At the recent CMO Club Innovation and Inspiration Summit in New York City, some of the world’s most innovative CMOs came together to share their insights on the biggest challenges facing today’s brands. In a session entitled, Can You See Each Customer in Your Analytics? I was joined on stage by Jeffrey Boorjian, VP Marketing of MetLife and Rex Briggs, Founder & CEO of Marketing Evolution, for a discussion around making sense of customer data, which we’ve distilled for you in four quick tips below.
#1: Customer Intimacy and Analytics Are Like Wet Clay
Even with the most advanced real-time data, we won’t have a flawless portrait of the consumer. Our reliance on data should to be balanced with a more consumer-oriented logic: We need analytics that are customer-centric and experience-focused, allowing us to measure messaging impact at every touchpoint.
#2: Demand Analytics That Are Smarter Than the Numbers
We must understand the real-life drivers of purchase at the intent level. This goes beyond mere “purpose” to the heart of who your customer is as an actor in the market and what types of customer experiences inspire sustained trust.
#3: The Best Data Helps You Anticipate Consumer Expectations
Consumers in the digital age are used to personalization; if your data isn’t informing your consumers’ experiences through customization, you’ll lose brand affinity. Optimize the value of your data by incorporating the ability to respond to consumer needs intuitively as much as possible into your product or services, even if it is limited to the initial customer service touchpoint.
#4: Test Your Tests and Then Some
Your data isn’t perpetual, like TV, it’s a snapshot (even when in real time). You’ll need to test and retest your strategy to find out which data actually matters in context of your messaging and how your consumer really lives outside the confines of their prescribed segment. Your consumers’ motivations may change as swiftly as their perception of your brand’s value proposition and that means you’ll have to develop a small team approach to new ideas. You’ll have to “test to learn” until your results become consistent and your interpretation of your first party data mimics the relevance of those fancy Big Data insights we all love so much.
They don’t call it “machine learning” for poetic reasons: the world’s smartest algorithms can’t escape the need for trial and error.