Imagine that you hire a marketing intern, and on day one of the job, they make an odd request: “I want to know everything there is to know about this company’s customers. What can you tell me?”
How are you supposed to answer? The scope of the question defies any coherent analysis. Your marketing team may collect data about customers, but if it knew “everything,” surely it would have violated basic norms of privacy getting there.
That is the problem we face with marketing analytics. If our analytics are broken, it is because we haven’t addressed this conflict between what we want to know and what we should know.
Spending for what?
With ever vaster quantities of data, we expect, like the intern, that analytics can tell us whatever we want to know. In reality, the analytics underperform.
Duke professors Carl F. Mela and Christine Moorman recently wrote in HBR that, according to Duke’s CMO Survey, marketers plan to increase their marketing analytics budget by 198 percent in the next three years. Yet, the self-reported performance impact of analytics – 4.1 on a 7-point scale – hasn’t increased in five years.
Say the co-authors, “How can it be that firms have not seen any increase in how analytics contribute to company performance, but are nonetheless planning to increase spending so dramatically?”
An illusory solution
No surprise, the Duke professors ask a good question and point out a reasonable explanation: poor data and poor analytical talent. More specifically, they note a few issues with marketing data: a lack of integration and standardization that slows processing; “too much data and too little information” due to a lack of overlap in data and fields from various sources, and a widespread inability to distinguish cause from correlation.
They suggest that “First, rather than create data and then decide what to do with it, firms should decide what to do first, and then which data they need to do it.” Fair enough. If we forget to ask meaningful questions and instead keep collecting data in anticipation of wanting to know something, the data will have little impact on marketing performance.
“Second,” add the authors, “companies should create an integrated 360-degree view of the customer that considers every customer behavior from the time the alarm rings in the morning until they go to bed in the evening.”
Maybe that second suggestion gave you pause. While it sounds like the common marketing jargon we digest daily, it returns us to that conflict between what we want to know and what we should know.
The Truman Show
Effectively, Mela and Moorman suggest that you should try to meet the intern’s demand: to know everything there is to know about your customers. What is a customer besides a set of complex behaviors that range from breathing, pumping blood, and regenerating cells to buying goods, clicking smartphone screens, and falling in love?
The sunrise-to-sunset study of customers sounds nice but is culturally tone deaf. What customer would knowingly subject themselves to 24-hour, 360-degree laboratory observations? What sounds so useful to marketers sounds dystopian to the non-marketer.
What does “every customer behavior” even mean? Conservatively, “customer behavior” could be interpreted as interactions with the company. I go to www.buystuff.internet, I search for Stuff, and I buy it. Each step is an interaction.
What about the random Googling I do in the morning before I visit buystuff.internet and the LinkedIn articles I read afterward? If the buystuff.internet drops cookies and follows my subsequent actions, is that “customer behavior” too? If I’m a customer, isn’t anything I perform between the alarm chime and bedtime fair game?
We risk defining “customer behavior” as anything a human being does. If that is how far marketing analytics must go to perform well, then in the future of marketing, every customer will star in their own Truman Show with an audience of algorithms.
Resolving the conflict
Even if a tech company could pull off 24-hour observations, it would likely result in the same problem the Duke professors called out: too much data and too little information.
Maybe the authors don’t mean to take marketing analytics that far. Either way, marketing jargon like “customer behavior” doesn’t have obvious boundaries, but should.
I agree with the authors on at least one point: that the design of an analytics system and choice of research questions should align with a definable business goal.
It’s too easy to say, let’s collect tons of data and see what we discover. A research inquiry without questions has little hope of providing answers.
If marketing analytics are broken, it’s because we have too much data and not enough analytics in the true sense of the word. Maybe better questions about attribution, funnels, and sentiments can solve the problem and limit our research within boundaries that don’t violate consumer privacy.
If we try to track every customer behavior – if we try to tell that intern all there is to know about customers – we’re going to analyze our way off an ethical cliff. Let’s not.
Full Circle Insights delivers marketing and sales performance measurement solutions to optimize a company’s marketing mix and drive more revenue, helping revenue marketers Plan, Achieve, Optimize, and Evaluate.
Bonnie is a 5-time VP of Marketing at Genesys, Netscape, Oracle (Network Computer Inc.), Stratify, and VoiceObjects. While valuing the creative side of marketing, Bonnie’s real love is marketing operations — measuring marketing investments and determining investment optimization. Now as CEO of Full Circle Insights, Bonnie is working to help fellow marketers get the data needed to succeed. In 2000, Bonnie was named one of the “Top 20 Female Executives in Silicon Valley” by San Jose Magazine. Bonnie holds a B.A. in biology from Princeton University.