In the business world, decentralization is key to achieve economies of scale, and a well established cure to avoid diseconomies of scale. When Alfred P. Sloan took the position of CEO of General Motors (GM) in 1920, his secret sauce to push GM ahead of the leading American car manufacturer at the time, Ford, a corporation fundamentally centrally organized, was what he called "decentralization with co-ordination". Fast forward to today to read anything from leading business consultancies such as McKinsey, and you can be sure that decentralization is almost always the guiding principle to organize businesses - only given up for centralization if really necessary to gain the effects of synergy, e.g. in the areas of purchasing, legal or human resources.
When we, however, turn our gaze towards corporations partaking in what is often called "Digital Transformation", a completely different picture emerges. Here, centralization is increasingly adopted as the guiding principle of organization. Not only do these "digitally transforming" companies simply add centralized "digital organizations" to implement anything digital, but in addition, data is also centralized whenever possible in so-called data warehouses, data lakes, or lakehouses.
Hence, the same executives who are naturally embracing the lessons of Mr. Sloan and the McKinseys and live in a company composed of decentralized business domains, those same executives seem to be completely oblivious to the fact that they are, at the same time, accepting a practice of centralizing their digital organizations and also their data. What a contradiction!