Landing page copy converts when it uses the customer's own words about their problem rather than your words about your product. The procedure is to mine your interview transcripts for exact phrases, draft the page bottom-up from the offer, write the headline last, then cut every adjective that a competitor could also claim. The most common structural failure is a headline naming the product category instead of the outcome — it's twenty minutes to fix and the single largest lever on the page.
Open your interview notes, your support inbox, the reviews of the tool your prospects use now, and the threads where they complain. Copy out the exact sentences — verbatim, misspellings and all — where someone describes the problem.
You're looking for three things:
Paste them into a document. This is your copy source, and every important line on the finished page should be traceable back to something in it.
The reason this works isn't a trick. Copy written from your own vocabulary describes your product to yourself. Copy written from theirs describes their situation to them, and the second one is what makes a stranger think this is for me in the two seconds before they leave.
Write the page in the reverse of the order it's read. The headline last, because you can only compress what already exists.
The offer. One sentence: what they get, what it costs, what happens when they click. If this is fuzzy, nothing above it can be sharp.
The proof. Whatever you actually have: a number, a customer quote, a specific result, a screenshot of the thing working. Real and small beats impressive and vague. If you have no proof yet, say what you're doing about that — "we're taking our first ten customers" is honest and converts better than fabricated social proof, which experienced buyers detect instantly.
The objections. List the five reasons a person who wants this wouldn't buy. Too expensive, too new, migration pain, doesn't integrate, might disappear. Answer each in one line, on the page. Objections you don't handle get handled in their head, and you lose that argument because you're not in the room.
Then the structure: problem → outcome → how it works in three steps → proof → price → objections → CTA. It's boring, and it's boring because it works.
Now compress. Twenty headlines, no filtering. Most will be bad; the exercise is what gets you to number fourteen.
Test each against three questions:
Then write the subhead to do the job the headline couldn't: who it's for, and how it works, in one sentence.
And add one line naming who it's not for. It costs a sentence, and it's the fastest way to make the right person feel found — a page that promises to serve everyone reads as serving nobody in particular.
Read the draft and delete every word that would survive being pasted onto a competitor's page.
The standard casualties: powerful, seamless, intuitive, innovative, cutting-edge, robust, enterprise-grade, best-in-class, revolutionary, effortless. These are free to claim, therefore worthless as signal, and their presence tells a reader you had nothing specific to say.
Replace each with something specific or nothing at all. "Powerful analytics" → "See which of your posts brought paying customers." "Seamless integration" → "Connects to Stripe in one click." "Effortless setup" → "Live in four minutes."
Then run three more passes:
Then paste the URL into the free roast below. It reads the live page as a first-time visitor would and ranks the three highest-leverage fixes. Do the first one. Ship. Come back next week.
One last pass before you publish: read the whole page out loud. Every sentence you stumble over is one a reader will stumble over silently and then leave. It takes four minutes and catches more problems than another hour of editing on screen, because your ear notices what your eye has already learned to skip.
kitty.build writes and publishes the page from your validated positioning, versions every revision so a rewrite that loses conversion is one click to undo, and hosts it on your own domain and Cloudflare account.
Feed her an idea