Guides · 5 min read
The five-second first-impression test
Three questions that tell you whether your home page says what you think it says. Run it today with synthetic users or with the next person who walks past your desk.
Someone lands on your home page from a search result. They have never heard of you. Within a few seconds they decide whether they are in the right place. If they guess wrong, nothing else on the page matters, because they are already reaching for the back button.
The first-impression test checks that moment. It is the cheapest usability test there is, and the one teams skip most, because everyone inside the company already knows what the product does.
The three questions
Show the page for about five seconds, hide it, and ask:
- What is this? A shop, a service, a tool, a blog? For whom?
- What can I do here? Buy, book, compare, sign up, read?
- Where would you click first? And what do you expect to happen when you do?
Write the answers down word for word. Do not correct people or explain. The gap between their words and yours is the finding.
What good answers sound like
You do not need people to repeat your tagline. You need them to be roughly right. “It’s a place to rent bikes in Valencia, I’d click the prices” is a pass. “Some kind of cycling company, maybe tours?” is a fail, even if tours are mentioned somewhere on the page.
Common reasons for a fail:
- The headline describes a feeling (“Move freely”) instead of the offer (“Bike rental in Valencia, from €12 a day”).
- The largest element on the screen is a photo that does not explain anything.
- The main button uses internal words (“Explore the fleet”) instead of the visitor’s (“See bikes and prices”).
- On mobile, the useful part of the page starts below the first screen.
Run it with people
Ask three to five people who do not work with you. Friends of friends are fine, as long as they have not seen the site. Use their own phones if most of your traffic is mobile. Ten minutes in total is enough.
Run it with synthetic users
Every UX Surgery study starts with this test. Each profile looks at the page before doing anything and answers the same three questions. Profiles with no context, reduced vision or little time tend to fail first, which is exactly who you want to hear from. The answers are hypotheses: if several profiles misread the page the same way, check it with a couple of real people before rewriting everything.
What to change
Usually very little. Rewrite the headline so it names the offer, make the first button say what it does, and move the price or the main action above the fold on mobile. Then test again. If the answers improve, you are done with the first impression and can move on to the tasks.