Retail is tough. Customers have endless choices, expectations are sky-high, and even small hiccups can send them elsewhere. But smart retailers thrive by listening carefully to their customers, and using their insights to boost satisfaction, loyalty, and revenue.
Voice of the Customer (VOC) metrics are a core part of the arsenal retailers use to better understand their customers. These metrics aren’t just nice-to-have analytics; they directly impact a retailer’s bottom line. High NPS and CSAT scores correlate strongly with increased customer loyalty, repeat purchases, and positive word-of-mouth. On the other hand, negative customer experiences can lead to churn, costly returns, and damage to brand reputation.
In short, measuring and acting on VOC insights is essential for any retailer looking to remain competitive today.
At Chattermill, it’s our mission to help organizations better understand how to approach their customer feedback and how to take effective action. Alongside our platform, the CX Intelligence Academy contains a number of resources to help retail professionals understand what goes into an effective VOC strategy. Everything that we’ll cover in this article is provided by two of our retail courses - VOC Metrics Analysis and How to Analyze Customer Feedback in Retail Organizations - giving you a low-down of both the theory and practice of how to effectively improve your north star metrics.
Why measure your VOC metrics at all?
So you want to start improving your team’s north star metrics - but why bother?
VOC metrics help retailers understand customer experiences clearly and consistently. The two primary VOC metrics retailers track are Net Promoter Score (NPS) and Customer Satisfaction Score (CSAT).
NPS measures customer loyalty by asking how likely they are to recommend your brand, whilst CSAT gauges satisfaction after a specific interaction or purchase.
More recently, Sentiment Analysis has been made available due to machine learning analysis of feedback; this metric reveals the extent to which your customers are feeling positively, negatively, or neutrally about aspects of their experience.
As Dave Ascott, Strategy Lead at Chattermill explains in the video above, there are clear benefits to monitoring and improving these metrics:
- Increased retention: loyal customers buy more often, reducing churn.
- Lower acquisition costs: positive customer experiences drive referrals and word-of-mouth marketing.
- Reduced operational costs: happier customers mean fewer returns, refunds, and less customer support overhead.
The second module of our VOC Metrics Analysis use case course is a great 101 on this.
Laying the groundwork
Tracking a metric is one thing, but having confidence that it’s accurately reflecting your customers is another. This is why it’s crucial to ensure you set up an effective feedback architecture to ensure you’re collecting accurate, relevant feedback that can lead to measurable improvements.
At Chattermill, we’ve spent the last decade working with some of the world’s most customer-centric companies to help them better understand their feedback, and have picked up a good amount of expertise on what works and what doesn’t.
Here are some essential steps we’ve found to ensure your VOC metrics deliver actionable insights:
- Gather feedback across the entire customer journey: Comprehensive feedback from each stage, not just isolated touchpoints, provides clarity and accuracy in understanding customer sentiment. For example, instead of just collecting feedback post-purchase, you might include your collection strategy to include in-app or cart abandonment touchpoints.
- Craft clear, neutral questions in your surveys: Ambiguous or biased questions yield superficial or misleading responses. Carefully designed prompts ensure reliability in your feedback. This means switching from asking “How much did you enjoy using our sizing tool?” to “How helpful was our sizing tool in choosing the right fit?”
- Enhance feedback with relevant metadata: Contextual data such as region, customer segment, and product type transform generic feedback into meaningful insights, helping you achieve targeted action.
- Ensure consistent feedback classification: Using automated tools or clearly defined tagging structures prevents inconsistent categorization, which can obscure true customer insights. This avoids ending up with with theme structures containing duplicates, eg “Shipping delay” and “Late delivery” - which masks the true significance of one issue.
- Evaluate statistical significance: Decision-making should be informed by sufficiently large datasets to avoid reacting to insignificant fluctuations or outliers. Just because you spot a large drop in your NPS, doesn’t mean you should rush to sound the alarm. If it came from just a few respondents, it’s worth waiting until you draw real conclusions.
These principles will give you a robust foundation for effective VOC analysis. For more detailed insights into each step, you can explore our VOC Metrics course available in the CX Intelligence Academy.
Thankfully, we’ve condensed all of this information into the first half of our VOC Metrics Analysis course.
Dmitry Isupov, Co-Founder of Chattermill and Dave Ascott, Strategy Lead, talk you through all of the variables to monitor to ensure you’ll be heading for success. You can head to the Academy to get the full walkthrough, but these videos include advice such as:
- You need coverage across the full journey: meaningful VOC analysis depends on collecting feedback at every touchpoint.
- The way you ask questions matters: A poorly phrased open-text prompt can lead to biased or shallow responses.
- It’s important to enrich your feedback with metadata: like region, product type, or customer status, is crucial for surfacing real insights.
- Accurate, consistent classification of free-text feedback is essential: If your tagging is manual or biased, you’ll get misleading results.
- Statistical significance matters: a dip in NPS based on 15 responses isn’t something to act on. You need a reliable volume of feedback to make good decisions.
Finding actionable insights
Now that you understand the foundations of VOC analysis, you can start to analyze your VOC data. As Dmitry explains in the video above, the key to making sense of VOC data is to approach it with three questions:
- Where did the metric change happen?
- Who was affected?
- And why?
Say your NPS just dropped. Don’t panic; start by checking the geography. Did scores fall globally, or only in a specific market? That’s your “where.” Then dig into segments. Was the drop driven by new users, returning customers, or one-off buyers? That’s your “who.”
Once you’ve narrowed that down, you get to the most important part: the “why.” This is where verbatim analysis comes in. You’re not guessing—you’re reading what real customers said. Maybe it’s broken discount codes. Or a buggy checkout. Or a confusing return process. Understanding the “why” behind a score means going beyond the score itself and digging into structured themes, especially using things like sentiment analysis and theme categorizsation (e.g. “Sizing,” “Checkout,” “Functionality”).
Once you’ve got this far, it’s important to make the move from looking at high level themes to actually reading verbatim comments, because that’s where your actionability comes from. You don’t just want to know “sizing is a problem”; you want to know exactly what your customers are saying.
A great example of approaching this ‘why’ part of the process is shown in our walkthrough course, How to Analyze Customer Feedback in Retail Organizations. Throughout this course, Tom Whitney, Chattermill’s VP of Solutions Consulting, walks through different use cases for a fictional retail organization, Chatterstyle.
In the video, Tom clearly spots a huge drop in NPS for the month of June for Chatterstyle. In a single click, he can dive into the exact data point to see summarized highlights for that date. It’s through this that he’s able to quickly understand some of the feedback behind this drop - ‘sizing inconsistencies associated with dresses that don’t fit as expected, exacerbated by the lack of free returns and accessible customer service’. Tom could also quickly dive into the ‘show feedback’ button to view verbatim comments which make up this summary too.
To get a broader view of what’s driving NPS across all time, Tom breaks down the report by categories, which immediately shows that ‘Product part’ category is having an outlying negative effect on the score overall. And by diving deeper into this, he can see that it’s ‘Zippers’ in particular that are causing the most issues.
Improving your metrics
Once you know what’s driving the issue, the next step is deciding what to do about it.
Dmitry introduces a practical framework: fix what’s broken, and scale what’s working. Not every insight deserves equal attention. Your job is to identify changes that are high-impact, feasible, and worth the effort. Start by looking at segments with low NPS or CSAT. If their feedback points to something fixable, like a recurring support issue, shipping delay, or missing product detail, it makes sense to prioritize. Especially if the segment is large.
At the same time, look at what’s working. Which segments are happiest? What are they praising? These strengths can often be amplified across other touchpoints or customer groups.
A fast way to understand what’s positively and negatively affecting your north star metrics, and to effectively prioritize what to work on, is to take advantage of Chattermill’s Impact Analysis functionality. In video 2 of the How to Analyze Customer Feedback in Retail Organizations course, Tom switches to an Impact Analysis report to get clarity over which themes are causing the spike in NPS he located.
In the example above, ‘Shipping speed’ and ‘Price’ are clearly a detriment to the NPS and could be areas to focus on, whilst themes like ‘Style & Design’ are solidly green, and are factors to work on preserving. It’s here that you can turn the feedback associated with your VOC metrics into something operational that better informs your roadmap in retail.
Monica Vinader’s story
We’ve talked a lot about how to improve your VOC Metrics, but we’re also all about delivering real results.
Take Monica Vinader, a global jewellery brand, whose partnership with Chattermill transformed their VOC program into measurable results.
After unifying all of their sources of feedback, including NPS surveys, CSAT, Trustpilot reviews, and post-purchase surveys - all into one place, Monica Vinader was able to access a single, unified source of customer truth. It’s here that they were able to focus on their north star NPS metric, and noticed a clear problem with customer churn.
To tackle the churn problem, the team used Impact Analysis to find out why their NPS dropped or increased, drill down into specific NPS drivers, and find the underlying causes.
"Impact Analysis allows us to understand why our NPS score dropped last month or what's causing our first-time buyers to become detractors. It's a great way to understand which trends impact our NPS and understand the "why" behind our score," says Jade.
Today, the team prides itself in having a consistent NPS score above 70 and a 4.9 average score at Trustpilot, with over 93% of 5-star reviews out of 12,550 comments. Click here to read the whole story.
Next Steps
Want to transform your approach to your VOC metrics in retail?
Become an expert by signing up to the CX Intelligence Academy today. VOC Metrics Analysis and How to Analyze Customer Feedback in Retail Organizations are just two courses in our roster of 7, with the others providing industry-recognized expertise across a range of use cases and best practices. All courses are completely free, and award a certification upon completion of a short assessment.
Book a free demo to see how Chattermill helps leading retail brands like Monica Vinader, River Island, Faire, and MusicMagpie reduce churn through actionable CX insights.
