Welcome to the age of customer analytics. The enormous volume and diversity of customer data has given rise to demands for data-driven business decision-making. More and more businesses are leveraging customer analytics to find effective ways of engaging customers on an individual level.
But companies may not have the resources to analyze their data. Enter the University of Pennsylvania's Wharton Customer Analytics Initiative (WCAI). Recognizing the growing demand for this field of analytics, the WCAI was founded five years ago to foster the development and application of customer analytic methods. Additionally, the WCAI was a recent recipient of the non-profit Marketing EDGE's Education award.
1to1 Media spoke with WCAI Executive Director Colleen O'Neill and Co-Academic Director Eric Bradlow about the WCAI's role in helping businesses find customer insights and the Internet of Things' potential impact on customer analytics. Below is a lightly edited transcript of the conversation. For more, check out a podcast of the interview on SoundCloud.
1to1 Media: What is your organization's mission as a research center and how do you define customer analytics?
Colleen O'Neill: We define customer analytics as people doing things over time. At a basic level, we look to marry the researchers with business problems companies are trying to solve and the data assets that they have [with which] to solve them. We use a crowdsource model to allow researchers to opt in or reply to get access to those company data assets to solve business problems for those companies.
Eric Bradlow: A lot of people use the term "business analytics" and we're focused on a subset of business analytics which is data at the granular or individual level. In the field of marketing, people have wondered when we spend on advertising, how much more do we end up selling? What happens when we change our price? That's analytics, it's not customer analytics. Technology has changed the face of marketing where you can target individual customers, compute a customer's lifetime value, and run experiments at the individual level.
Where do you get the data that you make available to researchers?
CO: We talk with companies about their business problems and what data assets they have that they can put forward toward solving them and then we typically receive a sample of their data assets. These are anonymized so we don't receive any personally identifiable information. And internally we organize those and conduct our own data audits, produce a data key, and make the data from the company as turnkey as possible for academic researchers.
EB: What I would add to what Colleen said is that almost every company that collects data has a CRM system where they're collecting data at the level of the individual customer. And unless you're a Google or Facebook that has hundreds of data scientists, everyone wishes they had this type of talent. That's what we provide-This is how we act as matchmaker and why it's win-win from both sides [business clients and researchers].
Is there any overlap in your role as a professor and what you do with the WCAI?
EB: It's obviously not a surprise that I would start a center on analytics since that's where my own area of research lies. I'm a researcher too. For example, there's a recent data set from a corporate partner that I wanted to use for my own research. So just like everyone else, I had to submit a proposal, I talked to the corporate partner about what I wanted to use the data for and so the nice thing is, I have the same rights to access the data like other researchers and I have a pretty good sense that my proposals will get accepted since I co-direct the center. It's been tremendous because I have learned a lot about the new kinds of datasets that are out there. I'm building equity with companies for the products of the future. Definitely, there are synergistic effects between my own research and WCAI.
Who are some of the companies that have worked with WCAI and what type of research did they look to do?
CO: We do all kinds of types of research. Our most recent webinars were about datasets and business problems related to risk scores for analysis for a credit agency and for developing a confidence score for an individual credit score. That is a little out of our traditional niche.
We've been more marketing oriented with a great number of projects where we're looking at attribution modeling and customer lifetime value. For example, one is Bazaarvoice, the ratings and review platform. The dataset was from one of their retailers, which has allowed us to use their data. So we're not just looking at the full path to purchase but also looking at the impact of ratings and reviews, [including] sentiment analysis of the text of those reviews, and the impact of the questions and answers that you find in online ratings and reviews. Do the numerical ratings match up with the sentiment of the text? And how all of that may or may not impact an individual's purchasing behavior.
What do you think the Internet of Things' impact will be on customer analytics as more touch points emerge?
EB: I always say it's not about more rows, it's about more columns. Over the last five to 10 years we've been able to track people for a while through loyalty card data or an IP address or cookie, etc. The part that's changed is we get to better measure things about people to really understand who they are and the kind of products they want.
And that's what the IoT is going to do. Now that I'm wearing a wearable watch, now that my alarm system is linked to my computing. All of a sudden everything is measurable. That's not adding more people, it's adding more columns. When that happens you can predict what people are going to do better and since I'm in the business of building mathematical models for new data structures it keeps me in business.
What advice would you give to companies on how to best leverage data to gain actionable insights?
EB: Get the data scientist out of the back room into the front room. For years, data scientists have been thought of as doing the analysis you do in the back. And now, it's not about analytics or business, it's about analytics and business.
We're strong believers that the best trained people to be in the C-Suite of the future will be people who know business strategy and analytics. We're trying to layer analytics on top of business education. I think that will put our students in a unique position. WCAI's view has always been analysis for a business reason. Framing the business problem, understanding data assets, understanding how those things interact is as important a skill set as knowing how to write code like R, Phython, or some other programming language. Framing problems and understanding the kinds of data you need to collect to solve those problems- those skills will never go away.