Voice of Customer (VoC)
Voice of Customer (VoC) is a research discipline that captures what customers actually say about a product, problem, or category — in their own words — and uses those verbatim phrases to guide messaging, product, and positioning. Unlike surveys that force customers into predefined answer boxes, VoC listens to open-ended language: reviews, support tickets, sales calls, social posts, interviews, and Slack messages.
Voice of Customer (VoC) is a research discipline that captures what customers actually say about a product, problem, or category — in their own words — and uses those verbatim phrases to guide messaging, product, and positioning. Unlike surveys that force customers into predefined answer boxes, VoC listens to open-ended language: reviews, support tickets, sales calls, social posts, interviews, and Slack messages.
Why It Matters
Most marketing language is written by marketers for marketers. "Supercharge your workflow" or "unlock insights" sound clean in a meeting and land flat with buyers because no real customer says those words. VoC closes that gap: when a brand's website copy uses the exact phrases customers themselves use, conversion rates routinely jump 20–50% without any change in product or audience. The most effective copywriters, PMMs, and founders treat VoC as the highest-leverage research they do — it's why "write like your customer talks" has become a standard maxim and why tools like Dovetail, Gong, and EnjoyHQ exist.
Where VoC Data Comes From
Customer interviews: One-on-one conversations with existing users, recorded and transcribed. The richest source.
Sales call transcripts: Tools like Gong, Chorus, and Fireflies capture thousands of prospect conversations. Pure unfiltered language.
Support tickets: How customers describe their problems when something goes wrong.
Product reviews: Trustpilot, G2, Capterra, Amazon, App Store — public language at scale.
Reddit / Quora / forums: Long-form discussions in user-chosen words, especially valuable for category framing.
Win/loss interviews: Why did this customer buy, and why didn't that one?
Churn interviews: Why are they leaving? Often the most honest answers.
Social listening: Mentions on X, LinkedIn, TikTok, YouTube comments.
Onboarding "Why are you here?" surveys: Free-text responses from new users explaining what they hoped to accomplish.
How to Do VoC Research
1. Define what you're looking for: Why does my ICP consider this category? What pains drive them? What words do they use?
2. Gather raw text: Collect 50–200+ verbatim examples from the sources above. Copy-paste into a single document or research tool.
3. Code by theme: Tag each excerpt with themes (pain, solution, objection, alternative tried, outcome). Use sticky notes, spreadsheets, or tools like Dovetail.
4. Find repeated language: Words and phrases that appear multiple times across independent sources are signal. One-offs are noise.
5. Surface the patterns: Write up the top themes with verbatim quotes. Quotes matter more than your summary — the raw words are the data.
6. Feed into product, marketing, and sales: Update website headlines, sales decks, ads, and product copy with the exact customer language.
VoC vs Surveys
| Aspect | Surveys | VoC |
|---|---|---|
| Input | Multiple choice + Likert scales | Open-ended text |
| Structure | Predefined | Emergent |
| Effort to analyze | Low | High |
| Depth of insight | Shallow | Deep |
| Bias risk | Leading questions | Sampling bias |
| Output | Percentages | Verbatim quotes |
Surveys tell you how many. VoC tells you in what words. Mature teams use both.
What VoC Changes
Website copy: Replace marketer-speak with customer phrases. "Content calendar that doesn't feel like homework" beats "streamlined editorial workflow."
Positioning: Customers may describe your category in words you'd never choose. Often those are the right words.
Ad headlines: Testing ad variations with customer-language phrases vs marketer-language phrases routinely shows 2–5× CTR differences.
Product roadmap: Repeated themes ("I need to do X before Y") reveal feature priorities that user request lists don't.
Sales objections: Knowing the exact pushback language lets sales preempt it.
Onboarding flows: Word each step in language customers already use, not internal feature names.
Common Mistakes
Assuming you already know: Every founder thinks they know what customers say. Actually pulling 100 quotes always surfaces surprises.
Editing quotes: "Cleaning up" a customer's phrasing destroys the data. Verbatim or nothing.
Sampling only happy customers: Churn and lost-deal VoC is often the most valuable.
Mistaking opinions for data: A single loud customer is not a pattern. Look for repetition.
Never closing the loop: Research that doesn't change copy, product, or sales behavior is overhead.
Running it once: VoC drifts as market, product, and audience evolve. Refresh quarterly or semi-annually.
Sources: