Data Analytics: Why is more important than How.

13th March 2017

Data Analytics is a big topic, with buzzwords flying left and right, with software dedicated to it, whole divisions of people focused on it, and every company out there wanting to “get in on it”. So here is the thing, why is it that not many people seem too clued up on what data analytics actually means?

So like I mentioned, there is software available, depending on the size of data Excel can be used (who doesn’t love a Vlookup and a Pivot Table?), fancy graphs can be drawn, and data consolidation can be a big step in the journey of analysis. The thing that often seems to be overlooked though is: Why?

Why are we analyzing data? Why are we consolidating everything into one place? Why are we producing these fancy pie charts? With data analytics, it is very important to know the how. Do I have multiple data sources? How do I consolidate them? How do I then split the data to show what I need? Etc. Without “Why” though, we are just gathering data for the sake of it.

Let’s use a practical example to explain. I’m going to use Banking as the example as that’s where my history and experience lies, and it’s somewhere I have seen a lot data analytics gone wrong.

Take a large Retailer, with hundreds of stores nationwide. Each store has its own set of financial data that tells you how many deposits they make each, how many withdrawals they make for Floats each month. Head office then has a set of logistical data for all of the stores, showing how many times the physical cash in store is collected and taken to the bank. (Please bear in mind this is a much dumbed down simplified example for the purposes of explaining the process and mistakes made).

So someone takes the financial data of each store (let us say 200 stores) and they merge it into one master Financial Data Sheet (let us say in Excel to keep it simple). They then take the logistical data and run a Vlookup to get all of the relevant stores data into the relevant row. So then, we have something like this:

And time after time I have this be the deliverable out of an analytics team. “Here we go; we consolidated all of the data for you into one sheet!” My answer to this is “So what?” This table means nothing, unless we start to actually look at why this data might be useful, what manipulation is required to answer the why, and what do we hope to achieve by all of this?

Looking at this simple example, we can see that maybe the data can be worked with to understand the correlation between Cash In/Out and Collections. Then maybe we can understand the cost of the collections in relation to the Cash. Then maybe we can look at why certain Cash values require more or less collections. From this the “Why?” comes. Using this data we can see if the total cost of Cash Management for the entire network of stores can be reduced (and believe me when I tell you from experience that it almost inevitably can!)

So before any consolidation, VLookups, Pivots, Graphs, presentations etc. are done, the simple question of “Why?” needs to be asked, and it really is as simple as understanding what the client wants to achieve (in this case, reduction of costs). When you have the “Why?” there are then many more “How’s” that present themselves.

I ran an exercise like the example above a few years ago for one of South Africa’s largest Corporate Retailers. Using their store’s financial data, their geographic data, their services utilized per store data and some other bits and bobs along the way, I managed to reduce their costs per annum by upwards of R10m. All because I understood the “Why?” And when it came to the “How?”, I used an Excel sheet. Because Why is more important than How.

If you or anyone you work with is interested in discussing Data Analytics, or require some assistance with delivering tangible results using Analytics, then contact either myself directly at jrd.whitaker@litech.co.za or send through a general query to info@litech.co.za