About This Episode
In Season 3, Episode 2 of the Data Management Podcast, I joined the hosts for a wide-ranging conversation about data warehousing methodologies — from the classical approaches of Kimball and Inmon to the modern highly normalized methods (Data Vault and Anchor Modeling) and the emerging Data Mesh paradigm.
What We Discussed
The Classical Approaches — We started with the foundations: Kimball’s dimensional modeling (star schemas, bus architecture, conformed dimensions) and Inmon’s corporate information factory (3NF enterprise model with dependent data marts). How they differ, where each works best, and why the “Kimball vs. Inmon” debate still matters.
The Highly Normalized Wave — The conversation moved to Data Vault and Anchor Modeling — what I call “HNHM” (Highly Normalized Hybrid Model). We discussed why these approaches emerged, what problems they solve that classical methods don’t, and the practical trade-offs of extreme normalization.
Data Mesh Enters the Chat — The final part of the conversation touched on Data Mesh and how it changes the modeling conversation entirely. When ownership is decentralized, does the choice of modeling methodology still matter? (Spoiler: yes, but differently.)
The Human Factor — A recurring theme was that methodology choice is as much about people as technology. Team expertise, organizational culture, and hiring market all influence which approach will succeed. The “best” methodology is the one your team can actually execute.
Why Listen
This podcast format allowed for a more conversational, nuanced exploration of topics that conference talks compress into 30-minute slots. If you’re evaluating DWH methodologies for your organization, this episode provides context and perspective that helps frame the decision.
