Data Governance – A Bright Shiny Object

I once sat in a get-to-know-you meeting when a company started a lengthy project with a consulting firm.  We went around the room saying why we were excited about starting this project.  A number of people from the company said, "Because we're getting our data put into the cloud!"  They had no idea what that meant.  They had a lot of problems, but not being in the cloud wasn't one of them.  They had heard a buzzword, a bright shiny object, and thought they needed it.

I've spent my entire career working with data.  I've seen things come and go and then sometimes come back again.  I've seen new names being given to things we've been doing all along.  What is consistent is that there are people getting distracted by the latest buzzword in the market and think they need it.  They don't think about what it really means or how it will benefit them.  Maybe it's a form of ADD.  Maybe it's fear-of-missing-out.  Maybe it's good sales reps?  They think that if they don't grab it when people are talking about it, then they'll miss the boat.

The problem is that if they always jump at the latest buzzword, they probably are missing the boat.  Most companies don't have well designed data environments.  If that foundation isn't there, then anything put on top of it has a high risk of failure.  It's just like building a house.  Build a house without a foundation and what do you have?  At best, a shack that blows down in the first storm.

Data lake.  Data analytics.  Data science.  Machine learning.  Insert buzzword here!  Just make sure that you have your data foundation nailed down first and then you can jump into these initiatives.

Garbage in – garbage out.  That was a phrase that seemed to start in the 1960s, although a version of it was 100 years earlier than that with Charles Babbage.  Unfortunately, it is still true today.  If you have bad data, your results generated on that data will obviously be bad.  I am always baffled at how companies want to spent tons of money on things like data science initiatives when they have failed to get their data house in order first.

I've spent years working with data.  I learned good data management practices in university.  It was commonsense to do some of the things I've done.  Although, I think there are some people more geared towards orderly systems and others more geared to stuffing it all in a data lake and assuming someone will be able to do something with it.  I've had people coming to me with a data lake asking what's in it, which is hard to figure out after the fact.  If it was done right the first time, they would have a great data lake that they were using successfully.

If you have a data governance program that is governing the management of your data, you'll get to what you need.  You'll have the right people in the company making accurate decisions about the data.  You'll understand what you have and where you have it.  You'll nail down what level of data quality you need and make sure you get it.  You'll master what needs to be mastered.  Not to mention the security and privacy of your data.

It's time for many companies to take a step back.  Look at their data strategy.  Figure out where they want to go.  And then figure out if they have the data infrastructure that will help them get there.  If they don't, that needs to be the bright shiny object they latch on to.  Get your data house in order and move on to greater things from there.