The Homepage Squint Test

There’s a trick painters use to check a composition: step back and squint. Squinting strips out the detail and leaves only the big shapes and the contrast, so you can see what the picture emphasizes rather than what it contains. The same trick works on homepages, and it finds problems that no amount of careful reading will, because the problems it finds aren’t in the words. They’re in what the page emphasizes, which is a separate channel of communication, and one that most site owners have never once checked.

Here’s the procedure. Open your homepage and squint until you can’t read any text. Now look at what’s left. Where does your eye go first? What’s the biggest thing? The brightest thing? How many things are competing to be looked at? Then, still squinting, guess what you’d click. When you open your eyes, you’ll know two things: what your page is actually saying, and whether that’s what you meant it to say. The two are different more often than you’d think.

The reason this works is that visitors process a page in the same order the squint reveals. Before anyone reads a word, they’ve taken in the layout: what’s large, what’s colorful, what’s isolated by space, what’s repeated. This pre-reading pass assigns importance, and the reading that follows mostly obeys those assignments. Big things get read; small dense things get skipped. So the visual hierarchy isn’t decoration around your message. It’s a message of its own, delivered first, and if it contradicts the words, the hierarchy usually wins. A page can say “the free scan is the main thing” in its copy while saying “this stock photo of a woman pointing at a whiteboard is the main thing” in its hierarchy, and the visitor receives the second message.

The squint test’s most common finding is exactly that: the most visually dominant element is decorative. An abstract illustration, a looping animation, a photo that took days to choose, occupying the position and scale of the most important thing on the page while containing no information. I understand how it happens. Images are what people fiddle with, because fiddling with images is more fun than rewriting a value proposition. But emphasis is a budget. Every unit of visual weight you spend on decoration was taken from something that had a job to do.

The second most common finding is a tie. You squint and see four blobs of equal weight, arranged with perfect symmetry, none of them winning. Designers sometimes produce this deliberately, because balance looks composed, and screenshots of balanced pages look good in portfolios. But a page where everything is equally emphasized has no hierarchy at all; it’s a page that refuses to have an opinion about what matters. The visitor, offered four equal claims on their attention, must do the prioritizing themselves, and prioritizing is work, and strangers don’t do work. Somewhere in your page there is a most important thing. The squint should find exactly one winner, and it should be that.

The third finding is the invisible button. Squint, and the primary call to action vanishes into the page, the same tone as everything around it, while some minor element, a cookie banner, a chat widget, a partner logo, blazes away. This one is insidious because at full attention the button is obviously findable, and everyone who works on the site views it at full attention. But strangers don’t grant a page full attention until it has earned some, and during that low-attention pass, contrast is the only thing that gets through. If the action you most want is not among the few things a squint reveals, then for a meaningful fraction of visitors it isn’t on the page at all.

What I like about the squint test is that it’s a null test, like listening to a mix on cheap speakers. It doesn’t tell you if your page is good. It tells you whether the emphasis agrees with the intent, and it does so in ten seconds, with no tools, no meetings, and no expertise required. It also settles arguments unusually well. Debates about design are usually taste against taste, and the loudest taste wins. But “squint and tell me what’s biggest” has an answer that two people can agree on even when they agree on nothing else. It converts an aesthetic argument into an observation.

There’s a blunter cousin of the test worth doing too: blur, don’t just squint. Take a screenshot of your homepage and blur it heavily, then look at the shapes on their own. When I was building GazeSite I spent a long time looking at homepage screenshots, because the scanner captures one for every page it audits, and I started blurring them almost by reflex. It changed how I think about clarity. I used to treat clarity as a writing problem, and mostly it is; plain words beat fog, and no layout rescues an empty sentence. But the words only get their chance if the hierarchy delivers a reader to them. Layout is the usher; copy is the play. When a page fails, everyone blames the play. Squint first. Half the time the play is fine and the usher is seating people facing the wrong way, and unlike rewriting a play, turning the chairs around is an afternoon of work: shrink the decoration, let one thing win, and make the button look like the point of the page. Then open your eyes and read the words, which is what your visitors, at last, will do too.