The Clarity Checklist for a New Landing Page

Before a pilot takes off, they run a checklist. Not because they’ve forgotten how to fly, but because the failures that kill you are the obvious ones you skip when you’re confident. Landing pages are like this. Nobody launches a page thinking it’s confusing. The confusion is invisible to the person who wrote it, which is exactly why it needs a checklist rather than a feeling.

Over years of shipping pages and then auditing other people’s, I’ve settled on a sequence of questions I ask before anything goes live. They’re all versions of one master question: will a stranger get this in five seconds? But the master question is too abstract to answer directly, the same way “is the plane safe?” is. You have to decompose it.

The first question is whether the headline states what the thing is. Not what it feels like, not what it enables, but what it is. “Effortless growth for modern teams” fails this test. “Email marketing for small stores” passes. I read the headline alone, covering everything else on the page with my hand if I have to, and ask whether someone could repeat back to me what’s being sold. If the headline only works with the subheadline’s help, the subheadline is the real headline and the current one is decoration.

The second question is whether the page names its audience. A stranger deciding whether to stay is really asking “is this for people like me?” and the fastest way to answer is to say so. Pages fear this because naming an audience excludes everyone else, but the exclusion happens anyway; it just happens later and more expensively, after a mismatched visitor has wasted their time and yours. Better to lose them at the headline than at the signup form.

The third is whether there’s exactly one obvious next step. Not zero, which happens more than you’d think, and not five. When a page offers “Start free trial,” “Book a demo,” “Watch video,” “Read docs,” and “Contact sales” with equal visual weight, it has delegated a decision to the person least equipped to make it. The visitor doesn’t know which door is right for them. You do. Pick one, make it prominent, and demote the rest. And the button should say what happens when you press it. “Get started” is a shrug. “Run a free scan” is a sentence.

The fourth question is about vocabulary. I go through the page hunting for words that only insiders use. Every industry has them, and they’re invisible to insiders precisely because they’re the water you swim in. “Omnichannel.” “Orchestration.” “Solutions.” The test I apply is whether the word would survive being said out loud to a customer at a dinner table. Most jargon wouldn’t. Plain words feel less impressive to write, but impressiveness is not what a confused stranger rewards. They reward recognition, the small click of “oh, that’s my problem, in my words.”

The fifth is the mobile question, which is really the fold question in disguise. On a phone, the fold is brutal: often just a headline and part of an image. So I load the page on an actual phone and ask whether the visible words, alone, orient a stranger. If the headline is fine but the screen is dominated by an illustration of abstract shapes, then on mobile my page effectively opens with abstract shapes. The layout that seemed balanced on a designer’s monitor becomes an ambiguity machine on the device most visitors use.

The sixth question is the one people skip: what does the page look like before it finishes loading, and what does it look like when something breaks? A hero headline rendered by JavaScript is a blank page for however long the script takes. An image that carries the meaning of the page carries nothing if it fails to load and the alt text is empty. Clarity has to survive the unglamorous conditions, because that’s where a meaningful share of first impressions actually happen.

The last item isn’t a question but a procedure: show the page to someone who has never seen it, for five seconds, then take it away and ask them what the product does and who it’s for. This test is old, it’s low-tech, and it’s undefeated. It fails pages that pass every internal review, because internal reviewers can’t unknow the answer. One honest stranger outweighs a dozen colleagues.

I’ll be honest about why this list exists in the form it does: I got tired of running it by hand, so I built GazeSite to run checks like these automatically, looking at a page the way a first-time visitor would, on desktop and phone, and flagging where comprehension breaks. But the tool is the smaller point. The larger point is that clarity failures are systematic, which means they’re checkable, which means shipping a confusing page is mostly a choice to skip the checklist. Pilots don’t skip theirs. The stakes here are lower, but the discipline transfers. Run the questions. Fix what fails. Then launch.