What a first-time visitor needs to know
Someone who lands on your page from a search, a launch post or a directory listing has never heard of you. Before they scroll, they want quick answers to a few questions: what is this, is it for me, what do I get, why this one, and what do I do next. If the first screen doesn’t answer them, most people leave without telling you why. The grader asks those questions of your page’s text, one by one.
The nine checks, with examples
The examples below are made up to show the difference. They aren’t real products.
- Says what the product is. The headline or the line under it names the kind of product and the job it does. Weak: “The future of work, reimagined.” Better: “Invoicing software that chases late payments for you.”
- Says who it’s for. A role, industry or situation the right visitor recognises. Weak: “For teams of all sizes.” Better: “For freelance designers who bill clients monthly.”
- States a concrete outcome. What changes for the customer, not what the software contains. Weak: “Powerful automations and smart reminders.” Better: “Get paid on time without writing awkward follow-up emails.”
- Says why it’s different. What you do that the usual alternative doesn’t, including a spreadsheet or doing it by hand. Weak: “The best invoicing tool.” Better: “Unlike accounting suites, it only does one thing: follows up until the invoice is paid.”
- One clear call to action. One main next step, with the strongest button. Weak: “Start trial”, “Book a demo”, “Join the community” and “Download the app” side by side. Better: one “Start free trial” button, repeated lower down, with a quieter text link for demos.
- Shows social proof. Evidence that other people use it: a short quote with a real name, customer logos you have permission to show, ratings from a review site, or a usage figure you can back up. Weak: “Loved by users.” Better: a named quote from a customer that says what changed for them.
- Handles a main objection. The worry that stops people clicking: price, risk or effort. Weak: nothing near the button. Better: “Free for 14 days, no credit card, set up in one sitting”, if that’s true for your product.
- Specific headline. A headline that could only belong to your product. A useful test: could a competitor, or a company in a different industry, paste it onto their homepage unchanged? “Work smarter, not harder” fails. “Chase unpaid invoices without the awkward emails” passes.
- Overall clarity. A five-step score from “confusing” to “very clear” for the headline, subheadings and opening text read together, as a stranger would read them.
How the grade works
We fetch your page once, the way a crawler would, and pull out a compact summary of what a visitor sees first. Jev 1.13, TypeSafe’s decision model, answers all nine questions about that summary in a single call, each with a probability. The grade out of 100 is a weighted average of those probabilities. “Says what it is” and overall clarity count the most, and “handles an objection” the least, because a page can work without it while a page that doesn’t say what it is rarely does.
For every check that isn’t a likely yes, Claude Haiku 5.5 writes a short fix based only on the text of your page. It’s a starting point for your own edit, not copy to paste in blind.
What the grader can’t tell you
- How your page looks. Layout, contrast, images and video are invisible to it. A product screenshot can answer “what is it” for a human even when the text doesn’t.
- How your real visitors behave. The probabilities are a model’s reading of your copy, not conversion data. Watch what visitors do, and ask a few people who fit your audience to explain the page back to you.
- Whether it ranks. For titles, descriptions and indexing, use the meta tag checker or the SEO checker.
After you fix the page
Test a few headline options in the headline analyzer, then run the grader again. When the page is ready for visitors, launch it free on LaunchRanked to get a permanent product page and a followed link with our badge.