How to Use Research in Copywriting

As an email copywriter connoisseur, research is 80% of my process. This is what people typically think research looks like: - Look at ten customer reviews - Maybe watch a demo video - Glance at past emails and brand guide This is what my copywriting research process looks like: - Analyze thousands of reviews from competitors and the brand I'm working with - Review mine in Reddit and other forums that may provide insight into customer data - Check for existing surveys that the brand may have used - Check for heatmap on existing landing page or website - Watch and rewatch as many demo recordings as possible - Interview the founder, sales team, and marketing team - Audit past email content to see what's performed the best - Cross-analyze best-performing blog content, landing page, and social content - Evaluate the stage of awareness of the audience from best-performing content - Analyze email data based on seasonality and segmentation - Analyze what has the highest CTR for different emails - Identify what copywriting framework to use and why - Analyze brand voice, value props, features, and listed outcomes - Analyze the ICP of the brand and common objections - Analyze the voice of the customer based on data This is just the research part. The actual writing part is a whole nother level. But I take pride in my research process. Because it's exhaustive. And I'm determined to find information and know WHY I'm writing it. It's why health tech and SaaS brands love working with me. They aren't just getting good copywriting from me that sounds good. They are getting deep-researched, data-driven copywriting that is built to convert.

Everyone thinks copywriting is clever hooks and persuasion frameworks. I've found the opposite is true: Once you've done the research, your ads are already 80% written. The market already knows exactly how to sell your product. They're screaming it at you through every review, comment, and Reddit thread. The problem is most marketers are too busy crafting "perfect hooks" to listen. Instead of diving into another copy template, try this: Get obsessive about your product documentation. Not the marketing fluff - the actual technical specs and functionality. Those become your best converting points. Dig into your demographic data and look at actual behavior patterns. Where do people hesitate? What makes them convert? You figure all this out by hanging out where your customers actually talk freely. Look at Reddit threads and Facebook groups where people are having conversations with their peers. That's where you find out what they're actually thinking. Then, track your competitors like a hawk. If you know what to look for, you can find the positioning gaps they're missing. Opportunities are in the questions nobody else is answering. Instead of overfocusing on slick copywriting tactics, learn to let the research write the copy for you.

Most content writers start their research with Google. That’s a problem. If you research first, before forming your own opinion, you: Regurgitate what’s already been said. Default to “safe” takes instead of original thinking. Sound like every other keyword-driven blog post on the internet. Here’s what to do instead: 1. Start with your own take Before you Google it, or ask ChatGPT, ask yourself: - What do I already know about this topic? - What’s my gut instinct or first reaction? - If I had to explain this to a friend, how would I say it? Why? This forces you to think critically, not just accept what’s already out there. It surfaces unique perspectives you might overlook if you go straight into research mode. It helps you see where your knowledge gaps are, so research becomes intentional, not just a content-filling exercise. For example, let's say you are writing about “Why Most B2B Content Doesn’t Convert” ❌ Google: “Why B2B content fails” → See generic listicles → Regurgitate same tips. ✅ Instead: Ask yourself, What have I personally seen that makes content flop? ➡ Maybe you realize: “Most B2B content is too safe and lacks a real point of view.” That’s a unique angle—now you research to back it up. 2. Find content gaps Once you have your own thoughts, now it’s time to see: Where do others agree? Where do others disagree? What’s missing from the conversation? Where to look for content gaps: Reddit & X → See what real people are talking about. Podcasts & YouTube comments → Experts share insights they don’t write about. Sales & support teams → They hear real questions from customers daily. For example, if every article says, “B2B buyers don’t read long content,” …but a Reddit thread has 50+ comments from marketers saying they do—boom, there’s your contrarian take. 3. Use AI as a thought partner, not a crutch Don't let ChatGPT replace your thinking. Use it to pressure-test and refine your ideas. Better AI prompts for original insights: ❌ “Write a blog post on why B2B content fails.” ✅ “What are 5 unconventional reasons why B2B content doesn’t convert?” ✅ “Challenge this idea: ‘TOFU content is more important than BOFU.’” ✅ “Give me 3 counterarguments to the idea that ‘SEO is dead.’” Instead of accepting AI’s output, you’re interrogating it. It forces you to think beyond the obvious. You get nuanced arguments, instead of generic summaries. 4. Research to validate and strengthen your take Find data to back it up. Find real-world examples. Find an expert. Talk to them. 5. Then, write it like no one else has Frame your idea it in an unexpected way. Be bold enough to spark a reaction. If nobody disagrees with you, your take isn’t strong enough. If your post could’ve been written by anyone else, go back and make it more you. Google can give you facts. 𝗢𝗻𝗹𝘆 𝘆𝗼𝘂 𝗰𝗮𝗻 𝗴𝗶𝘃𝗲 𝘆𝗼𝘂𝗿 𝗽𝗲𝗿𝘀𝗽𝗲𝗰𝘁𝗶𝘃𝗲.

Stop trying to write clever copy. Start mining your reviews, support tickets, and Reddit threads for the exact phrases your customers already use. THEN build your headlines, emails, and product descriptions around those words — not yours. Here's the principle underneath this ↴ Copywriting isn't about inventing language. It's about curating it. You've gotta realize your reader's brain is constantly running a background check on everything it reads. Does this sound like me? Does this person get my problem? When the answer is yes, two things happen fast: trust goes up and friction goes down. Both of those move people toward a click, a cart, a conversion. When the answer is no — when your copy sounds like it was written by a marketing department instead of a human who understands the problem — you lose them. Doesn't matter how polished the headline is. This is why "use customer language" isn't just a style tip, but rather a conversion principle. And it governs three rules you've probably heard before: 1) clarity over cleverness, 2) specificity over vagueness, and 3) benefits over features. All three are downstream of the same idea: Copy converts when it mirrors how your reader already thinks, and not when it tries to teach them a new way to think. So, next time you sit down to write, open your VOC data before you open a blank doc. The best copy is already written — you just have to find it. 🔍 #copywriting #conversioncopywriting #dtcmarketing #emailmarketing 💡 I put together a free resource called the Copy Principles Swipe File, Vol. 1 — 11 real ads, emails, and landing pages, each annotated by the principle that makes it work. If you want to start seeing the WHY behind great copy instead of just the what, drop "SWIPE" in the comments or DM me and I'll send it your way. (P.S. If this kind of thing makes you cringe, I’m sorry. Experimenting with some new ways to grow my email list. Never know what will or won’t work, and you don't know what you don't know until you try.) — 👋 I'm Matt Snyder, Director of Copywriting at Homestead Studio and the writer behind The Copy Minimalist. I write about conversion copywriting and the principles that govern it.

Stop writing copy based on gut feel. Here's our AI-assisted copywriting framework for website/product sprints you can use for your next research phase: 1. Discovery research: - Send and analyze client discovery questionnaires (brand + product/website). Ask about their ICP, pain points they solve, and what makes them different. - This will give you their internal perspective before you can validate it externally with research. 2. Run a competitor analysis - Look at how competitors frame their value props and what language they use. See what resonates. - Scrape online reviews with AI from forums, social media, support tickets, app store reviews to build extra problem-solution context. 3. Screenshot user flows from competing products (use Mobbin for this). - See how competitors structure their onboarding, messaging, and feature explanations. Look for repeating patterns to follow or break. 4. AI-assisted testing: - Use AI to generate copy and apply it to 15-20 hero variations. Test different value props, frameworks, and angles. - Save the best performers for A/B testing later. You're not trying to pick the winner now - just narrowing from 20 to 5 best versions. 5. Strategic validation: - Cross-check your top copy variations against research findings. Does this match what users said in reviews? Does it address the competitor weaknesses you found? Does it align with the questionnaire insights? - Use AI to help - feed it your research summary and top 5 copy options, ask it to identify gaps or misalignments. 6. Apply to all sections/screens Once you've validated your hero messaging, follow the same pattern for every other section and app screen. Each section should map to one of your core values (time, money, performance) and use the appropriate framework (PAS, AIDA, PSB). 7. Mix and match frameworks - For features, lead with benefits (PSB works well). - For social proof, use AIDA to build desire. - For conversion points, PAS creates urgency. Use AI to generate variations based on the winning hero framework, then validate against your research. 8. Be consistent Every section should reinforce the same core USP in different ways. Don't forget microcopy (buttons, error messages, empty states). These should follow the same voice and logic. "Get started" vs "Start saving time" vs "Start free" - all imply different value. Most startups can't afford to spend months on messaging. This used to take days, not it takes hours with AI and will easily fit your sprint timeline. Combine client insights, market research, and AI-assisted testing. Try this for your next discovery session :)