This is a companion paper to the Managing Data Lakes Spotlight we published in 2017. In that paper we discussed why data lakes need to be managed and the sorts of capabilities needed, in this paper we outline a methodology for building a business case in support of implementing suitable data lake management software.
Considering data lake management, and the costs thereof, is important for two reasons. Firstly, in its own right, as opposed to simply implementing a data lake – more likely a swamp – without any management or governance capabilities. And secondly, and probably more importantly, because there is a significant choice to be made between adopting a platform-based approach as compared to best-of-breed solutions. This is a recurring question, but it is particularly acute with respect to data lakes, because you could easily end up with six or eight different tools if you took a best-of-breed approach.
In practice, when building a business case, there are multiple issues to consider. On the one hand there are the positive benefits that are expected to accrue from your investment, typically captured in return on investment (ROI) but also in other measures such as time to value (TTV). Conversely, on the other side of a business case there is – or should be – total cost of ownership (TCO). We say: “should be”, because in our experience there are too many companies that do not approach TCO in any systematic way. Each of these issues is discussed in this paper.
Author/s: Philip Howard