5 mistakes to avoid when implementing data lakes by Matt Maccaux via @infomgmt

Many organisations are making debilitating mistakes that will ultimately hinder their ability to have a scalable, elastic data-monetisation platform.

I have to agree with his comments on too much Hadoop and not enough governance.  The data lake is only going to be useful if it is efficient and contains good and proper data - the old adage about garbage in and garbage out can definitely apply to a data lake without data stewards and some kind of of quality control.