Today when you need to prepare company information for decision making, you may well end up with a Data Warehouse project. Two things will come to mind: Always late. Always more expensive than planned. But there’s a new kid on the block that might make a Data Warehouse look old fashioned…
Let me first side track about how I manage mail. Because I think there’s a strong parallel with the way companies build (or at least try to) a Data Warehouse.
Sorting through my mail
For years and years, I have been spending lots of time arranging my mailboxes. Mail folders per customer, market segment, products, partners and so on. Initially I did this clean up job once a week, but it quickly became a daily activity, and took so much valuable time away from me actually doing my work.
But the number of emails I’m receiving now has exploded. Not only because the higher number of customers, partners and co-workers I’m dealing with, but also emails coming from different forums and notifications from social media postings. It was so time consuming that I decided to stop losing my time doing so.
I found another way to search through my emails. Search has become so powerful now that I can retrieve my mails by subject search, or any possible combination of words. I can find the information now even quicker than before. Faster and more complete, as earlier I was searching in specific folders - and not necessarily the right ones. Now the search capabilities of my email tool help me getting access to all information, wherever it is stored.
The same goes today for Data Warehouse and Data Mart projects. Wouldn’t you prefer spending your time making better decisions rather than wasting too much time building the Data Warehouse?
The good old Data Jailhouse …
Today when you need to prepare company information for decision making, you may well end up in a Data Warehouse project. You’ll need a lot of time to conceptualize it, design it, build it (or have it built). One thing is for sure: by the time it is (partially) finished, your business will have changed. And what is worse: you will be left with unanticipated questions that aren’t planned for in the structure your data warehouse builders have been sweating on.
… or the new-kid-in-town Data Lake
That’s why we’ve recently seen things happening in this space: more and more companies start giving up on building a nice and clean Data Warehouse (because they’ll never get there). They rather focus on putting an architecture in place that moves the effort away from sorting out a nice clean data warehouse, to technology that helps them search information in the – less structured – data pool. The concept is to bring all data in one place – often called a Data Lake – and give IT the responsibility to properly govern this data – including data accuracy, historic data and versioning, user access rights and segregation of duties. They also need to (help) chose a powerful tool that will give business users the capability to visualize and discover data, associate data, analyze it, predict behavior, and get the information they want right when they want it!
Data Governance
A crucial element is data governance. Since you will give more power to end users to browse through their data, it is even more important to have a solid data governance strategy in place. Otherwise you end up with anarchy. Without this governance strategy, your end users might be challenged with counter flows or swirls, and that’s not really what you want…
Do I need to chose?
Well, not really. Even if you have a Data Warehouse or Data marts in place, you can well start the Data Lake concept to get a shorter time to value. Essential is to have the right data governance strategy, to assure that your business users are using the right data, in the right way. This data governance function has become now more important than ever before.
To download the white paper ‘Evaluating governed Data Discovery ‘, push the button below.
