Implementing Voice of Customer Programs

I was chatting to a CX director last week who proudly said their VoC programme had high response rates and monthly reports shared with leadership. But when I asked "What do you know about the customers who never responded?" He couldn't say a word. Most VoC programmes I see require the customer to actively choose to respond. The problem is that for every customer who complains, 26 others stay silent and just leave.. → They skip a renewal and say nothing.  → They abandon the checkout and say nothing.  → They browse your cancellation page three times in a week and say nothing. But all of these behaviours are signals. Silent feedback that your VoC programme was never built to capture. So while companies keep optimising based on the minority willing to talk, they are missing a huge chunk of what is actually happening. AI changes this entirely, by making customer signals readable at scale: ■ Instead of asking why customers left → predict who is about to ■ Instead of waiting for NPS to drop → catch the pattern before it compounds ■ Instead of channel-by-channel reporting → one intelligence layer connecting every signal The real VoC data isn't in your survey responses. It's in everything your customers did before they stopped showing up. ▶︎ Chattermill mapped out the six shifts changing what VoC actually means in the AI era. If your programme is still waiting for customers to raise their hand, start here: https://bit.ly/4rrcQqT The customers who never complained are the most expensive ones you'll ever lose. And right now, your VoC programme has no idea who they are. The real question is what are our customers telling us through everything they do? #ad #VoC #cx #customerexperience 

When organizations launch a Voice of the Customer (VoC) program, there is often a disproportionate focus on launching surveys. But launching surveys is only half the job; acting on the feedback is where real value is created. Even if the program is rolled out in phases Stage 1, Stage 2, Stage 3, it should not be only about what surveys we launch and when. Each phase must clearly define what we will do with the feedback collected. In most conversations we have with organizations, the plan is very detailed on: a.how many surveys will be launched, b. where they will be deployed, c. how dashboards will be reviewed. But there is very little clarity on: 1.who will own the actions, 2.what actions are realistically possible, 3.whether the organization has the manpower, governance, and decision structures to act. Saying “we will launch 70 surveys in a year” sounds ambitious, but an important question follows: Will we realistically be able to take action on insights from all 70 surveys? If the answer is no, then the program risks becoming a reporting exercise rather than a transformation initiative. The better approach is simple: launch less, act more. Start with fewer, high-impact surveys, build strong action loops, assign clear ownership, and prove business impact. Once action becomes a habit, not an exception, scaling the VoC program actually makes sense. Thoughts? #customerfeedback #voc

Most product roadmaps are optimized for feature velocity. Very few are optimized for problem resolution efficiency. Here’s the gap: Structured data tells you what users are doing Unstructured data tells you why they’re struggling And most teams underutilize the second. Customer voice is not just feedback. It’s a high-dimensional signal space. Hidden inside it are: Latent problem clusters Friction points across journeys Early indicators of churn and drop-offs But extracting this requires more than tagging or keyword tracking. A technical approach to converting feedback → roadmap looks like this: 1. Semantic Normalization via Embeddings Raw feedback is noisy: “App is slow” “Takes forever to load” “Performance is laggy” Using transformer-based embeddings, these are mapped into a shared vector space, enabling: 👉 Semantic clustering of problem statements 👉 Reduction of linguistic variance 2. Unsupervised Problem Discovery Instead of predefined categories: Apply clustering (HDBSCAN / density-based methods) Identify emergent issue groups This surfaces problems you didn’t explicitly define. 3. Signal Weighting (Beyond Frequency) Not all feedback carries equal importance. A robust system weights signals based on: Intensity (language strength, sentiment gradients) Recurrence (cross-user + cross-channel repetition) Temporal patterns (sudden spikes vs steady signals) 4. Cross-Channel Correlation The same issue appearing in: Support tickets Reviews Sales conversations …indicates systemic product friction This is where prioritization becomes clear. 5. Mapping Feedback → Business Outcomes This is where most systems stop, and where the real value begins. Link problem clusters to: Activation drop-offs Feature adoption gaps Churn signals Now roadmap decisions are not based on volume. They’re based on impact correlation. What this enables: Prioritization of high-impact issues over high-volume noise Identification of hidden blockers in the user journey Alignment between product, CX, and revenue teams The shift: From: ➡️ Feature-driven roadmaps To: ➡️ Signal-driven product systems At scale, this requires infrastructure that can: Continuously process unstructured feedback Update clusters dynamically Align insights with business metrics What is one product decision you made recently that wasn’t backed by customer signals? Would you still make the same call today?

We've processed over 4 million customer feedback. Here are 5 patterns that changed how we think about Voice of Customer: 🔍 The paradox of the obvious cluster The biggest complaint category? It's rarely your biggest problem. We found that 70% of the actionable insights hide inside large, recurring clusters. Example: 400 feedbacks about "slow performance" actually contained: A critical mobile-specific bug (27 mentions, buried) A feature gap causing workarounds (81 mentions) A documentation issue (53 mentions) Don't skip the elephant. Dissect it. 💡 Pre-sale meets post-sale = reality check Combining marketing/sales data with support/product feedback reveals the expectations gap. What customers thought they were buying vs what they actually experienced. This gap predicts churn 3x better than NPS alone. 🤖 Bots hear the unfiltered truth Customers are more honest with chatbots than humans. We see consistently more negative sentiment and more specific complaints in bot conversations. Why? No social pressure. Your bot logs are your most honest feedback source. 📊 One feedback = multiple insights "Love the features but setup was confusing and nobody replied to my email" That's not "mixed sentiment." That's: Product ✅ Onboarding ❌ Support ❌ Breaking down multi-topic feedback increases insight density by 3-4x. ⏰ Timing beats templating Smart, in-app surveys at trigger moments (right after a key action) get 5x better response rates than generic email blasts. And the quality? Not even comparable. These patterns repeat across industries. The companies that structure unstructured feedback win. P.S. Want vertical-specific benchmarks or see how your metrics compare? Drop a comment or DM.

90% of data is unstructured, but most businesses ignore it. Here’s how I turn it into retention gold. When I look at tickets, emails, CRM notes and call transcripts, I see more than noise. I see early churn signals hiding in plain sight. 🚨 Most teams wait until customers leave before they act. By then, it’s too late. I’ve learned to flip the script. Here’s the workflow I use to turn unstructured data into action (and real retention gains): 1️⃣ Signal hunt → Pull every ticket, email, CRM note and call note into one place → STRUCTURE the information → Tag accounts with these signals 2️⃣ Act fast → Route flagged signals to the right team member → Respond within hours, not days → Give teams scripts to handle each signal type (don’t guess!) 3️⃣ Measure and improve → Track which signals predict churn best → Review outcomes weekly → Adjust detection rules and playbooks as you learn I’ve rolled out this system at several SaaS companies. Each time, churn dropped and retention climbed. The secret? Don’t wait for a survey or a renewal date. Listen to what customers are saying right now—even if it’s hidden in the noise. OK. STOP HERE! What's the problem then? The data is messy. Really messy. Turning raw customer conversations into actionable insights requires extensive cleanup, classification, and analysis. Analytics teams spend weeks wrestling with data extraction and preparation, often for minimal payoff. The technical complexity of processing unstructured text—from parsing different formats to creating meaningful embeddings—creates massive bottlenecks. The good news? Turning unstructured data into business impact is what qwe do at Flexor. It’s a game-changer! 🚀 How are you surfacing hidden signals in your customer data? Would love to hear what’s working for your team. #Retention #Churn #UnstructuredData #AI #SaaS #CustomerSuccess

I am saying it for the 10000th time: NPS is not VoC, and VoC is useless without support conversations. There's more to Voice of the Customer (VoC) than collecting just NPS. 1. The NPS Trap: Sure, Net Promoter Scores (NPS) have their place, but they're not the be-all and end-all. NPS can give us a snapshot but often misses the bigger picture—the why behind the score. 2. Support Conversations are Vital: Real-time, nuanced feedback from support conversations uncovers valuable insights. Incorporating them into your VoC strategy is a comprehensive understanding of customer needs. 3. VoC is an Ongoing Process: It's not a one-time project; it's a continuous effort. To truly understand customer needs and expectations, ongoing conversations and insights at every touchpoint are vital. 4. VoC Informs Long-Term Strategy: It goes beyond immediate fixes. Analyzing support interactions and conversations provides valuable insights guiding long-term strategic decisions and customer-centric initiatives. Let's avoid the misconception that VoC is solely about feedback collection and NPS. By including support conversations as a part of VoC, we capture the authentic voice of our customers and drive meaningful improvements. Are you including insights from support in your #VoiceoftheCustomer ?

Your content isn’t converting enough visitors into demos. Here’s how to use ChatGPT deep research + real lead data 𝘁𝗼 𝗶𝗺𝗽𝗿𝗼𝘃𝗲 𝗰𝗼𝗻𝘃𝗲𝗿𝘀𝗶𝗼𝗻 𝗿𝗮𝘁𝗲. Your leads are literally telling you how to sell to them. You just aren’t listening. If you don’t already have them, start by adding two unstructured fields to your forms: 1. “How did you hear about us?” 2. “What’s your biggest goal or need right now?” Next, on every demo or discovery call, ask prospects: 1. "What was your process to find us"? 2. "What pain points made you reach out"? Write these answers down. Every time. Now drop all that into a spreadsheet. Strip out any sensitive info. Fire up your favorite Deep Research GenAI tool and feed it this prompt: —--- “You are a data-driven B2B marketing strategist with expertise in qualitative analysis. Your task is to analyze a dataset containing unstructured lead and sales prospect information, including where the lead was sourced and open-ended responses about their biggest pain points, needs, and goals. When finished, produce a copy of the sheet with a new column that lists each response row's new cluster name. 𝗢𝗯𝗷𝗲𝗰𝘁𝗶𝘃𝗲: Perform a cluster analysis to identify common themes in how prospects describe their challenges and desired outcomes. Provide structured insights that can be used for targeted marketing messaging. 𝗜𝗻𝘀𝘁𝗿𝘂𝗰𝘁𝗶𝗼𝗻𝘀: Ingest Data: Analyze the provided spreadsheet. Focus on the unstructured text fields where prospects discuss their pain points, needs, and goals. Cluster Identification: Identify groups of similar responses based on recurring language, themes, or sentiment. Look for patterns in how leads describe their problems and what solutions they seek. Cluster Naming: Assign a clear, memorable name to each cluster. Example Quotes: For each cluster, provide the top 3 most representative examples (verbatim quotes) from the dataset. Marketing Insights: Summarize key takeaways from the clusters, including potential messaging angles, content ideas, and value propositions tailored to each segment. Output Format: Cluster Name: [Descriptive name] Definition: [Brief description of this cluster’s common pain points, needs, and goals] Quotes: "[Direct quote from a prospect]" "[Direct quote from a prospect]" "[Direct quote from a prospect]" Marketing Insight: [Actionable insights on how to tailor messaging, positioning, or content] (Repeat for each identified cluster)” —---- Now use those insights to rewrite your core messaging around what’s actually being said. Think: 👉🏾 Calls to action 👉🏾 Linkedin posts 👉🏾 New solutions/use case pages on your website 👉🏾 Blog post topics/titles 👉🏾 Youbute Ad creative 👉🏾 Newsletter 👉🏾 Google Ads ad groups Group messaging by source. Pay attention to what the impact is. Repeat this process every 3-6 months. ------- I hope today is the best day of your entire life. Cheers. 🚀 #chatgpt #seo #content #ai

The Problem with Solving the Problem in CX Jumping to conclusions. We think we know what the problem is. We assume we’ve seen it before. And we act fast. But sometimes speed is the enemy of accuracy. Let me share an example. Once we noticed a consistent spike in one category of customer queries: “Where is my order?” The apparent assumption we had was that Logistics delays would occur. We went full swing on working with delivery partners, streamlining SOPs, and keeping tight follow-ups. But even after trying to solve this for a few days, the volume was not coming down. We had to dig deeper to find the root cause. The data showed that only 28% of these queries were actually due to logistical issues. So why were so many customers still reaching out? When we dug deeper, the real issue came to light: Order status updates weren’t refreshing on time. The delay wasn’t in the delivery; it was in the communication. The app said “Order Confirmed” for days. Naturally, customers panicked. What we had wasn’t a logistics problem. It was a visibility problem. And the only way we discovered that was by listening to our customers and conducting a thorough root cause analysis. This happens far more often than we admit in CX. - We launch a chatbot to reduce query volumes, but the real issue is broken flows. - We offer discounts to retain customers, when the real issue is poor product quality. - We train agents to improve CSAT, but the real issue lies in complex backend processes. You invest time, energy, and budget, yet still achieve the same outcomes. Here’s my takeaway: - Before solving, slow down. - Trace the issue, find the root cause - Read the comments. - Talk to your customers. - And solve the actual problem, not the most obvious one. Because in CX, symptoms may lie. But the customer’s voice doesn’t. Have you seen examples where teams solved the wrong problem? Would love to hear your take. #CustomerExperience #CXStrategy #RootCauseAnalysis #VoiceOfCustomer #CustomerSupport #DesignThinking #CustomerCentricity

Most VoC meetings aren’t built for change. They’re built for applause. Here’s the cycle you’ve probably seen: CX teams spend weeks preparing beautiful decks, dashboards, and top themes by sentiment (using AI). They tell emotional customer stories to humanize the data. Executives nod and say, “Great visibility. Keep up the good work.” Then everyone leaves—and nothing changes. This isn’t customer centricity. It’s a reporting theater. We’ve been taught to believe that sharing insights is enough. That if people see the data, they’ll act. But the reality? Visibility does not equal accountability. Awareness does not equal outcomes. Here’s why this cycle persists:  • VoC has been positioned as a support function, not a strategic one. It delivers “insights” but doesn’t own outcomes.  • Meetings lack operational teeth. There are no clear owners, no timelines, and no follow-through.  • Leaders feel informed, but not responsible. They applaud the insights and move on to the next fire. The cost? Customer pain festers. Churn quietly rises. Teams stay busy but disconnected from impact. Real VoC work begins where reporting ends. If your meeting doesn’t:  • Prioritize pain points tied to churn, revenue, and engagement  • Assign owners and deadlines  • Track progress on past commitments …it’s not a VoC meeting. It’s a show. Customers don’t care how many insights you present. They care about what you do to alleviate their pain. Next: I’ll share how to turn your VoC meeting into a decision engine—one where insights don’t just inform, they drive action.

Working with unstructured data can be slow and messy. But with the right AI setup, you can stop drowning in open-ends and video feedback, and start surfacing real insight—fast. Agentic AI is the key here. This is not just about smarter tools. I'm taking about specialized agents that handle different research jobs (summarizing, probing, curating, reporting) and work together like an orchestra. It’s modular, scalable, and built for speed and rigor. So, how do you make agents for you? Here are some best practices that we share with our clients at Rival Technologies and Reach3 Insights: 📱 Capture better input. Rich insights start with authentic, thoughtful responses. Use mobile-first, conversational methods to get customer feedback in the moment—not hours later when the magic’s gone. 🧠 Let AI do the heavy lifting—strategically. Automate the tedious parts: clustering, summarizing, follow-ups. But always keep the human in the loop to catch nuance and avoid hallucinations. 🧩 Orchestrate, don’t just automate. Agentic frameworks shine when agents work in concert — each with its own task, collectively building something greater. 🎯 Measure thoughtfulness, not just word count. AI can detect if a response is vague or surface-level. At Rival and Reach3, we have an AI tool called Thoughtfulness Scoring that specifically measures how (as the name suggests) thoughtful a response is. The benefit? It guides probing features to follow up only when needed. So you can get richer insights without annoying your participants. The shift to agentic is not about automation for automation’s sake. It's about giving researchers superpowers — more time for strategic thinking and more compelling ways to deliver value to the business. If you want more tips on using agents to elevate your qual research, check out this new guide from our team: https://lnkd.in/gSNN3Sey #QualResearch #AI #marketresearch