Editor’s note: This article was originally written by Mark Harrington in 2013 and has been updated in April 2026 to reflect the latest developments in digital marketing and media.
-
- Tension: Companies hoarded unprecedented volumes of customer data yet grew more disconnected from the people behind the numbers.
- Noise: The analytics industry convinced marketers that more data points would automatically translate into deeper customer understanding.
- Direct Message: Knowing someone requires interpretation and empathy, and no spreadsheet has ever contained either one.
To learn more about our editorial approach, explore The Direct Message methodology.
Imagine you’re standing at a crowded farmers’ market in San Francisco on a Saturday morning. You’ve been watching the same woman visit the same olive oil vendor for six consecutive weekends.
You know her purchase frequency. You know her average spend. You know she arrives between 9:15 and 9:40 a.m., usually wearing running shoes, which suggests she jogs beforehand. You’ve cataloged all of this.
Now she walks past you on the sidewalk, and you have no idea who she is. You couldn’t tell her what kind of week she’s had, whether she’s buying that olive oil for herself or for her mother, or whether she’d be delighted or disturbed to learn how much you’ve tracked. This is, in miniature, the state of modern marketing.
The industry spent more than a decade building the most sophisticated surveillance apparatus in commercial history, and the result was a strange paradox: companies that could describe their customers in granular statistical detail but couldn’t truly describe them at all.
The Widening Distance Between Data and Recognition
I grew up in a small town in Oregon where the nearest mall was two hours away. The people who ran local shops knew their customers by name, by story, by the way someone’s face changed after a hard winter. There was no CRM platform involved. There was no segmentation model. There was presence, attention, and a willingness to listen. When I later spent six years as a growth strategist at a Fortune 500 tech company, I watched the industry move in the opposite direction with breathtaking confidence. The belief was straightforward: if you collected enough behavioral signals, you could predict what people wanted before they knew it themselves.
The data poured in. Click paths. Scroll depth. Cart abandonment rates. Email open times down to the second. Location pings. App session durations. And yet, something troubling emerged in quarterly reviews and campaign post-mortems. The campaigns built on this data often missed. They felt tone-deaf. They targeted people with products they’d already bought or with messages that seemed to come from a company that had never actually met them.
Research from Baylor University revealed one of the reasons: marketers consistently project their own preferences onto customers, a cognitive distortion known as the false consensus effect. The study found that even highly trained marketing professionals assumed their target audiences thought, felt, and valued what they themselves did. The data was supposed to correct for this bias. Instead, marketers often used data selectively to confirm the assumptions they’d already made. The numbers became a mirror rather than a window.
This is the gap that rarely gets discussed in industry conference keynotes. The tension was never between “data” and “no data.” It was between accumulation and comprehension. Companies got extraordinarily good at the former while neglecting the latter. They built vast digital profiles of individuals and still couldn’t answer the simplest question a corner shop owner in rural Oregon could answer intuitively: What does this person actually care about right now?
The Analytics Industry’s Loudest Promise
For years, the dominant narrative in marketing technology was seductive in its simplicity: invest in data infrastructure, and understanding will follow. Vendors sold platforms on the premise that integration alone would produce insight. Connect your email platform to your CRM, your CRM to your ad network, your ad network to your content management system, and suddenly you’d have a “360-degree view of the customer.”
This belief was often framed in sweeping terms. As Mark Harrington, Chief Marketing Officer at ListenLogic, wrote at the time:
“Having the ability to visualize millions of consumers based on their needs, attitudes, actions, and experiences delivers multidimensional insight to drive critical marketing components, ranging from promotions to product innovation. Marketers can gain deep understanding of what prospects and customers want, need, like, and dislike without ever asking a question. And they can do this on a continual basis to track markets shift in the always-on world.”
It’s a compelling vision. It’s also where the distortion begins.
What I’ve found analyzing consumer behavior data sets over the past several years is that the most common failure mode looks the same across industries. A company collects millions of data points. It builds segments. It automates messaging. And then it watches engagement plateau or decline, because the segments are statistical clusters, not portraits of living people. The customers in “Segment 4B: High-Value Repeat Purchasers, Age 35-44, West Coast” might share demographic coordinates, but they don’t share motivations, anxieties, or the particular circumstances of their Tuesday afternoon.
The oversimplification was seductive precisely because it was measurable. You could put a number on data volume. You could chart the growth of your customer database. You could present dashboards to leadership that showed rising collection rates and expanding profiles. What you couldn’t easily measure was whether any of it translated into genuine recognition of the people you served. And in the metrics-driven cultures of most marketing departments, what can’t be measured tends to get ignored.
The Insight That Changes the Equation
There is a question at the center of this entire problem that most organizations never stop to ask:
If your customer walked into your office today and sat down across from you, could you have a real conversation with them based on what your data tells you? Or would you be reciting statistics to a stranger? The companies that learn to interpret data through the lens of human context will be the ones that finally close the gap between collection and connection.
This reframe matters because it shifts the standard of success. The goal of customer data was never supposed to be completeness. It was supposed to be clarity. And clarity requires something that no platform can automate: the willingness to ask what the numbers mean in the context of a specific human life.
Building Recognition Into the System
I keep a journal of marketing campaigns that failed spectacularly. I call it my “anti-playbook.” The entries share a common thread. Almost every spectacular failure involved a team that had abundant data but lacked a framework for translating that data into empathy. One campaign targeted new parents with luxury vacation packages three weeks after their due dates. The data said they were high earners with travel history. The data said nothing about sleep deprivation, shifting priorities, or the emotional complexity of new parenthood. The campaign flopped. The unsubscribe rate was brutal.
The companies starting to get this right are approaching data differently. They’re layering qualitative research on top of quantitative signals. They’re investing in ethnographic studies, customer interviews, and community listening. They’re asking their analysts to spend time in customer service departments, reading support tickets, hearing the actual language people use when something matters to them. Emerging market analysis increasingly emphasizes this blend of predictive modeling with on-the-ground cultural understanding, recognizing that statistical patterns only become actionable when they’re interpreted through human context.
In behavioral psychology, there’s a well-established distinction between “thin” and “thick” data. Thin data tells you what happened. Thick data tells you why it happened and what it felt like. Most marketing departments are drowning in thin data while starving for thick data. The correction doesn’t require abandoning analytics. It requires surrounding analytics with the kind of interpretive intelligence that turns pattern recognition into person recognition.
What this looks like in practice is less glamorous than a new AI dashboard. It looks like a product manager who reads fifty customer reviews before building a feature brief. It looks like a brand strategist who spends a Saturday at the kind of store where their customers actually shop. It looks like leadership that rewards qualitative insight with the same enthusiasm it reserves for conversion rate improvements. I learned the hard way, years into my career, that data without empathy creates products nobody wants. That lesson cost real money and real trust before it sank in.
The path forward is simple to describe and difficult to execute: treat data as the beginning of a conversation, not the conclusion of one. The numbers can tell you where to look. They cannot tell you what you’re seeing. That part still requires the oldest technology in marketing, which is genuine curiosity about the person on the other end of the transaction.
