They should be; part of Data Architecture is defining what those reusable business rules are and if so needed by a wider should be modelled as business rule outcomes where columnar storage will not help you.
A good place to start as Kimball.

Don’t forget that even in DAMA’s DMBoK2 (data modelling knowledge are) the two main participants in producing data models are data modellers and data analysts. It’s not just me, but the professional I work in that agrees with me.