Making sense of a fractured analytics ecosystem
One of the things we hear repeatedly from clients is, “well, there are just too many possibilities for our analytics, where do we even start?”
This has become such a common refrain in the office that we refer to this attitude simply by its acronym: “TMO.” A not-uncommon interaction would be an email stating that, “Client X is having a TMO issue with deployment, could you please advise?”
And, to be fair, without question this is a major issue in analytics circa 2016: between the various platforms for analysis employed by analytics professionals — primarily Python and R, but there are a host of proprietary platforms as well — and the more consumer-facing alternatives such as MS Excel, Power BI, Tableau, and many, many others, where does one begin?
Well, and we say this to clients all the time, ideally you should begin at the beginning.
Does that sound pat?
It does. Because it is.
But let us explain ourselves….
One of the major challenges we, as an analytics firm, face with clients is helping organizations embrace the era of data-driven decision making we are well into.
And that’s hard, especially considering that virtually no organization or firm exists without legacy systems in place, and many — though it’s questionable if one could say most — of these systems simply do not jibe with today’s analytics offerings. Additionally, in many cases, the organizational structures of a company/institution are not structured (we often use the term “stacked”) to facilitate the integration of data analysis-generated insights. And structural changes to a company or organization happen slowly, and are often a rather painful process.
Add to this the proliferation of analytics offerings on the marketplace today and (often) a data scientist or two screaming about changes that need to be made now, now, now and, well, is it any wonder we here at Konteksto find ourselves uttering “TMO” quite often?
It’s actually quite understandable, and that’s why — yes — it’s best to begin at the beginning.
In today’s market simply having an “Excel jockey” or two isn’t really an option. A company or organization needs a workable data analytics strategy, and an organizational structure that allows for quick and easy dissemination of insights. This is more than just choosing Python or R as your flavor of choice, or deciding that Plotly rather than Tableau will be your preferred visualization provider.
It’s about a deep integration of data analysis into a company or institution, and, on a more granular level, about actually utilizing the insights generated.
So “the beginning,” both fortunately and unfortunately, is (most often) much more complex than a simple paralysis at the profusion of data analytics technologies on the market today. It cuts to the core of the “what” your company or organization is.