We Ran the 5-Second Test on 30 Startup Homepages. Here's What Almost All of Them Got Wrong.

The "five-second test" is an old idea in UX research: show someone a homepage for five seconds, take it away, then ask what they remember. If they can't say what the company does, who it's for, or what to do next, the page has failed at its most basic job — regardless of how it looks.

We wanted to know how often that failure actually happens on real startup homepages, not hypothetical ones. So we ran the test on 30 live startup sites — a mix of early-stage SaaS, local service businesses, and DTC brands — using five people who had never seen any of the sites before. Each person saw a homepage for five seconds, then answered three questions: What does this company do? Who is it for? What am I supposed to do next?

The topline numbers

Across 30 homepages and 150 individual test responses, the pattern was more consistent than we expected going in:

That last number is the one worth sitting with. Four out of five homepages in this sample left a first-time visitor unsure of at least one of the three most basic things a page can communicate — what, who, and what next. None of these were bad businesses. Several had real traction, real customers, and real revenue. The website simply wasn't doing the one job it exists to do in that first moment.

Three failure patterns that showed up again and again

1. The "vibe headline"

The single most common failure was a hero headline built entirely around tone rather than substance — clever, on-brand, and completely unclear to anyone outside the company. Phrases built around abstract outcomes ("Built for what's next," "Where growth happens") tested almost identically to a blank hero section. Testers could describe how the line made them feel. They couldn't say what the product was.

2. Audience left to inference

Very few homepages stated who they were for. Most assumed the visitor already knew, or would figure it out from context — logos, imagery, pricing tier names. But a first-time visitor doesn't have that context yet. Sites that explicitly named their audience in the first screen ("for solo founders," "for clinics with 5+ locations") were answered correctly almost every time. Sites that didn't were guessed at, often incorrectly.

3. Competing calls-to-action

The 37% of pages with no clear next step almost always had the opposite problem of having none: too many. "Book a demo," "Start free trial," "See pricing," and "Watch video" all present with equal visual weight in the same screen. Testers didn't fail to find an action — they failed to identify which one mattered, and in that hesitation, most simply picked none.

The one variable that predicted a pass

Across the sample, the single strongest predictor of passing all three questions wasn't visual polish, load speed, or even industry — it was whether the homepage stated its value proposition as a literal sentence, in visible text, above the fold. Not implied through imagery. Not embedded in a tagline. An actual sentence a human could read and repeat back.

The six homepages that passed cleanly all shared this. Several were visually simpler than pages that failed. Design quality and clarity turned out to be only loosely related — plenty of polished, expensive- looking sites failed, and a few plain ones passed easily.

Why this keeps happening

Most founders and teams have read their own homepage hundreds of times. Context that's obvious to them — what the product does, who it's for — gets mentally filled in automatically, which makes a vague headline feel perfectly clear from the inside. The gap only becomes visible when someone with zero context looks at the page cold. That's exactly the position every new visitor is in, and exactly the blind spot this kind of test is built to catch.

A homepage doesn't get graded on how good it looks. It gets graded on how fast a stranger understands it — and most homepages are graded by strangers, not by the people who built them.

The fix, in nearly every one of these 30 cases, wasn't a rebuild. It was a rewrite of the first screen: one literal sentence about what the business does, one clear signal of who it's for, and one action given more visual weight than the rest. Small changes, tested the same unforgiving way they'll actually be encountered — in five seconds, by someone who owes the page nothing.