Krannert Walmart Data Dive at Purdue April 2-3, 2016

Since I am currently a graduate student, I was eligible to represent my school, and attend the Krannert Walmart Data Dive at Purdue this past weekend. It was a fun experience, and if given the chance, I would do it again. I learned a lot, and the experience was invaluable.

Here’s what I considered to be the top takeaways.

1. Your organization doesn’t need to buy data in order to come up with powerful and actionable insights that you can use.

The data we were given contained all kinds of info, internal and external, on every individual. With so little time with which to come up with something, my team couldn’t afford to bring in very much. We stuck to the basics, namely purchasing totals, RFM and CLV.

2. Keep it simple.

The simplest insights are going to be the most powerful. Especially when your audience are non-technical. Data mining is all about reducing the noise.

3. R is king. Coding is king.

The team that was adept in R went the farthest and the fastest. They also won. My team wrote sql and exported csv’s to use with SAS. You can probably imagine how that worked: back and forth between the two applications as we tried different approaches. An odbc connection in R would’ve been amazing. I’m going to get that all going on, very soon.

4. WalMart data is massive.

You wouldn’t believe it.

5. Presentation and being able to sell your ideas is just as important as the work you do.

This has been a hard lesson for me, as I’m an introvert and I have not ever enjoyed getting out in front of people very much. But if I am having good ideas, and I’m not ever selling them, what good are my ideas? As I’m realizing the folly of my ways, and getting out there, I’m finding that I don’t dread it as much, because I’m speaking about things I’m passionate about, anyway.

6. I have a lot more to learn.

There are always more approaches and more tools to find out about, and implement. Growth is happening at an exponential rate. We are living in an exciting time, and we’re in the most amazing field. Hearing about how 19 teams answered the same 2 questions in completely different, and yet valid ways, was pretty darn cool.