I'm back in the office today after three days of nonstop data and analytics at the SAS Global Forum and SAS Global Forum Executive Conference events in Orlando, Fla. I listened in to some great presentations, watched cool demos, and -- what I always like best -- got the chance to speak with analytics practitioners one-on-one. Now I need to decompress.
OK, done -- who, after all, can take off for more than a second in this data-crazed, results-hungry world!
I've got a ton of great insight and information to share. Unfortunately, I can't do that all at once, so I'll be spreading out highlights over the next few weeks, both here in my blog as well as through video.
I want to start with insights gleaned from Scott Friesen, senior director of analytics in consumer insights at big-box retailer Best Buy. Friesen participated in an Executive Conference panel discussion on integrated marketing management, and I had the opportunity to sit down with him for a brief chat on his strategy and vision.
At Best Buy, Friesen's group handles customer analytics -- econometrics, market share measurement, predictive modeling, optimization, segmentation, and so on -- as well as operational elements such as data enhancement and tools management. He reports into the CMO, but doesn't consider customer analytics as a marketing function because his group also services the channels, the business units that roll together by product type, the merchant organization, Geek Squad services, and business-to-business operations.
He describes his group's role in this way: "We're a domestic capability and an enterprise center of excellence."
Best Buy knows its customers extremely well at a transactional level; it can assign about 60 percent of its transactions and 80 percent of revenue to a customer, Friesen says. However, he admits, identification falls off drastically when trying to associate Web behavior with a customer. And he feels that being able to make that connection is where the real opportunity lies.
"I've stopped even thinking about the transactional element. Instead I spend all my effort to understand the stuff that's not the transactions. The fact that we know that a customer bought something is, yes, fine -- but that's yesterday's insight... We have to move past the notion that we should only concern ourselves with the transaction as an important event."
Toward that end, Best Buy is in the process of setting up a new customer analytics master for the enterprise. The goal is getting a global view of all events and interactions associated with a customer, Friesen says, noting that this will require a mind shift as well as a change in root systems.
"We have to stop thinking about transactions and start thinking about events, and we also have to stop thinking about products and start thinking about items," with the latter referring to digital, ethereal things like music downloads and monthly services with variable pricing. "These need some sort of generic framework in our root systems to allow for flexibility."
The new customer analytics master, he says, will give Best Buy a better sense of events and items then it's ever had. "The world has gotten fuzzier, which means we have to be clearer, and we have to accept the lack of definition. You can't fight that reality too hard, or your systems will have no realistic representation of the real world."
Take customer IDs, for example. "Abstracting a human being and putting that representation in a database is a really fascinating challenge that gets really interesting, especially the digital age."
Consider a customer named William, who sometimes is known as Bill, Friesen says. He might sometimes use his home phone and sometimes his cellphone. Sometimes he uses his work email address and sometimes his personal email address. "We would like people to be these wonderfully consistent beings, but people aren't that consistent. So you have to start constructing your systems and the way you do your customer analytics to handle layers of certainty... as opposed to a thing that's absolutely known."
Events and items, layers of certainty -- how do notions like this fit in your organization's customer analytics vision? Share on the message board below.