Application-Centric vs Data-Centric Approaches to IT

Application-centric thinking is fully entrenched in today's organizations. In short, applications are viewed by business managers as solutions to problems. This results in the organization building, buying, or renting (SaaS) new applications to solve individual business problems. Unfortunately, each application comes with its own unique data model (aka database) which must be integrated with the other applications – often via brittle ETL and/or the rekeying of data. These integrations often introduce errors and limit flexibility. In addition these applications typically have limited customization capability, which results in a variety of manual processes surrounding each application (i.e. partial automation with significant gaps).

A data-centric approach is one where all application functionality is based on a single data model (aka single database). Business managers view data and information as the critical resource, and an application is nothing more than a means to manipulating that data. For this to work, applications will have to be easily and rapidly created on top of the single data model and traditional application logic will need to be pushed down into the data layer (so that apps do not conflict). This requires not only new thinking, but also new tools that enable the approach. Most integration headaches are eliminated before they occur. The data model and applications are specific to the business processes being automated, which results in far fewer manual processes, greater flexibility, and lower total overall costs.