Companies will have to define a semantic layer, no matter what. If you don’t assign experts on your data and analytics team to do this, then end users will do it for themselves in Tableau, PowerBI, Qlik, DataRobot, or whichever business intelligence (BI) or machine learning (ML) front end they are using.
On the business side, besides being a colossal waste of time, allowing BI users and data scientists to each create their own metrics and business terms creates chaos and inconsistency.
On the technical side, requiring data engineering to support numerous extracts and ETL pipelines to feed BI/ML tools is a maintenance headache and adds complexity for governance professionals.
For data and analytics leaders on both the business and technical sides, this whitepaper will explain how a semantic layer: