Why is creating an analytics culture so hard?

A personal experience at a great company and lessons learned as a consultant

I started my career at FedEx in the marketing analytics group. It happened completely by chance and I had little idea what was expected of me as a fresh MBA. I was an engineer with some fundamental quantitative and programming skills and was assigned a mentor who was an expert in analytics and database marketing. Within one year I was programming in JCL (Job Command Langauge to extract operational data from mainframe systems and looked extremely scary with green blinking characters on black screen and had no tolerance for the slightest formatting error), transforming and analyzing the data in SAS on Unix, putting them in a multi-dimensional cube for spinning around and then finally reporting in Excel and Powerpoint. In my second year I built the first k-means clustering based holistic customer segmentation at FedEx, built many predictive models for customer acquisition using look-alike modeling on external and internal data (with significant uplift), identified high potential customers to drive incremental sales (with huge ROI) and got to present all this to Senior VP of Marketing.

So how did this happen so quickly? Is it because I’m a genius? Although my mother likes to think that I am, I think it has everything to do with the environment I was in. I didn’t know the title then  and how sexy it was going to become but I was data scientist in making. There were tons of people I could learn from in addition to my mentor, and I quickly became one myself. We were sent to training programs starting in our first month that gave us the confidence with tools and techniques. We got external help (i.e. consulting) on very specific topics and only for handholding rather than outsourcing the entire thing. We were given the freedom to try the new and innovative, for example if there was an OLAP tool out there that we thought could be more efficient, we would test it head to head with existing solutions and see what came out on top. And most importantly we knew the senior leadership cared and truly believed in the importance of data driven decision making. It was easy to see when during the segmentation presentation, I was asked about the methodology, I repeat THE METHODOLOGY, by the marketing executives.

Now looking back I see that it was magic. Magic that could turn an inexperienced engineer into a data scientist that was right at the heart of business. We did not worry about (or really know about) the gap between business and technology because we were operating entirely in that gray area. In the process I was part of a team that built one of the world’s most sophisticated data warehouses. We desgined experiments to make sure that the company’s marketing dollars were spent wisely and the customers were happy.

I then moved on to customer marketing and finally e-commerce before I left FedEx. I spent the following 14 years as a management consultant, mostly as a Partner at Peppers & Rogers Group advising companies on strategy, analytics and customer relationship management. It was during those years that I realized how lucky I was, because creating that kind of environment is unbelievably hard. I saw it first-hand with companies struggling to create it and to maintain it.

By definition culture is engrained into the fabric of the organization. Analytics culture does not depend on the talent of a few individuals, it is multi-generational. The analytics group at FedEx was thriving when I arrived, continuing to grow stronger long after I left. As I hope you can see in my personal story, it requires many ingredients, and it’s not just about getting smart people, having the latest technology or skateboards in the building. It is the vision and leadership driving the organization relentlessly and sincerely and executing passionately.

I will follow this up with some specific recommendations for organizations wanting to create a culture of analytics. See you soon!

About Hamit Hamutçu

Co-founder at Analytics Center
Share this post:

Leave a comment

You must register to leave a comment