Voice of Customer Analysis: What Your Data Isn’t Telling You

Voice of customer analysis is the process of capturing what customers actually say, think, and feel about a product or service, then using that intelligence to inform strategy. Done well, it closes the gap between what a business assumes its customers want and what those customers are genuinely experiencing.

Most companies collect some version of this data. Very few use it well. The gap between gathering customer feedback and acting on it in a commercially meaningful way is where most voice of customer programmes quietly die.

Why Most VoC Programmes Produce Insight and No Action

I have sat in enough agency review meetings to know the pattern. A client commissions customer research, the findings are presented in a beautifully formatted deck, everyone nods, and six weeks later the marketing strategy looks exactly the same as it did before. The research becomes a line item on a project tracker rather than a catalyst for anything.

This is not a data problem. It is a process problem, and often a political one. Voice of customer analysis tends to surface things that are uncomfortable. Customers describe the product differently than the brand team does. They use language that does not match the campaign messaging. They complain about things that internal teams consider solved problems. Acting on that feedback requires someone to say, out loud, that the current approach needs to change. That is harder than it sounds in most organisations.

Early in my career, I worked with a client who had spent a considerable budget on brand tracking research. The findings were clear: customers did not associate the brand with the premium positioning the company had been investing in for three years. The research was filed. The premium positioning continued. Two years later, the brand was discounting to compete on price. The data had told them what was coming. They just did not want to hear it.

Voice of customer analysis is most useful when it is built into decision-making cycles rather than commissioned as a standalone project. If the output does not have a named owner and a defined decision attached to it, the probability of it changing anything is low.

Where the Most Useful Customer Language Actually Lives

There is a version of VoC that relies almost entirely on surveys. Net Promoter Score, customer satisfaction ratings, post-purchase questionnaires. These have their place, but they are also the most filtered version of what customers think. People respond to surveys in the context of being surveyed. They tend to moderate their language, round their scores, and give answers they think the company wants to hear.

The unfiltered version of the customer voice lives somewhere else entirely. Support tickets are one of the best sources of raw customer language available to any business. When someone is frustrated enough to contact support, they describe their problem in their own words, without a survey prompt shaping the response. Sales call recordings are equally valuable. The objections a prospect raises on a call, the questions they ask before committing, the comparisons they draw to competitors: all of that is genuine, unscripted customer language.

Review platforms are another underused source. One-star reviews in particular tend to be brutally specific. Customers who are angry enough to leave a negative review often describe their experience with a precision that no survey would capture. The same applies to community forums, social comments, and anywhere else customers talk about a category without a brand moderating the conversation.

When I was running an agency and we were pitching for new business, I made a habit of reading every public review I could find about a prospective client’s product or service before the first meeting. It was consistently the most useful thirty minutes of pre-pitch preparation. Not because the reviews told me what to promise, but because they told me what the client’s marketing was failing to address. That gap between what customers were experiencing and what the brand was saying was almost always the most productive place to start a conversation.

If you are building a VoC programme, the data sources worth prioritising are the ones where customers chose to speak, not the ones where your business asked them to. The distinction matters more than most marketers acknowledge.

For a broader view of the research methods that sit alongside VoC, the Market Research and Competitive Intelligence hub covers the full landscape, including how to sequence different research types across a planning cycle.

The Segmentation Problem Nobody Talks About

Averaging customer feedback is one of the most reliable ways to make it useless. If you take the responses from your most loyal customers, your recent churners, your occasional buyers, and your one-time purchasers and blend them into a single dataset, you will end up with a picture of a customer who does not actually exist.

The most commercially valuable VoC work I have seen consistently segments the data before drawing any conclusions. What your highest-value customers say about why they stay is a different signal from what your churned customers say about why they left. Both are important. Neither should be averaged with the other.

Forrester has written about the challenge of hidden personas in customer research, particularly the way that aggregate data can obscure the behaviour of specific customer segments that are disproportionately valuable or disproportionately at risk. The same principle applies across industries. When you treat your customer base as a single audience, you lose the signal in the noise.

There is also a temporal dimension worth considering. What customers say in the first thirty days of using a product is often very different from what they say at the twelve-month mark. Onboarding friction, initial value realisation, long-term satisfaction: these are distinct experiences that require distinct analysis. A VoC programme that only captures feedback at a single point in the customer relationship is missing most of the story.

When I was scaling an agency from around twenty people to over a hundred, we ran a client satisfaction process that, in hindsight, was too blunt. We surveyed clients annually and reported an average score. What we should have been doing was segmenting by tenure, by account size, and by the type of work we were doing for them. The clients who had been with us for three or more years had very different views about what made the relationship valuable than clients we had onboarded in the previous six months. Treating those as the same signal cost us some early warning signs we should have caught sooner.

How VoC Should Shape Marketing Messaging

There is a specific and underappreciated use case for voice of customer analysis that most marketing teams miss: using the exact language customers use to describe a problem as the foundation for campaign copy.

Brand teams and agency copywriters tend to write in polished, aspirational language. Customers tend to describe their problems in blunt, specific terms. The gap between those two registers is often where campaign performance leaks. An ad that describes a problem the way a marketing director would describe it will consistently underperform an ad that describes the same problem the way a frustrated customer would describe it at the end of a bad day.

This is not a new idea. Copyblogger has written about the importance of writing in language that resonates with readers rather than language that sounds impressive to the writer. The principle is straightforward. The execution is harder than it looks, particularly in organisations where marketing copy goes through multiple rounds of review by people who are not the target customer.

The practical application is to pull verbatim quotes from customer interviews, reviews, and support tickets and use them as raw material for messaging development. Not to copy them directly, but to test whether your current messaging reflects the same vocabulary, the same emotional register, and the same framing of the problem that your customers use. Where there is a gap, that gap is a messaging opportunity.

I have judged marketing effectiveness work at the Effie Awards, and the campaigns that consistently stand out are the ones that demonstrate a genuine understanding of what the customer is experiencing, not just what the brand wants to say. The connection between customer insight and creative output is visible in the work. You can tell when a campaign was built on real customer language and when it was built on internal assumptions. The former tends to perform. The latter tends to win internal approval and underperform in market.

Building a VoC Process That Does Not Collapse After the First Quarter

The most common failure mode for voice of customer programmes is that they start with genuine ambition and fade within a few months. The initial research gets done. A few changes get made. Then the next campaign cycle starts, the team gets busy, and VoC becomes something that will be revisited when there is more time. There is never more time.

Building a sustainable VoC process requires treating it as infrastructure rather than a project. That means identifying the data sources that will be monitored on an ongoing basis, assigning clear ownership, and connecting the outputs to specific planning milestones rather than leaving them as a general resource.

A minimal viable VoC infrastructure for a mid-sized marketing team might look like this: a monthly review of support ticket themes, a quarterly analysis of review platform data, and a rolling programme of customer interviews targeting two or three conversations per month. That is not a large commitment. But it requires someone to own it and a process for turning the findings into decisions.

The BCG research on decision-making in uncertain conditions makes a point that applies here: organisations that build structured intelligence-gathering into their operating rhythm consistently outperform those that commission research reactively. BCG’s work on strategy under uncertainty frames this as a capability question rather than a resource question. The same is true of VoC. It is not about budget. It is about whether the organisation has built the habit.

There is also a question of what you do with the findings. The most effective VoC programmes I have seen have a simple output format: a short document that captures the key themes from a given period, the commercial implication of each theme, and the decision or action it maps to. Not a hundred-page research report. A working document that a planning team can use in a meeting. The format matters because it determines whether the insight gets used or filed.

The Uncomfortable Truth About What VoC Reveals

I have a view that I have held for most of my career in marketing. If a company genuinely delighted its customers at every meaningful touchpoint, it would need to spend considerably less on marketing. A lot of what marketing budgets are used for is compensating for product or service experiences that are not good enough to generate organic advocacy. VoC analysis, done honestly, tends to confirm this.

When you systematically listen to what customers are actually saying, you often find that the problems are not primarily messaging problems. They are product problems, service problems, or pricing problems. Marketing can paper over those cracks for a while, but it cannot fix them. And a VoC programme that is designed to surface messaging improvements while carefully avoiding product feedback is not a VoC programme. It is a validation exercise.

The most valuable VoC work I have been involved in over the years has been the work that made someone in the room uncomfortable. Where the customer language contradicted the brand narrative. Where the churn data pointed to a product gap rather than a marketing gap. Where the support ticket themes revealed that a process the operations team considered fixed was still causing real friction for real customers. That discomfort is the signal. If your VoC programme is not occasionally producing it, you are probably not listening to the right data.

BCG’s broader writing on customer-centric strategy, including their work on retail banking transformation, makes the point that the organisations which use customer insight most effectively tend to be the ones willing to let it challenge internal assumptions rather than confirm them. That is a cultural point as much as a methodological one.

Copyblogger’s writing on building credibility through genuine understanding reflects the same principle from a content perspective. Authority in a market comes from demonstrating that you understand your customer’s situation better than they expect you to. That understanding has to come from somewhere. VoC is where it comes from.

If you are building out a research capability and want to see how VoC sits alongside competitive analysis, customer segmentation, and market sizing, the Market Research and Competitive Intelligence hub brings those methods together in a way that is designed for working marketers rather than research specialists.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions