A five-second test shows people a design for a few seconds, then hides it and asks what they remember. It tells you what makes an impression straight away: whether people can tell what the page is for, who it's from, and what the main message is.
That matters because people decide very quickly whether a page is relevant to them. If the key message doesn't land in the first few seconds, many visitors never get to the rest.
When to use it
- Homepages and landing pages: is it clear what you offer?
- Headlines and value propositions: which version is understood and remembered?
- Brand recognition: do people notice and recall the brand name?
- Ads, emails and campaign pages: does the main message survive a glance?
A five-second test isn't the right tool for checking whether people can find something or complete a task. For that, use a first-click test or a tree test.
How to run one
1. Pick the design
Use the screen exactly as people would first see it. Only the top of the page will be visible, so make sure what you care about is there. You can upload an image or capture a live page by its URL.
2. Set the context (optional)
A short scenario helps people look at the page with the right mindset, for example "Imagine you've just landed on this site while looking for a winter coat." Keep it neutral, so it doesn't give away what you want them to notice.
3. Choose how long to show it
Five seconds is the classic length and works for most pages. Use a little longer (7 to 10 seconds) for dense pages, or shorter for simple ones like an ad.
4. Write the questions
Two to four questions is plenty. People forget quickly, so ask the most important question first. Good ones:
- "What do you remember about what you just saw?" Open and unprompted, the best measure of what stood out.
- "What do you think this page is for?" Checks whether the purpose is clear.
- "Who do you think this page is from?" Checks brand recall.
- "What could you do on this site?" As multiple choice, to check understanding.
- "How clear was what this company offers?" As a 1 to 5 rating.
For open questions, decide in advance which words would show someone got the message, such as the brand name or the key benefit. In ClickTree these are called expected words, and the results show how many people mentioned them.
5. Collect responses
Aim for 30 to 50 people. Five-second answers vary a lot from person to person, so you need enough of them for patterns to emerge.
Reading the results
- Recall: the share of answers that mention the words that matter. Below about half means the message isn't landing.
- Most mentioned words: what people actually took away. If a secondary element dominates, it may be competing with your main message.
- "Don't know" answers: many of these suggest the page is too busy to take in quickly, or the key content is below the first screen.
- Choice and rating questions: a quick, comparable measure of understanding and clarity.
Common fixes
- Make the main message bigger, simpler or higher on the page.
- Remove or tone down elements that compete for attention.
- Say what you do in plain words, rather than a clever tagline people have to decode.
Then test again with the same questions, and compare.
You can create a five-second test in ClickTree on the free plan.