KDnuggets Home » News » 2014 » Jul » Opinions, Interviews, Reports » Interview: Amy Gaskins, AVP, MetLife on Smarter Analytics through Qualitative Research ( 16:n16 )

Interview: Amy Gaskins, AVP, MetLife on Smarter Analytics through Qualitative Research


We discuss the relevance of qualitative research for customer intelligence, MetLife Infinity, and the increasing trend of behavior-based customer segmentation.



Amy GaskinsAmy Gaskins is an Assistant Vice President in MetLife’s Global Technology & Operations. She currently serves as the principal subject matter expert for data analysis and business intelligence. As a member of the SWAT Team, MetLife’s internal IT consultancy, she explores data science applications for emerging technologies across multiple lines of business worldwide. Prior to joining MetLife, Amy was a senior intelligence analyst supporting various agencies within the United States Intelligence Community and Department of Defense.

Here is the first part of my interview with her:

Anmol Rajpurohit: Q1. During your presentation at Sentiment Analysis Innovation Summit 2014 you mentioned that "knowing about your customers is not the same as knowing them". What do you mean by it?

Amy Gaskins: When I say that knowing about customers is not the same as knowing them, it’s really a statement about intentions. When we know about someone, we focus on easily quantifiable variables, such as demographic information or transactional information like purchasing behavior. This kind of information answers questions about what a customer does. What did they buy? When did they buy it? Do they have patterns of purchasing behavior? These types of data are a lot easier to digest and dashboard; that’s why people love NPS (Net Promoter Score) scores.

Qualitative ResearchThe idea of being able to ask a single question and determine customer loyalty is very seductive, but that one data point won’t give you the whole story because it can’t. It captures how a customer feels at a specific point in time after a single interaction, but is that how they feel about us all the time? I doubt it. The NPS score wraps up a lot of assumptions in a tidy package. I’m not saying companies shouldn't use it, just that it can’t be their sole measure because it still only answers a what question.

When I say that we know someone, our research is focused on variables that are harder to quantify, like motivation. Call it qualitative. We’re looking for why our customers do what they do. Why did they choose one option over another? Why did they choose us over a competitor? Why do they show certain purchasing patterns?

When we introduce qualitative research, it enables us to get past the “correlation isn't causation” statement. Quantitative research alone can’t get you to causal analysis. The only way to truly understand the root cause of behavior is to do qualitative study. Combining qualitative and quantitative research together is really powerful.

Sometimes it simply confirms what we already assume, but sometimes it yields really interesting insights, particularly about things like values. Either way, we gain a much more holistic view of the customer that gets to the heart of the why question.

AR: Q2. What is the main idea behind MetLife Infinity? Why is there a different product strategy for China as compared to US?

MetLife InfinityAG: MetLife Infinity is designed around the concept of digital legacy. The ability to tell your stories after you’re gone. The Infinity app helps ensure that our customers have a safe location to digitally store their memories and the opportunity to share them with loved ones at preset dates in the future. It works like a time capsule; you save things now so that others can view them later.

Our product strategy for China was different for a few reasons. First, Chinese heritage is more strongly rooted in ancestry than in the US. It’s not that we don’t look to the past here, but Chinese families’ legacies can span thousands of years. It was important for us to honor that in the look and feel of both the website and promotional channels. You can see the difference in how we've explained Infinity’s functionality:

US: “MetLife Infinity lets you store your photos, videos and important documents in one secure location. Share them with your loved ones now or on a pre-determined future date. Infinity ensures your digital legacy lives on, in the right hands.”

China: “With MetLife Infinity, you can share your memories and send them to the people who matter most. And when they grow older, your memories can stay alive and be passed from generation to generation.”

They both convey the same message, but the US version is direct and focuses on how the app works, while the China version is more subtle and emphasizes why it works, such as for respecting heritage.

AR: Q3. What factors are driving the move from transaction-based segmentation to behavior-based segmentation of customer base?

AG: This goes back to the “what vs. why” problem. Transactions can help segment customers into tidy groups, where you can infer how they might behave based on statistical models. We can predict with some degree of accuracy what they might do, but only within that group. What happens when we want to understand what motivates them? That’s at the heart of the why question. This is especially true when trying to better understand outliers or groups for which you have no existing data.

Behavior-based segmentation groups customers into like-mindedness, rather than like-attributed-ness.Customer Segmentation That’s when you begin to see clusters of people who may not seem similar from the outset, but actually share intrinsic motivations. If you can figure out which group feels most strongly about the idea you’re selling, the information used for transaction-based segmentation matters less. It’s difficult to master, but can give you a much more accurate way to develop targeted messaging.

Second and last part of the interview.

Related: