?