Disruptor brands are built on breakthrough insights about their consumers. Liquid Death saw the tongue-in-cheek appeal of making water edgy. Hoka homed in on making comfort cool. As marketers, we’re constantly trying to figure out the secret sauce behind the next Stripe or Figma.
Audience research can deliver insights that can help unknown brands turn entire industries upside down, but it’s often relegated to measuring KPIs that don’t budge. Traditionally, these approaches ask our customers a lot of questions. But are we really giving them a chance to speak their minds?
When you look at your organization’s insights, are you listening to the voice of the customer or the customer’s voice?
From Metrics to Meaning
Here’s the key difference between the two: Voice of the customer research asks about the metrics that you and your organization have already determined. Meanwhile, customer’s voice research allows your audience to simply articulate what matters to them. Then, it turns their priorities into metrics you can use to evaluate your brand’s performance.
Take NPS questions and 5-point scales, for example — two fixtures of voice of the customer research. They both stem from a brand’s perspective, forcing customers to answer questions that the brand has already decided is important — whether that’s recommendation, or satisfaction, or brand fit. And while these methods provide useful metrics, we all know they only tell part of the story. Think of it this way: if you had delicious food at a restaurant, but the waiter was rude, what might you rate it on a scale of 0 to 10? It’s hard to say without the full story.
New opportunities for brand growth, improvement, and innovation lie in the nuances of customers’ experiences. The combination of AI and advanced analytics now allows us to crunch vast amounts of unstructured data, offering the ability to map customers’ minds from their stories, online comments, and other sources of unsolicited feedback. That’s how you can uncover the true customer voice and gain insights beyond the limits of voice of the customer research.
We’ve found three ways to get this level of breakthrough insights, each offering varying levels of depth and their own tradeoffs. Bronze pulls from data loose-ends you already own. Silver explores data in the wild. Finally, Gold creates new data to extract unique insights – ones your competitors won’t have.
Bronze Standard
Existing Data Assets
Bronze-tier insights come from data you already possess—support calls and chats, and past research archives.
We used this approach to analyze open ends from a gaming company’s brand tracker. The results were still valuable, but without Narrative Intelligence follow-up probes, we couldn’t go as deep or uncover as many surprises.
This is bronze because: You’re limited by whether the past data is robust enough to meet your current needs, or whether the right questions were even asked in the first place.
Silver Standard
Mining the Digital Landscape
Silver-tier insights come from pulling narratives from the wild—social media, review sites, forums, and other digital conversations happening in real-time.
We did this for a telecom company that needed a fast read of their competitive landscape. By analyzing organic conversations, we identified which messages and positioning were owned by competitors. More importantly, we found how those messages were being interpreted by customers and prospects in the market.
This is silver because: Organic consumer conversations are unbiased by the “brand priority” language and questions found in traditional surveys. The conversations are authentic and unsolicited, but you can’t control the topics people discuss. You get what you get. Without a specific research question in mind, you may struggle to draw solid conclusions from consumer discussions peripheral to it.
Gold Standard
Purpose-Built Narrative Intelligence
The gold standard is collecting insights using Narrative Intelligence surveys with projective techniques. This multi-stage approach uses virtual agents to engage in real conversations with consumers, prompting them with open-ended questions that elicit rich stories. Then, analysts work with AI to translate these nuanced answers into familiar metrics and models informed by deeper, unmistakably human truths.
Unlike social listening, Narrative Intelligence puts you in the driver’s seat. You set the agenda, control the topic, the depth of the conversation, and the quality of the probing.
We took this approach recently for a major retailer who wanted to test aisle layout options for their bedding section. Using Narrative Intelligence, we uncovered surprises that traditional research would have missed. The things they assumed were important to customers barely registered, while factors they hadn’t even thought to ask about emerged as critical decision drivers.
Even better? We were able to explain what customers actually meant by vague terms like “style” and “calm.” Not just that customers mentioned them, but the implications those feelings had for their decision-making process.
This is gold because you get: Control over a conversational agenda that directly reflects your research goals, plus deep follow-up probing. Extracting both themes and dimensions offers limitless room to surface the unexpected.
“But Wait—Can’t Everyone Pull Themes?”
Yes. Absolutely, they can.
With today’s AI tools, almost anyone can analyze unstructured text and extract themes. You’ll get lists of topics, word frequencies, and maybe some sentiment scoring. You’ll know that 47% of people mentioned “innovation” and 32% talked about “customer service.”
But here’s what themes don’t tell you: How much does innovation actually matter to your customers? How are you performing on it compared to competitors? Is it a driver of purchase intent or just table stakes?
That’s the difference between themes and dimensions—and it’s everything.
Themes vs. Dimensions
The Critical Distinction
A theme tells you the number or percentage of people who mentioned a specific attribute, like “innovation.” It’s a frequency count. It’s what people are talking about, and how often.
A dimension gives you a performance score, competitive context, and predictive power. They could be a person’s likelihood to recommend product, or their overall evaluation of a company’s customer service. Dimensions measure not only how you’re doing on certain attributes, but also how much they matter in the big picture view.
Think back to the restaurant analogy. If you’re reading a restaurant review, themes are like knowing which ingredients appeared in the various dishes. Meanwhile, dimensions are like knowing how each dish actually scored on taste, presentation, and value, and how that influenced whether diners would return.
So How Do You Extract Dimensions from Narratives?
This is where the real magic happens, and where our methodology separates from basic theme extraction.
We ask the AI to act like multiple expert coders, predicting how each respondent would have answered a question about whether a particular dimension (or emotion or theme) applies to the brand. We do this in a two-step fashion, similar to how a skilled focus group moderator would probe for small differences in opinions, but done at scale.
The result? You get the richness of qualitative storytelling and the rigor of quantitative scoring. You move from simply hearing the voice of the customer (your agenda) to truly understanding the customer’s voice (their priorities).
Real Impact
When the Customer’s Voice Surprises You
We studied an iconic soft drink company to understand their brand positioning. They were focused on being seen as an industry leader. What emerged from listening to the customer’s voice? Nostalgia was the real emotional driver that brought people to their product—something far more powerful and ownable than generic leadership claims.
That’s the difference between asking customers to react to your priorities versus discovering theirs.
The Bottom Line
Collecting data from organic, unprompted sources like human narratives is the future of research and insights. To know what someone thinks, we need to hear their stories. To use those insights to make smarter decisions, we need to turn stories into reliable dimensions. But quality matters, too. Not all unstructured data delivers the same strategic value, and not all AI analyses get you to dimensions that drive decisions.
Whether you’re starting with gold-standard purpose-built conversations, mining silver-tier digital landscapes, or extracting value from bronze-tier existing data, the key is knowing what questions your data can actually answer—and what it can’t.
Ready to discover what your customers’ voice is really telling you? Want to move beyond themes to dimensions that drive strategy? Contact us—we’re listening.
