The Test Most Founders Fail and the Data Behind It
Microsoft research from 2015 put the average attention span on a webpage at 8 seconds. Jakob Nielsen's lab data is harsher: visitors decide whether to stay or leave within 5 seconds, and most of that decision happens above the fold.
I run the 5-second test on every positioning grade I do. Roughly 70% of indie SaaS homepages fail it. Not because the product is bad. Because the homepage is doing 4 jobs at once and none of them well.
Here's what the test actually measures, how to run it on your own site without spending a dollar, and how to fix the 3 most common failures.
What the 5-Second Test Actually Measures
The test is simple. Show someone your homepage for exactly 5 seconds. Take it away. Then ask three questions:
Question 1: What Does This Product Do?
Not "what category is it in." What does it actually do for the user? "It tracks my SaaS metrics." "It cleans my email list." "It generates social posts from my blog." Specific verbs. Specific objects.
If the answer is "I think it's some kind of analytics platform?" you failed.
Question 2: Who Is It For?
Not "businesses." Not "teams." A specific role, stage, or situation. "Solo SaaS founders." "Series A engineering managers." "Shopify store owners under $50K MRR."
If the tester says "it didn't say" or "everyone, I think," you failed.
Question 3: Why Would Someone Use It?
The benefit. The outcome. The reason this exists. "To find out why their MRR dropped." "To stop spending 4 hours a week on social media." "To get their first 100 paying customers."
If the answer is "to be more productive" or "to grow their business," you failed. Those are platitudes, not benefits.
The Bar
To pass, your homepage needs to communicate all three (what, who, why) in 5 seconds. Not 30 seconds. Not "if I read the whole hero section." Five seconds. That's the visitor experience.
How to Run the Test on Your Own Site
You don't need a research budget. You don't need a tool. You need 5 minutes and one of these three methods.
Method 1: The Friend Test (Free, Highest Quality)
Find someone who matches your ICP. A founder friend, a colleague at the right size company, a user from a competitor's product. Send them a screenshot of your homepage above the fold. Tell them to look for 5 seconds, close the image, and answer the three questions.
This is the highest signal version because the tester actually fits your audience. The downside: friends are nice. They might soften answers. Ask them to be brutal. Promise you won't get hurt.
Method 2: The Stranger Test (Cheap, Highest Volume)
Use UsabilityHub, Maze, or just post in an indie hacker Slack offering to swap 5-second tests with another founder. You test their homepage, they test yours. Honest feedback because you don't know each other.
Cost: $0 to $50. Time: under an hour. Run with 5 testers minimum. If 3 of 5 fail any question, that question is a real problem.
Method 3: The AI Test (Free, Instant)
Take a screenshot of your homepage above the fold. Drop it into ChatGPT or Claude. Prompt: "Look at this image for the equivalent of 5 seconds. Without reading carefully, just glancing. Then answer: what does this product do, who is it for, and why would someone use it. Be honest if anything is unclear."
AI tests aren't perfect (they read everything), but they're useful for catching when your copy is genuinely vague vs. when humans just didn't have time. Run this test in addition to method 1 or 2, not instead of.
Method 4: The Self Test (Free, Worst Quality)
You can't run the test on yourself. You wrote the copy. You know what it means. You'll fill in the gaps in your own head. Don't bother. Use methods 1, 2, or 3.
5 Examples of Homepages That Pass vs. Fail
Pass: A Hypothetical Cold Email Tool
Headline: "Book 10 sales calls a week without your domain getting blacklisted."
Subhead: "Cold email infrastructure for B2B SaaS founders doing their own outbound."
What: Cold email tool. Who: B2B SaaS founders doing outbound. Why: Get sales calls without deliverability problems. Passes all three in 5 seconds.
Fail: A Hypothetical "Modern Communication Platform"
Headline: "The future of work, today."
Subhead: "Powerful collaboration for modern teams."
What: Some kind of communication tool? Who: "Modern teams" (everyone). Why: Future of work (meaningless). Fails all three.
Pass: A Hypothetical Churn Analytics Tool
Headline: "See exactly which onboarding step kills your activation rate."
Subhead: "Built for B2B SaaS founders between $5K and $50K MRR."
What: Onboarding analytics. Who: Specific MRR-stage founders. Why: Find the broken step in activation. Passes.
Fail: A Hypothetical "AI-Powered Productivity Suite"
Headline: "AI that gets work done."
Subhead: "Boost your team's productivity with intelligent automation."
What: AI does something? Who: "Your team" (no role specified). Why: "Boost productivity" (vague). Fails all three. This is the most common failure pattern in 2026.
Pass: A Hypothetical Indie Hacker SEO Tool
Headline: "Rank your indie SaaS for 50 long-tail keywords in 90 days."
Subhead: "Specifically built for solo founders without an SEO budget."
What: SEO tool with specific outcome. Who: Solo founders. Why: Rank for keywords without spending. Passes.
The 3 Most Common 5-Second Test Failures
I've graded 240+ homepages this year. The failure modes cluster into 3 categories. Diagnose which one you have, then fix it.
Failure 1: The Category Fog
Symptom: Tester can't tell what the product does. Says "I think it's some kind of platform?"
Root cause: You used a category word ("platform," "suite," "solution") instead of describing what the product does. "Marketing platform" tells the reader nothing. "Email tool that warms cold outreach" tells them everything.
Fix: Replace the category word with an action verb. Bad: "AI marketing platform." Better: "AI that writes your weekly newsletter from your blog posts." Specific. Concrete. Passes the test.
Failure 2: The "For Everyone" Trap
Symptom: Tester can't tell who the product is for. Says "I guess businesses?" or "everyone."
Root cause: You hedged on the audience. You wrote "for teams," "for businesses," or "for modern companies." None of those describe a real human.
Fix: Name the role, the stage, the company size. Generic positioning kills conversion. "For B2B SaaS founders at $5K to $50K MRR" tells the reader instantly if they're in the right room. The wrong reader leaves. The right reader stays. Both outcomes are wins.
Failure 3: The Benefit Vacuum
Symptom: Tester can describe what the product does and who it's for, but can't say why anyone would use it. Says "I guess to be more productive?"
Root cause: You described features instead of outcomes. "Real-time analytics dashboards." "Automated workflow triggers." Those are features. The benefit is what the customer gets.
Fix: Reframe each feature as the outcome it delivers. "Real-time analytics" becomes "find out why MRR dropped before your next board meeting." "Automated workflows" becomes "stop manually copying leads from your form to your CRM." Outcome over feature. Always.
How to Fix Each Failure (with the BELT Framework)
The 5-second test diagnoses where your homepage is leaking. The BELT framework tells you whether your fixes will compound.
If You Failed on "What"
Your headline is doing description, not promise. Use the formula from the 30-minute hero headline playbook. Specific verb plus specific outcome plus specific audience plus specific timeframe. Apply BELT's B (behavior): does your headline name a behavior the reader already does? Attach the product to that behavior.
If You Failed on "Who"
Your audience description is too broad. Apply BELT's E (enduring problem): who has this problem regularly? Pick the specific stage and role of that person. Then put it in the subhead. "For [specific role] at [specific stage]." If you don't know who your ICP is, run the 7-day ICP sprint first.
If You Failed on "Why"
Your copy is feature-heavy. Apply BELT's L (lock-ins) and T (transient distractions): what's the durable benefit your product creates that the customer can't get elsewhere? Put that benefit in the headline. Move the features below the fold or onto a feature page.
FAQ
How many testers do I need for a 5-second test?
5 minimum. Less than 5 and you can't tell signal from noise. With 5, if 3 fail any question, that question is a real problem to fix. With 10, you get statistical comfort. Above 10 is overkill for early-stage SaaS.
What's the difference between the 5-second test and a usability test?
The 5-second test measures comprehension at first glance. Did the visitor understand what, who, and why before deciding to bounce? A usability test measures whether someone can complete a task on the site (sign up, navigate, etc.). The 5-second test runs first. If you fail it, the usability test doesn't matter.
Should I run the 5-second test before or after I redesign?
Both. Run it before to know what's broken. Run it after to know if your redesign fixed it. If your post-redesign 5-second pass rate is the same as before, you redesigned for the wrong reasons.
Can my homepage pass the 5-second test if it's text-heavy?
Yes, if the H1 and subhead are clear. The eye reads the H1 in roughly 1 second. The subhead in another 2. The CTA in another 1. That's the 5-second experience. If those three elements communicate what, who, and why, the rest of the page can be as long as it needs.
Does the 5-second test work for landing pages and product pages too?
Yes. Any page where you're trying to convert a stranger into a next action benefits from the test. The questions adapt slightly: for a feature landing page, "what does this feature do, who needs it, why does it matter" is the right framing.
What to Do This Week
Take a screenshot of your homepage above the fold. Send it to 5 people in your ICP. Ask the three questions. Track the answers. If 3 of 5 fail any question, you have a fixable problem and now you know exactly where it lives.
If you want a structured grade with specific rewrites, run your homepage through the Growth Pigeon Positioning Grader or get a full clarity map teardown with copy fixes for every section that fails the test. The BELT framework covers what to optimize after your homepage starts converting.