Navigating Data Modelling — What is the right track for your business?

4 min read
Thumbnail for Blog Post - Navigating Data Modelling — What is the right track for your business? || blog/navigating-modelling-maze/modelling-maze.png

In the ever-evolving landscape of data-driven decision-making, the strategic importance of data modeling cannot be overstated. At its core, data modeling is the exercise of simplifying complex data within possibly multiple information ecosystems of any business. Effectively, it allows a much wider audience to understand concepts which natively are very technical and complex. Data modeling is also laying the foundation upon which actionable insights and foresights are built.

However, one size does not fit all. With a plethora of methodologies such as Kimball, Inmon, and Data Vault, the high-level pros and cons of each become critical in making informed choices that align with business objectives and the unique data journey each company undertakes. As some of these concepts where introduced long before todays modern technologies, it’s also critical to know which elements are still relevant, why and how they’re best utilized.

The Imperative of Choosing Wisely

Data modeling is not just an IT concern; it’s a strategic business decision with far-reaching implications. The choice of data modeling approach can either unlock the potential of data to drive growth or become a bottleneck that hampers agility and insight. In the tech-agnostic philosophy of Tasman Analytics, we’ve carefully over many years evolved our own take by combining elements from different theories to form a best of all breeds structured process of how to tackle data modeling problems.

Regardless of how one would model their data it’s crucial to be consistent once the choice is made and the rules applied needs to be discussed, understood and agreed. Switching between methods on the fly will likely end up with chaos and anarchy with each individual stakeholder making their own interpretations and decisions. There is likely no better way for flushing a company’s trust in data out the drain. If there is something the industry can agree on is that there is usually nothing harder, or sometimes even impossible, than to rebuild decision makers lost trust in data. We’ve seen real life examples of consultancies building similar solutions to what we would’ve built, but their trust was lost somewhere a long the way so their contract was terminated swiftly.

Comparing the different approaches to data modeling: An Overview

In the realm of data modeling, there’s a variety of methodologies, each with its unique nuances, tailored to serve different business needs and operational environments. Understanding these concepts and being able to chose why, when and how to utilize them is essential for success for any business aiming to pivot to a more data driven strategy.

Starting with the Kimball’s Dimensional Modeling approach, it’s lauded for its user-friendly design, which caters directly to the needs of end-users, facilitating ease of data understanding and navigation. The model shines in speed of implementation and it’s low threshold for starting. However the flexibility and rapidness have a negative impact on the ability to define broadly agreed sources of truths and there is a high dependency on the end users to completely understand semi complex data concepts to ensure data quality.

Contrasting this is Inmon’s Corporate Information Factory, which offers a consistent data model and promotes a single version of the truth through a normalized structure. Its detailed level of data storage becomes a stronghold for intricate queries that require comprehensive data analysis. Yet, this increased level of detail introduces a degree of complexity that can result in a slower journey from data to insight.

The Data Vault methodology, especially in its updated incarnation Data Vault 2.0, brings agility and exceptional scalability to the table. It’s a resilient approach that adapts to changes in the business ecosystem with minimal restructuring. Nevertheless, it’s not without its challenges—there’s a steep learning curve associated with its implementation, and the initial setup demands a thorough understanding of business keys and relational intricacies, which can be resource-heavy

The landscape of data modeling is ever-evolving with new techniques emerging constantly. Such avant-garde methodologies can deliver sophisticated modeling capabilities but often with a tradeoff or the need for very specialized expertise.

In sum, the selection of a data modeling methodology is a strategic decision that must consider the specific context of a business’s data ecosystem and great focus should be given to determine the right fit for a business data strategy, infrastructure, and data maturity at hand — whether it be the structured approach of Inmon, the agility of Data Vault, the rapidness of Kimball, or potentially advanced capabilities of emerging approaches. Each offers distinctive benefits and presents unique challenges but when properly considered and applied they provide the groundwork for robust data strategies and transformative insights.

Real-world Applications: A Comparative Look

The proof is in the pudding, and nothing elucidates the strengths and weaknesses of these methodologies like real-world applications and examples. In coming blog posts we’ll deep dive into each of them to highlight illustrate the nuances of each method through the lens of actual business cases, highlighting successes, drawbacks, learning curves and what key features we extract to form the Tasman best practice of data modeling.


In conclusion, data modeling is an essential but complex element of a company’s data strategy. With no one-size-fits-all solution, it requires a nuanced understanding of various methodologies and a deep dive into the specific needs of your business. Tasman Analytics remains your partner in this journey, ensuring the path chosen is as unique as your business and as dynamic as the market you operate in.