Why five seconds?
Most visitors don’t read a web page. They scan it and decide, quickly, whether it’s for them. Usability researchers at Nielsen Norman Group found that people read about 20% of the words on an average page, and that many leave within 10 to 20 seconds.
That means the first screen does most of the selling. Five seconds is a strict version of the first glance: if your offer isn’t clear by then, a large share of visitors never gets to the part of the page that explains it.
How to run the test
- Open your homepage on a phone, since that is where most first visits happen.
- Don’t scroll. Only what fits on the first screen counts.
- Look for five seconds, then look away.
- Answer the five questions below out loud, using only what you saw.
- Better still, ask someone who has never seen your business to do it.
The five questions, with examples
Each example below is a rewrite for a fictional business, to show the difference between a line that fails and one that passes.
1. Who is it for?
Your ideal customer should recognise themselves in the first line. “Businesses” and “everyone” are not audiences.
FailsAccounting solutions for businesses of all sizes.
PassesAccounting for founders who’d rather not think about tax until it’s done.
2. What problem does it solve?
Name the problem in the words your buyers use, not the words your industry uses. Customers say “chasing late payments”, not “accounts receivable workflow”.
FailsStreamline your billing workflow with powerful tools.
PassesGet paid on time without writing the awkward email.
3. What do I get?
Say the outcome plainly: what changes for the reader after they buy. Features describe your product; outcomes describe their week.
FailsWe bake fresh cakes for all occasions using quality ingredients.
PassesThe birthday cake they’ll still talk about next year.
4. Why you?
Give one specific reason to choose you over the next open tab. “Quality”, “passion” and “experience” don’t count, because every competitor says them too. A number, a guarantee, a method or a narrow speciality does.
5. What do I do next?
One clear button with a verb, visible without scrolling. “Learn more” and “Submit” are weak. “Plan your cake”, “Book a call” and “Get the audit” tell the reader what happens when they click.
How to score it
| Questions passed | What it usually means |
|---|---|
| 5 of 5 | The first screen is doing its job. Look further down the page for leaks. |
| 3 or 4 | Close. One or two fixes, usually the headline or the button, will make it land harder. |
| 0 to 2 | The page is likely losing many visitors before they scroll. Start with the headline. |
What to fix first
Fix in this order, because each one decides whether the reader sees the next:
- The headline. It decides whether anyone reads on. Make it about the reader’s problem or outcome, not about you.
- The button. One action, with a verb that says what happens.
- The line under the headline. Use it to answer “who is it for” and “why you” if the headline couldn’t.
You can check a headline right now with the free headline scorer on the Sphota homepage. It flags filler words and runs five quick checks as you type.
Questions
Why five seconds?
Research by Nielsen Norman Group found that many visitors leave a web page within 10 to 20 seconds, and that people read about 20% of the words on an average page. Five seconds is a strict version of that first glance: if the first screen does not explain the offer by then, many visitors are already deciding to leave.
Who should run the test?
Someone who has never seen your website. You already know what you sell, so you will fill gaps the page leaves open. A friend outside your industry is ideal.
How many questions must the page pass?
All five is the goal. Failing two or more usually means the first screen is costing you enquiries, and the headline is the place to start.
Does this work for landing pages too?
Yes. A landing page has one job, so it should pass the test even more easily than a homepage. Run it on any page that gets paid or search traffic.
Related: What is conversion copywriting?