Voice-of-customer research: mine the words buyers actually use

Here is the skill that separates working copywriters from people who write nicely: voice-of-customer research. Buyers describe their problems in words you would never invent sitting at your desk. Those exact phrases live in reviews, forums, support tickets, and sales calls. Copy built from them reads like the customer's own thoughts — because it is. When a reader thinks "that's exactly my situation," the page is halfway to the sale.

Where to mine: product reviews for your client and their competitors — the five-star reviews give you outcome language, the three-star reviews give you objections with the emotion still in them. Forums and groups where buyers ask strangers for help. The client's support inbox and chat logs, where the same questions appear weekly. Sales-call notes and recordings, where you hear what people say in the moment before they decide. Cancellation and return reasons, which are objections that arrived too late.

The method: before drafting anything, collect 30–50 verbatim phrases into your notes doc, sorted by job — pains, desired outcomes, objections, and the words buyers use for the product itself. Keep the wording raw. "It saves me time" is mush; "I get my Sunday afternoons back" is a headline embryo. Your five-star reviews will often contain a sentence stronger than anything you'd write from scratch — your job is to find it and build around it, not to out-write it.

Three pitfalls: paraphrasing too early, which smooths the life out of the phrases before you can use them. Mining only happy customers, when the objections that kill sales live in the three-stars. And inventing phrases that merely sound right — invented customer language is how copy ends up sounding like everyone else's, and sounding like everyone else is the one thing clients pay a human writer to avoid.

Interviewing a client: proof, objections, and the real offer

Before you write a word for a client, interview them. Forty-five to sixty minutes, recorded with permission, questions prepared in advance. You are mining for three things: proof you can put on the page, objections the copy must answer, and the real offer — which is often better than the one the client has been describing out of habit.

The questions that do the work:

• Describe one real best customer. Who are they, and what was going on in their life when they bought?
• What do buyers say in the moment they decide — the sentence right before "okay, let's do it"?
• What almost stops them? Price, trust, timing, a partner who has to agree?
• After it works, what do customers say — in their exact words?
• What can you prove with numbers, and where does the evidence live?
• What do competitors promise that you believe is overclaimed?
• Walk me through what happens after someone buys, step by step.

Listen for stories, because stories become case-study material. Listen for numbers, because numbers become proof. And for every claim, ask "can you show me?" — the results screenshot, the testimonial email, the review. Claims you cannot verify do not go in the copy; Module 3 covers how to write strongly around a gap instead of papering over it.

The output of a good interview is an offer brief in your notes: audience, promise, proof list, objection list, and a handful of the client's customers' own phrases. Drafting from that brief is fast, and revisions stay small because the facts were settled up front. Clients feel the difference, too. Writers who interview get rehired. Writers whose first question is "so, what should the page say?" do not.

Swipe files done right: study structure, never copy lines

A swipe file is a working library: ads, emails, landing pages, and headlines that made you stop scrolling — saved with notes. Every serious copywriter keeps one. Amateurs keep a folder of five hundred screenshots they never open again. The difference is not the collecting. It is the annotation.

The rule that keeps the file both legal and useful: study structure and psychology, never lift lines. On every save, write down what you are actually extracting. The promise type — save time, avoid embarrassment, gain status, remove a risk. The proof device — a demo, a number, a testimonial, a guarantee. The objection it answers and where. The order of the sections and why that order works. One honest note per save ("works because the headline names the failed alternative first") is worth more than the screenshot itself.

Where the material comes from: your own inbox — join the email lists of brands in your niche and file what they send. The public ad libraries the platforms maintain, where you can read competitors' running ads. The pages you research during client work. Your own purchases — the emails that got you to buy are a masterclass addressed to you personally. Five annotated saves a week builds a real library in a few months.

How you use it: before drafting, pull three or four swipes in the same format and outline their skeletons side by side. Your page gets their proven architecture and your client's own voice-of-customer words — borrowed structure, original language. That combination is the whole method. And the practical reason copying lines fails, beyond the ethics: another brand's voice in your client's mouth reads wrong immediately, to everyone, including the client.