For too long we have asked people to go find their data in the warehouse, not find it in the store

Great webinar from the Eckerson Group on Data Products

The title of this post is a quote from Wayne Eckerson.

I just finished watching a great webinar from the Eckerson Group all about Data Products.

A Guide to Data Products: Everything You Need to Understand, Plan, and Implement.

You can watch a replay of the webinarover at:

https://www.airmeet.com/e/920ec780-13c2-11ef-b132-a51013087680

My key takeaways after listening:

Retail Data Stores over Warehouses and Catalogs

When we have implemented a Data Warehouse and a Data Catalog in the past, we have treated them just like a warehouse, a place to store and catalog data tables etc, which is fine for us as data professionals. We know what we are looking for and we know when we find it.

But for stakeholders and data consumers, we need to create and expose Data Products in the form of a Marketplace or a Retail store. We need to make it easier for them to find the Data Products they need and easily access those products.

Different Groups want Different Types of Products

I love the use of food as an analogy to the data space and its used a lot in this webinar.

One of the challenges in data is the variability of how people want their Data Products served.

While some people are happy to pick up a ready meal, some want the ability to make changes, to personalise their food.

Some want their burger without pickles.

Others want to replace their coke for a milkshake or a bottle of water.

Some what to be able to combine meals for their family, they need to grab burgers for the kids, Chinese for themselves, a kebab for their significant other.

Some want silver service for a special occasion, they want the food delivered to a table, in a resturant, with the right ambiance and the right background music.

Other want the food delivered to them at home.

Others want a smorgasboard where they can mix and match different foods on their plate, a little taste of what ever takes their fancy at that time.

How many food businesses do you know that deliver to every one of these use case?

I can think of very few.

The variability of product means they cannot delvier all those use cases efficiently.

They cannot make that business model work.

What makes us think as a data team we can?

What would happen if we gave Data Products SKU’s?

What would happen if every Data Product Manager in your organisation was forced to create a SKU for every data product they had in their portfolio?

And they were forced to record the cost of building and maintaining that product?

And they were forced to identify and record the business value that data product generated?

How many zombie unused Dahsboards do you you think we would have left in the organisation after that?

How many of those Jira tickets requesting “some data” would get fulfilled or declined?

Non Data Influencer Disclaimer:

I was not paid in anyway by the Eckerson Group for this post, I just found their content useful and so wanted to share it.