Here’s the problem:
Tactical approaches are the ones that will give you quick wins and win over your customer more often than not. That’s what they actually want. They’re short term and short-lived. Strategy on the other hand is an implementation of a long-term vision. So, it takes much more time but will survive the test of time. It’s a much longer term solution but unfortunately takes too much time to build.
Lets look at the “Art of War” according to Sun Tzu.
“Strategy without tactics is the slowest route to victory. Tactics without strategy is the noise before defeat”
So, a purely tactical approach is doomed to eventual failure whereas using purely strategy is slow and painful.
You really need both.
Let me say that again.
You really need both.
Enter the Data Vault 2.0 System of Business Intelligence. It provides both. You get tactical implementation guidelines which fit into the long term strategic vision. It was designed from the ground up (again) to deliver.
Before we even go into the differences, both Data Vault (now called DV 1.0) and DV 2.0 were created solely by my friend Dan Linstedt.
The difference between DV 1.0 and DV 2.0 starts with the coverage itself. DV 1.0 is restricted to a Data Warehouse Architecture. It includes a modelling paradigm and a methodology. The modelling paradigm, or DV 1.0 modeling was actually put in to the public domain by Dan. What this means is anyone can pick it up, use it, build trainings on it and even build derivative works out of it. You can too. But, only with the model parts. It’s a good practice to provide attribution if you do though.
The implementation methodology even for 1.0 is still copyrighted by Dan (This is pretty standard copyright law in terms of producing IP by the way. It was developed in 2003 by Dan and therefore he retains copyright. So, to use and propagate the methodology, and/or to build any derivative works, you would actually need his permission). So, even for any DV 1.0 architecture based solutions, you’d be in good company with the inventor.
Data Vault 2.0 is actually termed as a System of Business Intelligence. It’s no longer just a cog in the BI landscape, but designed for you to run your entire BI program using it. It encompasses many things and the DV 2.0 model has some subtle changes for parallel architectures.
But, now you have an established and proven method to design an entire BI program end-to-end from sourcing to delivery and everything in-between.
Actually, instead of me saying it, lets listen to the legend who created our industry (i.e. the way we earn our bread and feed our families):
Here’s what the legendary father of Data Warehousing Bill Inmon said about it, ” Over multiple years, Dan improved the Data Vault and evolved it into Data Vault 2.0. Today this System Of Business Intelligence includes not only a more sophisticated model, but an agile methodology, a reference architecture for enterprise data warehouse systems, and best practices for implementation.”
These are sometimes referred to as pillars, but I have a better analogy. Think of a standard 4-legged table and each of these things as one of those legs with your BI program as the table top. It’s optimized and already vetted and tested.
There’s no argument that it works great.
After all, it borrows many best practices, just like the model borrows the best from the available modelling methods of the time. The success stories for DV 2.0 are amazing.
Just in case you’re interested, here’s one:
We can think of the arguments against this which I will talk about soon, such as using only the model and your own tailor made methodology etc. It can be done but won’t be optimal. More on that is forthcoming.