Inmon’s approach is the top-down one. The whole database works in a centralized manner, and no information is developed separately by a single entity. According to Inuwa:
“Inmon’s idea of the data warehouse is to construct an atomic data warehouse which will later help in the creation of any number of departmental data marts.”
Kimball’s approach is the bottom-up one. Any department of a company designs its own database freely, and those data storage can be integrated at any time, in any manner. The structure of the data warehouse is flexible and decentralized.
The main difference between those systems is in the rigidity of the structure. As Aziz noted in his studies:
“Contrary to Inmon, Kimball views data warehouse as a collection of departmental individual data marts.”
Those methods are also particularly different in terms of building and management. Inmon’s system, due to its solid structure, is harder to build, but it is easier to maintain. Kimball’s system is easier to produce because of the local approach to data storage, yet it is harder to analyze and filter for redundant information.