Fast Food Data Warehousing

I’m sure you’ve been to a fast food place and stood in line to order your food. There’s a very limited selection which usually fits on a board on top where you can prominently see it. However, after you make a selection, your order is then ready in a jiffy.

We’ll call this option #1.

Compare this to a 5-course meal at an elegant restaurant which serves the dinner over a period of time. You get even more limited choices, but the food is usually tastier … much tastier, but it takes time to both prepare and consume. It also costs about 10 times more than the fast food.

Lets call this option #2.

The common thread between the above two though is that your choices are still limited whether you’re in a hurry or whether you want sit and have a nice elegant meal. The choices are governed by the preparers and even though you do have a little bit of customization available, it’s usually fairly insignificant.

Now, how do we improve upon this situation?

What if we can separate some of the repeatable mechanical tasks such as slicing, mixing sauces and prepping grain foods etc etc from the artistic tasks such as mixing all these semi-prepped ingredients into a perfect blend on the plate to make your mouth water with an exquisite combination of tastes.

Now, we have much more choices with an a la carte menu. No it’s not as fancy as the 5-course meal or as shabby as fast food, but somewhere in the middle, however it has more choices than both in a fairly elegant manner.

Lets call this option #3.

Now, because you separated the mechanical and repeatable tasks, you can easily delegate this to someone after training them and still count on them to consistently produce what you want. It’s still not something you can send to the table without prep though (Or it will come back twice as fast as it was sent out).

If you have a chef who is really good at mixing, she can focus her time on churning out really good food, really, really quickly as opposed to doing everything himself (The mechanical tasks would slow her down considerably).

Now, if you start comparing these 3 to common Data Warehouse architectures, you roughly could equate #1 with dimensional models, #2 with a 3NF DW data models and #3 perhaps with a Data Vault model (specifically DV 2.0).

So, all 3 options are workable and usable, but the DV 2.0 will give you more options with much more predictable and repeatable results … every time!

However, in the DW world, this gets even better because of the way it is architected. The repeatable patterns give you the ability to generate things much faster than option #1 albeit with the elegance of option #2.

The natural implementation which is borrowed from “agile” makes you avoid repetitive work which makes it the obvious choice. If you haven’t seriously taken a look at it, perhaps you should now. And if you already know about the DV and haven’t looked at DV 2.0, now is the time. DV 2.0 is like DV 1.0 on performance enhancing drugs. The speed of implementation when compared to traditional sports would prompt an investigation into the unfair advantage.

Thankfully, we’re just talking about technology.

And we’re talking about a “solved” problem.

And we’re talking about a real long history of success stories.

If you haven’t reserved your seats yet, now would be a good time to get on board for the World Wide Data Vault Consortium where there’s a presenter this time so rare, we don’t yet have permission to even tell you. 

You can learn more here:

WWDVC Special Guest Presenter

Most of our presenters as well as audience return again and again for good reason. After all it is the ONLY Data Vault focused event in the world and includes all sorts of interesting presentations like customer case studies, tips and tricks, what’s new, amazing networking and so much more … all at an unbelievably low price.

You won’t want to miss out!