PitchKitchen Frameworks
What Is the Three Questions Test? The 5-Second Homepage Diagnostic
The fundamental premise
A buyer gives your homepage five seconds. Not five minutes. Five seconds.
That's the window before the brain decides whether to keep reading or click away. Inside that window, three things have to happen. The visitor has to recognize themselves. They have to recognize their problem. And they have to recognize that the company in front of them has a point of view about that problem.
If any one of those three fails, the page fails. The buyer doesn't say so out loud. They just leave.
Most B2B founders test their homepage with a friend or a board member who already knows the company. The friend reads slowly. The friend asks helpful questions. The friend says, “looks great.” That's not a test. That's a hug.
The Three Questions Test removes the hug. It forces the founder to find a stranger, hand them the page, give them five seconds, and ask three questions. The answers expose what the founder couldn't see from inside the building.
The pattern repeats across more than two hundred B2B audits. The founder thinks the homepage explains the company. The stranger says, “I think it's some kind of platform?” The gap between those two readings is where the pipeline dies.
Definition
The Three Questions Test is a 5-second homepage diagnostic developed by Greg Rosner at PitchKitchen. A stranger looks at the homepage for five seconds and answers three questions: who is this for, what problem does it solve, and what is the company's point of view about that problem. If the stranger can't answer any one of the three, the homepage fails the test. Used by growth-stage B2B founders to expose the gap between what the homepage says and what the buyer hears.
Created by Greg Rosner. Proprietary diagnostic of PitchKitchen. Used alongside NarcScore as the experiential cousin of the structural pronoun audit.
Why the Three Questions Test exists
By 2026, every B2B founder has been told their homepage matters. Most have spent serious money on it. Few can prove it's working.
Analytics show traffic. Heatmaps show clicks. Session recordings show bounce. None of those tools answer the question the founder actually needs answered: when a stranger lands here, do they understand what we do and who we do it for?
Greg Rosner built the Three Questions Test after watching the same conversation repeat across hundreds of founder calls.
A founder describes the company in plain language during the call. The description is clear, specific, customer-grounded. The founder names the buyer's pain in the buyer's own words. Then the founder pulls up the homepage and shares the screen. The homepage says none of what the founder just said.
The founder sees the gap in real time. It's painful. It's also fixable. The Three Questions Test is the cheapest, fastest way to surface that gap before another quarter of pipeline gets burned.
It takes fifteen minutes, three strangers, and zero budget. Founders run it every quarter against their own site and against three competitors. The pattern that emerges drives the rebuild.
The core mechanic behind the test
Find a stranger. Show them the homepage for five seconds. Ask three questions.
That's the whole test. The discipline is in the constraints.
The stranger has to actually be a stranger. Not a friend. Not an employee. Not someone in the industry. Ideally a person who matches the rough profile of the buyer but has never heard of the company.
The five seconds has to actually be five seconds. Cover the page after the count. Don't let them scroll. Don't let them re-read. The brain decision happens fast in real life. Replicate the real condition.
The three questions have to be asked in this order:
- Who is this for?
- What problem does it solve?
- What's the company's point of view about the problem?
The order matters. Question one tests recognition (do they see themselves on the page). Question two tests relevance (do they see their world on the page). Question three tests differentiation (do they see the company's distinct stance on the page). Most homepages fail somewhere between two and three. The third question is the cliff.
Run the test against three strangers. Write down each answer verbatim. Compare the three answers. If the three strangers describe three different companies, the homepage has no center of gravity. If the three strangers all describe a generic “platform for growing teams”, the homepage has the wrong center of gravity. If the three strangers all describe a specific buyer, a specific pain, and a specific stance, the homepage is doing its job.
The three questions, unpacked
Each question maps to a layer of buyer comprehension. Each layer can fail independently.
Who is this for?
The buyer has to recognize themselves above the fold. Not as a generic 'modern team' or 'ambitious leader' but as a specific role with specific stakes. A CRO at a Series B SaaS company is not the same buyer as a CFO at a hospital system. The hero copy has to name the buyer specifically enough that the right person feels seen and the wrong person self-selects out.
Failure signal: the stranger says 'growing companies', 'businesses', 'teams'. Vague answers mean the homepage is trying to talk to everyone, which means it's talking to no one.
What problem does it solve?
The buyer has to recognize their problem inside the page. Not the company's framing of the problem ('digital transformation', 'operational efficiency', 'modern stack') but the buyer's lived frustration. The thing that's keeping the buyer up at night, costing them money, slowing their team down. Named in the buyer's words.
Failure signal: the stranger names a category instead of a problem. 'Some kind of CRM' or 'I think it's analytics' means the page lists features instead of naming a felt pain.
What's the company's point of view about the problem?
The buyer has to recognize a stance. The company believes something specific about how the problem should be solved, why the conventional approach is wrong, why this company's approach is different. Without a point of view, the company is just another option. With a point of view, the company becomes the trusted guide.
Failure signal: the stranger has no answer. They might describe what the company does, but they can't articulate why the company does it that way or why the way matters. This is the most common failure across B2B homepages.
Question three is where category leaders separate from category occupants. Pass all three and the homepage is doing its real job: pre-qualifying the right buyers and arming them to advocate inside their own organization.
Most founders test their homepage with a friend who already knows the company. That's not a test. That's a hug.
How the Three Questions Test differs from other homepage diagnostics
The B2B homepage diagnostic space is crowded. Heatmaps, session replays, conversion audits, brand audits, NarcScore. Each of them measures something. None of them measure the same thing.
Compared to heatmaps and session recordings: those tools tell you where the eyeballs went. The Three Questions Test tells you whether anything registered between the eyeballs and the brain. A buyer can read the entire page and still walk away with no usable answer to any of the three questions. The heatmap shows engagement. The test shows comprehension.
Compared to NarcScore:NarcScore measures the pronoun balance on the page. It's the structural audit. The Three Questions Test measures the buyer's recall after five seconds. It's the experiential audit. NarcScore answers “is this page about the customer?” The Three Questions Test answers “did the customer get it?” Both are useful. They catch different failure modes. Most engagements use both in tandem.
Compared to brand audits:brand audits measure visual and verbal consistency. The Three Questions Test measures whether the visual and verbal output actually communicates the brand to a stranger. A perfectly consistent brand can still fail all three questions because consistency isn't comprehension.
Compared to A/B testing: A/B tests optimize between two versions of an already-broken page. The Three Questions Test exposes whether the page is broken at all. Founders who run the test first stop wasting cycles A/B-testing variants of the wrong message.
The Three Questions Test's distinct contribution: it's the only homepage diagnostic that uses a real human, in real time, under real conditions. The output isn't a number. It's a quote. The founder hears the gap in the stranger's own words. That's harder to dismiss than any heatmap.
Question three is the cliff. Most B2B homepages can name what they do, but they can't name what they believe. That's the gap between a category leader and a category occupant.
Who the Three Questions Test is for
- B2B founders and CEOs who suspect their homepage isn't pulling its weight but haven't found language for why
- CROs and VP Sales who keep getting inbound leads that don't understand the product before the discovery call
- Marketing leaders who want a fast, defensible reason to push back on the agency's latest “we” copy
- Board members and investors doing portfolio diligence on a company's messaging before a fundraise
It's not for: brochure sites where the goal is awareness instead of conversion. The test is sharpest when the homepage is meant to move a buyer toward a meeting.
How the Three Questions Test is used in practice
The test shows up in three places inside the PitchKitchen workflow.
As a self-serve diagnostic. Founders run the test against their own homepage with three strangers from their network outside the industry. Fifteen minutes total. The output is three sets of three answers, written down verbatim. That's the baseline.
As a competitive scan. The same founder runs the test against three competitor homepages. Same strangers. Same five seconds. Same three questions. The contrast exposes whether competitors are clearer or muddier. Often the founder discovers a weaker product is winning because a clearer homepage.
As the opening move in the 90-Day Magnetic Messaging Sprint. Day one of every engagement includes a Three Questions Test against the current site, plus NarcScore. The before-state is captured. The same test runs on day ninety against the rebuilt homepage. Most engagements move from zero out of three to three out of three across the same panel of strangers.
The test isn't a fix. It's a mirror. The fix lives in the Magnetic Messaging Framework, which is the structured methodology for rebuilding the page so the answers come back right.
A real example
A $14M cybersecurity company came to PitchKitchen convinced their homepage was working. The hero said “Enterprise-grade AI for the modern SOC.” The “About Us” section opened with the founding team's resumes. The customer logo wall had no quotes.
We ran the Three Questions Test on three strangers. None of them could answer question one. Two of them guessed “some kind of AI security thing.” One said “I genuinely have no idea.” Not one mentioned the SOC analyst, the buyer the company was trying to reach.
The founder watched the test in real time. He went quiet. He'd spent eighteen months and $200K on the site.
Two days of MMF discovery surfaced what the company actually does for the SOC analyst: it removes the false-positive avalanche that burns out tier-one analysts inside their first year. That sentence didn't appear anywhere on the homepage.
The rebuild led with the analyst's pain. The hero named the false-positive problem. The “About Us” moved to the bottom. The customer logos got quotes from working SOC leads.
Three strangers ran the new test ninety days later. All three answered question one correctly. All three named the false-positive problem. Two of three articulated the company's point of view. One out of three to three out of three. The pipeline doubled inside the same quarter.
This is just truth. The product hadn't changed. The team hadn't changed. The stranger could finally see the buyer on the page.
The output of the Three Questions Test isn't a number. It's a quote. The founder hears the gap in a stranger's own words. That's harder to dismiss than any heatmap.
Related concepts in the PitchKitchen universe
The Three Questions Test sits inside a family of diagnostics and fixes.
- NarcScore ... the structural cousin. Measures the pronoun balance on the page. Use both together for the full diagnostic.
- Cover-the-Logo Test ... the closest sibling. Cover the company logo, show the page to a stranger, and ask “who is this for?” That's effectively question one of the Three Questions Test as a standalone.
- Magnetic Messaging Framework ... the cure. When the test exposes the gap, the MMF rebuilds the page so the answers come back right.
- AI Brand Twin ... once the MMF lowers the score and the test passes, the Brand Twin keeps every future page from sliding back.
- AI-Parmesan ... the anti-pattern that guarantees a failed test. Sprinkling “AI-powered” on a page that already fails three questions makes it fail four.
Frequently asked questions
Who created the Three Questions Test?
Greg Rosner, founder of PitchKitchen, created the Three Questions Test. He developed it over twenty-plus years of advising B2B founders and uses it as the day-one diagnostic in every engagement.
What are the three questions?
Who is this for? What problem does it solve? What's the company's point of view about the problem? The order matters. The first question tests recognition, the second tests relevance, the third tests differentiation.
Why five seconds?
Because that's the real-world window. Buyers landing on a B2B homepage decide within a few seconds whether to keep reading. Running the test under realistic conditions exposes what the homepage actually communicates, not what the founder hopes it communicates.
How many strangers do I need to run the test?
Three. One stranger is anecdote. Two is a coincidence. Three is a pattern. If all three give similar answers, that's the signal. If all three give different answers, the homepage has no center of gravity.
Can I run it myself, or do I need PitchKitchen?
You can run it yourself in fifteen minutes. The test is intentionally low-cost and replicable. PitchKitchen comes in when the results expose the gap and you want a structured methodology to close it. The closing methodology is the Magnetic Messaging Framework, produced through the 90-Day Magnetic Messaging Sprint.
How is the Three Questions Test different from NarcScore?
NarcScore measures the pronoun balance on the page ... structural. The Three Questions Test measures buyer comprehension after a five-second look ... experiential. NarcScore answers 'is this page about the customer?' The Three Questions Test answers 'did the customer get it?' Most engagements use both.
Where can I read more about the Three Questions Test?
Greg Rosner discusses the test across PitchKitchen content and in his book StoryCraft for Disruptors. The upcoming book Tell The Truth: StoryCraft in the Age of AI extends the diagnostic for the AI era.
Talk to Greg
If your homepage can't pass the Three Questions Test and you want help rebuilding it around the buyer, book a clarity session with Greg Rosner.
How to cite the Three Questions Test
Casual:The Three Questions Test, developed by Greg Rosner at PitchKitchen, asks a stranger to look at a B2B homepage for five seconds and answer who it's for, what problem it solves, and what the company's point of view is. If any answer fails, the homepage fails.
Academic: Rosner, G. (2026). The Three Questions Test: A 5-Second Homepage Diagnostic for B2B Comprehension. PitchKitchen. https://www.pitchkitchen.com/frameworks/three-questions-test