As an Analytics professional, I’m exhausted. I’m exhausted by the constant search for silver bullets (by others). And I’m even more exhausted of the notion upheld by every non-Analytics person around me that what I provide might as well come in the form of pixie dust. Product companies love to create this mystique around Analytics; making it look as difficult, magical, and opaque as possible. That helps them sell their software. However, they have all been so successful at this approach that it has made my job eternally harder. By convincing potential buyers that what they have to offer is magic, silver bullets in a black box, the software companies raise buyer expectations to unreasonably high levels. Purchasers of that software are also my clients; and they are often astounded that I can’t unbox these magical silver bullets and begin firing away like a gun slinger in an old western. As cool as that sounds, it just isn’t realistic. So, what is realistic? Well, to best answer that question, let’s start at the beginning…
As I see it, Analytics is fairly simple. It is the rigorous application of statistics, mathematics, common sense, and technology to arrive at valuable outcomes for clients. Hard work is rewarded with more hard work, and there are no short cuts! Period. And it all starts with data. Without data, you are merely performing an academic exercise, which can only be rigorously validated by…. Collecting data! Thus, Analytics in all its forms truly revolves around data. So, before one even bothers to read a pamphlet on the latest whizzy thing in the market, there are some basic questions to be answered.
1) What questions do you want to answer?
2) What data do you have to answer those questions?
3) Where are the data gaps, and what are your options for filling them?
Answering these questions can sometimes take longer than one might expect as there is real work to be done at each step–often Analytics professionals help organizations work through these steps. However, it is only after these questions have been effectively answered that an organization has the information necessary to make sound decisions regarding Analytics (e.g., Choosing the right software, etc.). And that is when the hard work begins.
Analytics professionals begin the task of understanding data through exploratory analysis and experimentation. Part of this work includes cleaning and preparing the data for analysis. Based on the results of this process, predictive models can be developed, improved, and automated. Whether the goal is to predict the next word in a sentence, risk of default on a loan, the next location of a crime, or the likelihood of fraudulent activity; the same general process still applies. No bullets, wands or pixie dust… Just blood, sweat, and … coffee. Lots of coffee.
analyticaltern 