Demystifying Data: Models

Few things give me the chuckles more than old-school data modelers holding forth on the arcane mysteries of their craft like a secret society of theurgists.

I have a different take.  Data modeling and querying are about the simplest things to teach any business user with an analyst mentality.  I've worked over the decades with very smart people who deliver value with Excel spreadsheets despite the numerous obstacles an IT-centric data management paradigm throws in their way.

Imagine what they could accomplish if their leadership demanded that those obstacles be removed, along with the veil of secrecy and exclusion that characterizes the traditional model data team engagement?

Bringing data to the point of business is a foundational principle of my data leadership philosophy.  Over the course of my career, the most value-added services I've delivered are to frontline analysts who have to spend weeks manually compiling, integrating and transforming data into Excel spreadsheets to simply do their jobs.  Meanwhile, IT launches a six-month-plus long project to build a star schema and a set of reports and dashboards that ultimately miss the mark.

Why?  Because they're building for the wrong audience, and they don't understand the underlying business processes well enough.

So, data models…  How important and complex are they, really?

The answer is very, and…  not so much.

Having a visual of your data concepts, how they relate, and what business outcomes you're chasing is a great reference to focus your analysis and development of any data product.

Keeping it simple and away from the data modeling purists' meddling is essential.

Here's my simple model to help you get started:

That's it.  Everything you ever do to deliver data value boils down to elaborating on this simple model.  It doesn't matter if you end up building a star schema, a data vault, or an Excel datasheet.  All that matters is that you take this simple formula and map out your target product concept.

The rest is discovery, experimentation, conversation, testing, and refinement.

Might you need a technologist or two to make it more robust, scalable, and consistently fresh?  Sure, but YOU, the business consumer, own the recipe.  It only takes an hour or two to bake, frost, and decorate a cake.  The real magic comes from the chef who understands how to prepare, combine, and transform the ingredients into something delicious.

This not me dismissing the relevance of technology contributions to delivering valuable data products, as much as me encouraging people who feel like they're spending their careers chasing data ineffectively. 

You understand more than you know, and you should be in charge of your own house.  I encourage the rank and file to send this message to their leadership - and the leadership to recognize a broken system and take action to fix it.