December 2, 2016
Has ‘Data Lake’ Idea Already Dried Up?
by Paul Roma
What ever happened to all those massive, megabillion-dollar data lakes we all kept hearing about over the past few years?
Well, that was fast.
Remember all those massive, megabillion-dollar data lakes we all kept hearing about over the past few years? With the exception of the U.S. government, we’ll probably never see their likes again. Many of the large organizations that were pursuing those data lakes (not to mention countless smaller ones) have largely changed course. Why? The answer is actually not so surprising, even if this particular outcome is.
Many of the CIOs I talk to these days are no longer thinking of their insight systems (analytics tools, data lakes, etc.) as separate from the rest of the business, or the enterprise systems that support them. They’re managing these insights systems more as a portfolio of analytics systems — a true play for return on investment (ROI). As a result, large investments are broken into smaller, more agile investments. The technology organization may be shepherding 500 analytics projects rather than just five high-profile initiatives — enabling and supporting 10,000 people, for example, rather than just 100. In that environment, a massive data lake starts to make less sense, even if all those 500 projects tap into it. A data lake is just too resource-intensive.
See also: Why Exactly Does Big Data Matter?
Meanwhile, the need to tap into a large volume of data isn’t going away. In lieu of a huge, proprietary data lake, what options are there? This is where CIOs are getting creative, creating a network of smaller, more manageable data lakes, for example, supplementing their data with that provided by other organizations.