When Your Copy Assumes Reading You Didn't Earn

There’s an assumption baked into most website copy that nobody ever states out loud: the assumption that the visitor will read it. Read it in order, from the top, sentence by sentence, the way the writer wrote it. Under this assumption, all the standard moves make sense. You build context in the first paragraph, develop the argument in the second, and land the payoff in the fourth. You introduce a term early and rely on it later. You save the best point for the end, like a good essay.

The assumption is false. Watch anyone use the web and you’ll see it immediately. Visitors don’t read pages; they raid them. They land, flick the thumb twice, let their eyes catch on a heading or a bolded phrase, sample half a sentence, and either commit or leave. Reading, real linear reading, is not the default behavior. It’s a reward the page earns after the raid goes well. Most copy is written for a reader the page hasn’t earned yet.

This creates a specific and common failure mode that I’d call unearned assumption of attention. The page is well written by every classroom standard. The argument is sound, the prose is clean, the structure builds. And it fails completely, because its virtues only exist for someone reading linearly, and the actual visitor experienced it as a random sample of three fragments from the middle. The careful setup in paragraph one was never read, so the payoff in paragraph four referenced context the visitor doesn’t have. The pronoun whose antecedent lived two sentences back pointed at nothing. The killer point saved for the end was saved for no one.

Once you accept that skimming comes first, the craft changes. The goal of the top of a page isn’t to begin an argument. It’s to survive a raid: to make sure that a visitor who reads ten percent of the page, chosen mostly by visual prominence, still comes away with the right idea. That means the first line of every section has to carry the section’s point, because it may be the only line of the section anyone reads. It means each paragraph should be comprehensible on its own, without the paragraph before it, which is why short paragraphs beat long ones; they’re not just less intimidating, they’re more self-contained. It means the important claim goes first and the elaboration after, the inverted pyramid newspapers used for a century, for the same reason: their readers skimmed too.

It also changes what earns the deeper reading. A skimmer decides to slow down when a fragment they sampled turns out to be worth it. Concrete beats abstract in this economy every time. “Set up in one afternoon” stops a skimming eye; “streamlined onboarding experience” slides right past it. Numbers stop the eye. Named specifics stop the eye. Abstractions are invisible at skimming speed, which means a page written entirely in abstractions is, for most visitors, functionally blank.

None of this means dumbing anything down, and this is the misunderstanding that makes writers resist it. Writing for skimmers isn’t writing for stupid people. Your smartest visitors skim hardest, because they’ve seen the most websites and have the best-trained filters. An engineer evaluating your product skims your docs. An executive skims your pricing. Writing that survives skimming is harder to produce than linear prose, not easier, because every sentence has to work locally and globally at once. It’s the difference between a building you can only enter through the front door and one that’s coherent from any entrance.

There’s a humbling way to test your own pages, and it’s the idea behind how GazeSite reads a page when it audits one: look at the rendered screenshot the way a visitor does, catching only what’s visually prominent, and ask what the page appears to say. Take your own homepage and read only the headings, the button labels, and the first line of each paragraph. That’s roughly the raid. Does the result make your case? On most pages the honest answer is no. The skimmable layer says something generic, and everything persuasive lives in the middles of paragraphs, invisible at speed.

The fix isn’t to write less. It’s to stop treating attention as a starting balance and start treating it as something the page earns line by line. Assume you have one sentence of attention. Use it to earn the next one. A page written under that assumption works for the skimmer, and it works better for the rare linear reader too, because prose that survives being sampled is prose with no dead weight anywhere. The discipline the skimmer forces on you was good discipline all along.