Voice-of-customer research
The best-converting line on your page is usually one you didn’t write — you quoted it. Here is how to find your customers’ own words and put them to work.
What voice-of-customer research is
Voice-of-customer (VOC) research is the practice of building your messaging out of the language your customers already use, rather than the language you’d prefer they used. Instead of inventing a clever way to describe the problem, you find how they describe it — and say that back to them.
It works because recognition beats persuasion. When a visitor reads their own half-formed thought stated cleanly on your page, they don’t feel marketed to; they feel understood. That’s the difference between a line that converts and one that gets skimmed.
The mistake: writing in your words, not theirs
Teams default to the language of the thing they built — “end-to-end workflow automation”, “unified data layer”. Customers don’t talk like that. They say “I keep copy-pasting the same thing between two tabs all day.”
Both describe the same product. Only one matches what’s already in the customer’s head, and that’s the one that earns the click. VOC research is mostly the discipline of catching yourself using the first kind and replacing it with the second.
What to mine from a real thread
Reading customer conversations for copy, you’re hunting for four specific things:
- Exact phrases — the literal words for the problem, ready to paste onto a page
- The verbs — what they’re trying to do (“keep track of”, “stop missing”), which become your value props
- The comparison — what they weigh you against, including “a spreadsheet” and “doing nothing”
- The metaphor — the image they reach for, which is often a better headline than anything you’d draft
- The objection — the “yeah but…” that your copy has to answer before they’ll act
From quotes to copy
The pipeline makes this concrete: every classified thread carries verbatim quote fields — the best line from the original poster, the best from the top reply, and the key quotes from the discussion — each with a link back to the source.
So the output isn’t a vague “customers care about speed”; it’s a sortable stack of real sentences, attributed and linkable. You write your page by promoting the best of them, lightly edited, instead of starting from a blank document and your own assumptions.
Put the words to work
Voice-of-customer mining feeds directly into landing-page and content writing — here’s that workflow end to end.
Research-backed writingKeep reading
Write essays and threads backed by real evidence
Draft essays and X threads grounded in real evidence, not vibes.
Read →Write content about what your audience actually asks
Write about the questions your audience is actually asking.
Read →Reddit market research: reading a market in its own words
A founder spends six weeks and a chunk of budget on a survey, then learns more from one afternoon reading a niche subreddit. This is how to do Reddit market research on purpose instead of by accident.
Read →How we score a thread for pain and willingness to pay
The exact schema that turns a messy Reddit thread into a rankable pain and willingness-to-pay score.
Read →The objections buyers voice when you’re not in the room
Buyers are polite to your face and honest behind your back. Here’s how to map the objections they only voice when you’re not in the room.
Read →