How to write landing-page copy that converts (with examples)

Most landing pages fail at the same spot: the top. A visitor lands, reads your headline, and within a few seconds decides whether to keep going or close the tab. If your copy makes them work to understand what you do, they leave. No amount of nice design saves a page that doesn't say anything.

Good news: copy that converts isn't about clever wordplay or a thesaurus full of power verbs. It's about being clear, specific, and honest about what you do and who it's for. Here's how to write it, with concrete before-and-after examples you can apply today.

Lead with the headline: clarity beats cleverness

Your headline has one job: tell a stranger what you do and who it's for. That's it. Cleverness is a tax on comprehension, and you can't afford it in the first five seconds.

Weak: "Reimagine your workflow." Strong: "Turn your invoices into paid bills in two clicks."

The first could describe a yoga app or a tax tool. The second tells a freelancer exactly what they get. Notice the strong version names the outcome (getting paid) and implies the audience (people who send invoices). You don't need to be poetic. You need to be unmistakable.

A simple structure that works: "[What it does] for [who it's for]." For example, "Bookkeeping software for solo consultants who hate spreadsheets." Boring? Maybe. But the visitor knows in one read whether they're in the right place, which is the only thing that matters.

Use the subhead to add the proof or the how

If the headline makes the promise, the subhead earns a little trust. Use it to explain how you deliver the outcome, or to handle the first objection that pops into the reader's head.

Headline: "Turn your invoices into paid bills in two clicks." Subhead: "Send, track, and auto-remind clients who forget. Most users get paid 11 days faster."

The subhead answers the immediate "okay, how?" and adds a specific number. Avoid restating the headline in different words, which is the most common subhead mistake. If your subhead is just the headline wearing a hat, cut it.

Start with the customer's problem, not your feature list

Founders love their product. That's natural, and also the trap. You know every feature you shipped, so you write about features. But the visitor doesn't care about your features yet. They care about their problem and the outcome they want.

Open with the world from their point of view. Before: "Our platform offers automated multi-channel scheduling." After: "Stop posting the same thing to five apps by hand." Same product, but the second version starts where the reader already is: annoyed at a chore. Lead with the pain or the result, then introduce the product as the bridge between them.

This is also why you should write the page before you fall too deep into design. If you're using Kalit Flow to build the page itself with AI, feed it copy that's already grounded in the customer's problem, not a feature dump dressed up as marketing.

Translate features into benefits (with a concrete rewrite)

A feature is what your product has. A benefit is what the customer gets. People buy benefits. The fix is mechanical: take each feature and ask "so what?" until you hit something the reader actually wants.

Feature: "End-to-end encryption." So what? "Your client data is encrypted." So what? "Your client data stays private, even from us." Benefit copy: "Client data only you can read. Not even our staff can see it."

Another one. Feature: "AI-sourced custom assets." Benefit: "Your page gets real images, icons, and a color palette picked for your brand, not stock photos everyone's seen." The feature is the mechanism. The benefit is the thing they'd pay for. Keep the feature as supporting detail, but make the benefit the headline of each section.

You don't have to delete every feature. Just make sure each one is attached to a reason the reader should care.

Add social proof, and make it specific

Strangers don't trust your claims. They trust other people's claims about you. Social proof is the cheapest trust you can add, and most pages do it badly with vague, anonymous quotes.

Weak: "Great product, highly recommend! — A happy customer." That's worth nothing; you could have written it yourself, and the reader assumes you did. Strong: "We cut our invoicing time from 3 hours a week to 20 minutes. — Mara K., freelance designer." A name, a role, and a number make it believable.

If you're pre-launch and don't have testimonials yet, use what you have: a count of people on your waitlist, logos of tools you integrate with, a founder's relevant credential, or a screenshot of a real conversation. Don't fabricate testimonials. Made-up proof reads as fake, and one whiff of fake poisons the whole page.

Pick one primary call to action

Every page should ask for exactly one thing. When you offer five buttons, the visitor has to make a decision, and a confused visitor does nothing. Choose the single action that matters most right now, and make that button the loudest element on the page.

Make the CTA describe what happens next. "Submit" is dead. "Start my free page" or "Get my early-access invite" tells the reader what they're getting. If you have secondary actions like "See pricing," keep them quiet and visually smaller. One bright button, repeated as the reader scrolls, beats a menu of equal options.

If you're launching before the product is fully ready, the one action is usually an email signup. The same clarity rules apply when you build a waitlist page that converts: one promise, one field, one button.

Cut the jargon, then run the 5-second test

Hype words are filler that makes your copy sound like every other page. Phrases like "next-generation solution" or "powerful platform" carry zero information. Delete them and the sentence almost always gets stronger.

Before: "A powerful, all-in-one solution that empowers teams to optimize their synergy." After: "One place for your team's tasks, files, and chat." Concrete nouns and plain verbs beat abstract adjectives every time. If you can't picture it, the reader can't either.

Then run the test that matters: show your page to someone who's never seen it, for five seconds, then ask what you do and who it's for. If they can't answer, your copy isn't ready, no matter how good it feels to you. Friends are too polite to be useful here, so push for the honest version. When you're ready to put it all together, Kalit Flow builds the page from a prompt, sources real custom assets, and publishes to a live URL, so you can iterate on the copy in plain language instead of fighting a page builder.

Here's the whole thing as a checklist:

1. Write a headline that says what you do and who it's for, plainly.
2. Use the subhead to add proof or explain how, not to repeat the headline.
3. Open with the customer's problem or desired outcome, not your features.
4. Turn every feature into a benefit by asking "so what?" until it lands.
5. Add specific social proof with names, roles, and numbers.
6. Choose one primary CTA and make it describe what happens next.
7. Cut jargon and hype, then run the 5-second stranger test.

Copy that converts is mostly copy that's clear. Say the true thing plainly, put the reader's problem first, and ask for one action.