Analytics In Action: Building a Customer-Centric and Data-Driven Culture
By Professor Eric T. Bradlow (@ebradlow) (GBK Collective and the Wharton School)
Whether an academic, marketing professional, or business leader, we all (should) know the importance of putting customers at the center of our brand strategy, and the application of data analytics to better understand and predict customer behavior. And yet, the reality is it’s not easy. It takes a company-wide commitment and data-driven culture to collect the necessary data, predict and respond to the customers' near and future needs, and to utilize it for strategic decision making.
Last month, I had the pleasure of co-moderating “Analytics in Action: It’s Still All About the Customer”, an event co-hosted by Wharton Customer Analytics and Teradata where we explored the topic of how to build a customer-centric and data-driven culture. The event included fireside chats with my long time business partner, GBK Collective President Jeremy Korst and Teradata CMO Martyn Etherington.
In this two-part blog series, I’ll share some of the key points that came out of the event including what goes into creating a customer-centric culture, and how to analyze and act on available data to create change.
Why understanding cross-person customer heterogeneity is no longer enough
As Vice-Dean of Analytics at Wharton, I’ve spent the past 26 years of my career on the application of data and analytics to better understand customer heterogeneity to predict individual-level customer behavior. In fact, the core of my work as a Bayesian statistician focuses on understanding the differences between customers, their preferences and price sensitivities and then optimally combining data on those differences to model and predict specific outcomes.
“With real-time data, we can go beyond cross-customer heterogeneity and recognize that each customer may have different ‘personas’ (and hence sensitivities) depending on the context.”
Traditionally the study of customer heterogeneity has been defined by your classic segmentation approach - with each customer segment targeted in different ways across each of the 4 Ps of marketing (product, price, placement and promotion) depending on their preferences. But I’m here to tell you that this approach is no longer enough.
In today’s world - where we can measure customer behavior or interactions in real time - we need to yield better data and insights to understand how an individual’s preferences change depending on the context. With real-time data, we can also go beyond cross-customer heterogeneity and recognize that each customer may have different “personas” (and hence sensitivities) depending on the context.
The future of modern marketing science is about understanding within-person heterogeneity
As marketers, we’re all familiar with the concept of personas representing different customer segments. The area marketers are less familiar with is what I call “within-person heterogeneity.” I'll use myself as an example.
It is true I'm a Bayesian statistician. I'm also a Professor at the Wharton School. I also happened to be a partner at GBK. I happen to be a father of three. I happen to be a husband. I'm also a huge sports fan with a weekly podcast called Wharton Moneyball, where I talk about sports and statistics with my co-hosts. So depending on the role I’m playing at the time or the specific context, my reaction to the four P’s of marketing might be very different.
For the past 25 years, the focus of marketing data science has primarily been understanding how I’m different from another customer. But what about the concept of there isn't just one me? Sometimes there are multiple “me’s” depending on the context.
Based on my research, I've now become a firm believer that understanding the heterogeneous “me” (the different ways a customer interacts with a brand depending on context) is every bit as important as understanding the heterogeneity of different customers or segments. In other words, sometimes I want certain things from products and services, and what I want another time could be wildly different depending on the specific context. Modern marketing science and richer data helps us understand this.
Define your brand strategy based on target customers, not all customers
Now that we’ve covered this conceptually, let’s talk about “where the rubber meets the road”. What do brands need to do to become customer centric?
As my partner Jeremy Korst shared in our fireside chat: “The most successful brands embrace the fact that every customer is different, with continually evolving tastes and preferences. Even more important is they define their brand strategy around the needs of their target customers or priority segments, rather than all customers.”
Based on his experience working in the C-Suite of major companies including T-Mobile and Microsoft, Jeremy also shared his experience of what it takes to create a more customer centric culture, where the company’s approach is truly customer led.
Example #1: T-Mobile’s move from value player to industry leader as the uncarrier
In the first example, Jeremy shared how T-Mobile pivoted to become the fastest growing mobile provider in the U.S. with the introduction of its uncarrier brand strategy.
“On the surface, eliminating contracts didn't seem to make sense. Well, don’t mistake a loyal customer for a hostage.”
“Today T-Mobile is arguably among the most-recognized brands in the world. But there was a long stretch where the company suffered – with consecutive quarters of net customer base declines. In fact, during the time I worked at T-Mobile as VP and General Manager of the Emerging Devices Group, the company went through four different CEOs, and in 2011, was almost acquired by AT&T.
The reason T-Mobile struggled in the years leading up to 2013 is quite frankly, the market continued to shift – focusing more on network quality and experience, rather than a value message. With the rapid rollout of 4G networks and the introduction of the iPhone, which wasn't initially available on T-Mobile, customers were easily lured away.
Things began to change in 2012 when the company rolled out its 4G network. Then in January 2013, T-Mobile shocked the wireless industry by eliminating contracts altogether, becoming the ‘uncarrier’. On the surface, eliminating contracts doesn't seem to make sense. After all, contracts are a proven factor in improving loyalty and increasing CLV. Well, don’t mistake a loyal customer for a hostage.
With customers leaving, we knew the price and value message around the functional benefits of products wasn’t working. So the team at T-Mobile dug into the data and conducted research to understand why consumers were switching. What would it take to appeal to high value customers and segments? Number one, effectively competing and retaining customers required major investments in network and product. Secondly, high value customers were inherently in long term contracts, but hated it.”
By understanding the biggest consumer pain points – those things that people complained about the most – and then systematically solving for those gaps in the marketplace, T-Mobile defined its brand strategy from the outside-in. Removing contracts was the first in a series of ‘uncarrier moves’ to differentiate its brand in the market and become the fastest growing mobile provider in the U.S., accelerated even further through its recent merger with Sprint.
“Defining a brand strategy from the ‘outside in’ starts with understanding your target customers or priority segments, and then applying insights and data to predict their future needs.”
Taking an "Outside In" versus “Inside Out” Approach
What does it mean to define a brand strategy from the “outside in”? As Jeremy pointed out, it starts with defining your brand strategy around the needs of their target customers or priority segments, and then applying customer insights and data to better understand and predict those customers’ needs.
Conversely, when a brand takes an “inside out” approach, it is essentially telling the world that they believe they have all the answers, or they know what customers want better than they do themselves. Not only that, companies that fail to apply customer centric thinking operate as if everyone should be their customers, which is patently implausible.
Another way to look at these two divergent schools of thought is on one hand there is the “learn-it-all culture” which is how “outside in” practitioners approach brand strategy. On the other hand you have the “know-it-all” culture which, as the name implies, believes it doesn’t matter what their customers say or the data show - they know it all.
Ultimately you need a commitment across the organization and investment in an external, customer-driven brand strategy. What does a day in the life of your target customer look like? What are their current and future needs? How does their individual heterogeneity vary depending on the context of where they are and what they're doing in the moment?
Example #2: How Microsoft redefined its brand strategy from the outside-in with Windows 10
A classic example of an ‘inside-out’ approach was the launch of Windows 8 in 2012. Despite Microsoft having invested millions of dollars into R&D and creating what they considered to be hundreds of innovative features, the platform failed to gain traction. Less than two weeks after its launch, Inc. magazine openly questioned if Windows 8 was the “epic fail of the decade.” In our next example, Jeremy shares the story behind Microsoft’s pivot to turn things around with the launch of Windows 10:
"Following my time at T-Mobile, I joined Microsoft as General Manager of the Windows & Devices marketing group, responsible for the launch and worldwide marketing of Windows 10. The Windows brand had been around for nearly 30 years and was used by billions of people worldwide on a daily basis. It was a very strong brand, but with Windows 8 failing to gain traction, Microsoft was at a critical, existential juncture where they needed to reinvent Windows and rethink their approach to the market.
“With the launch of Windows 10, Microsoft focused on over-serving the needs of its target customers. The result was the fastest adopted version of Windows ever.”
The problem with Windows 8 was Microsoft tried to market it to everyone instead of trying to build a differentiated experience that appealed to any specific customer segment. When you're dealing with hundreds of millions or even billions of customers, understanding individual heterogeneity or even cohort heterogeneity by segment is a hard concept.
With Windows 8, Microsoft found itself in this place where you don't want to be as a brand. Essentially they had developed a product for everyone and therefore it wasn’t unique or differentiated. In other words, what we found is ‘Windows for everyone is Windows for no one’.
We spent a lot of time conducting research and talking to actual users and found out that there were in fact seven groups of customers that we believed were most valuable for the Windows brand. We built a product based on those customers' needs and made some hard trade-off decisions. We knew we couldn’t serve everyone, so focused on over-serving the needs of our target customers. The result was Windows 10, which went on to become the fastest adopted version of Windows ever.”
Better data, not big data
To wrap up the event, I chatted with Teradata CMO Martyn Etherington to hear more from his perspective about what goes into creating a custom centric and data driven culture. Martyn reiterated the importance of developing an “outside-in” mentality and data quality.
“One of the things our customers struggle with is the multitude of data silos, each having thousands, if not tens of thousands, of frequent updates,” shares Etherington. “Taking an outside-in approach starts with asking what business outcomes you’re looking to achieve first, and then finding the data and insights you need to inform those decisions.”
This is such an important point. Too often we see companies collect data for data’s sake, attempting to integrate datasets before they even understand the customer outcomes or business problem they need to solve for.
Collecting more data doesn’t necessarily lead to greater business intelligence – and in many cases can expose the brand to issues that impact customer trust. Ultimately companies need to have the right approach and governance structure in place to apply the right data and insights based on a brand strategy that’s truly customer-led.
In part two of this blog series, I discuss what goes into implementing a customer centric approach at scale with insights from Wharton Professors Katy Milkman and Kartik Hosanagar.