For an online retailer, inventory represents a major source of cost. Every item of inventory represents an item that has not been sold and which therefore represents unrealized gains. Moreover, the greater the inventory a retailer must hold the anticipation of filling customer orders, the greater the amount of money it must have invested in something that cannot be used for other purposes. Two tenets of inventory management are to turn over inventory as quickly as possible and to hold the minimum amount of inventory necessary to fulfill tomorrow's orders efficiently. Those two challenges are related. If lower levels of inventory are needed, then less inventory will be on hand and it will be turned over more quickly. In competing with another online retailer, however, a company will be inclined to hold inventory of as many items as exist. If a particular type of item is not in stock at one retailer, a customer will turn to a competing retailer. Holding all of the possible stock items in stock adds to inventory cost. A solution to this issue has been previously for a company to have one central, monster-sized warehouse. Centralizing inventory allows a company to hold the widest possible range of items at the lowest necessary levels. But since customers also value speed of delivery time, the "monster warehouse model" has a flaw in that it involves shipping from a location which may not be as close to a customer's location as otherwise and which therefore would be vulnerable to a competitor who can deliver faster.
These considerations highlight the importance of information about consumer demand. At the basic level, knowing what items sell helps brick-and-mortar retailers determine what items to stock. In competitive online retail, companies with good data can stock minimum sufficient inventory levels. Even more significant, a national retailer that can forecast demand for specific items by region can move from the monster warehouse model to a system of regional warehouses, decreasing shipping time without increasing inventory costs.