This document discusses the concept of late binding in data warehousing and analytics. It defines late binding as binding data elements, rules, and vocabularies together as late as practical for analytic purposes. This allows for more flexibility and agility to address new use cases compared to early binding. The document also presents an eight level model of analytic adoption in healthcare and discusses how data governance and analytic complexity increase with each level of maturity. Late binding is presented as an important design principle to enable progression through the levels of analytic maturity.