Story-Driven HomepageThree Questions Test

Why does our message work in person but not on our website?

Greg Rosner

By Greg Rosner

Founder of PitchKitchen · Author of StoryCraft for Disruptors

· 7 min read

Hero image for Why does our message work in person but not on our website?

TL;DR

Your message works in person because you're finishing it live. In the room you read the badge and decide who this person is, you skip the part they already know, you reach for the one number that matches their job, and you answer the doubt on their face before they say it out loud. Your website makes all four of those calls in advance, for a stranger who will never ask you a question. Across 302 B2B homepages scored on the 19-signal Brand Signal Score, only 27 name the buyer, pass a 7-second read, and carry one liftable sentence. 131 score exactly half marks on all three, and 130 of those sit in the bottom two bands. Half marks is what a message that needs a person to finish it looks like once it's written down.

Your message works in person because you're finishing it live. In the room you read the badge and decide who this person is, you skip the part they already know, you reach for the one number that matches their job, and you answer the doubt on their face before they say it out loud. Your website has to make all four of those calls in advance, for a stranger who will never ask you a single question.

That's the mechanism. What's sitting on the page is your pitch with you subtracted, and most founders are surprised by how much of it they were carrying. The rest of this piece is the four jobs you do for free in the room, a way to test whether your page does them, and what actually changes when it doesn't.

What does the room supply that our website can't?

Four things, every time, and you do all four without noticing.

You target. You see a title on a badge or a role on a calendar invite and you pick which version of the company to describe. The economic buyer gets one opening, the person who'd use the product daily gets another.

You sequence. You start where they already are. If they name the problem in their first sentence, you skip the setup entirely and go straight to what you do about it.

You select proof. You've got a dozen numbers and you reach for the one that lands on this person's job. A CFO hears the one about cost per period. An operator hears the one about minutes per shift.

You handle the objection before it hardens. You watch the face change, and you answer the doubt while it's still a flicker rather than a written-down reason to pass.

None of that is documented anywhere. It lives in your reflexes and re-runs perfectly every time you're standing there. Take you out of the room and the page is left holding a general-purpose description of a product, which is the thing buyers scroll past. This is also why your happiest customers can explain what you do better than your homepage can: they learned the message from a version of you that was adapting live.

Isn't this just a copywriting problem?

It's a decision problem wearing copywriting's clothes, and the data says the failure is almost never a blank page.

This August we scored 302 B2B healthcare software homepages against the 19-signal Brand Signal Score, the same rubric behind our free diagnostic, and published the results as the 2026 Healthtech Messaging Index. Three of those signals measure exactly what we're describing here: the 7-Second Test (can a stranger tell what this is), ICP Clarity (does the page name who it's for), and LLM Quotability (is there a sentence a machine can lift and repeat back).

Outright failure is rare. Only 6 of 302 companies score a flat zero on the 7-Second Test, 6 on ICP Clarity, 29 on LLM Quotability. What happens instead is partial credit, on all three at once.

GroupCompaniesAverage total (of 38)
Full marks on all three travel signals27 of 30227.07
Half marks on all three travel signals131 of 30216.40
Whole field30219.18

130 of those 131 half-credit companies land in the bottom two bands. Only 27 companies out of 302 name the buyer, survive a seven-second read, and carry one sentence worth quoting. Reported as association, not cause: a page can be clear and still lose for a dozen other reasons. But half marks on all three is a recognizable shape, and it's the shape of a message that needs a person to finish it.

Read the scoring notes and you can hear the missing half. The page implies hospitals but never names the buyer's role, so nobody can confirm they're the intended reader. The hero sentence is clean and declarative with no number standing next to it, so there's nothing an engine or a champion can lift. In the room you'd close both gaps in your next breath, and you'd never know they were gaps at all.

Why do our best channels hide this?

Because you're in every one of them. A booth, a demo, a discovery call and a partner intro all return the same verdict, and all four verdicts are contaminated the same way.

A conference is the clearest example. You bought a room full of the right buyers, you had twenty good conversations, and not one of them tested your message, because you stood in front of it for three days. That's also why sponsorship rarely turns into pipeline: presence isn't preference, and the booth wall is usually the homepage with a carpet under it. The same contamination is why demos go great and deals still stall.

The channels that tell you the truth are the ones you're absent from. The page a stranger opens at eleven at night. The deck your champion forwards into a meeting you're not invited to. The answer an AI engine gives when a founder asks who fixes this kind of problem. Those are the moments that decide whether buyers remember what makes you different, and you have no ability to intervene in any of them.

How do we test whether our message travels?

Four tests, in the order they get uncomfortable. Run them this week and you'll know within an hour.

  1. 1The Cover-the-Logo Test. Hide your logo, show the homepage to someone outside the company, ask who this is for and why this vendor. If the page reads identically with a competitor's logo on it, there's no position on it.
  2. 2The Three Questions Test above the fold. Why change, why change now, why change with you. Five seconds, no scrolling.
  3. 3The 60-Second Test on a rep who joined in the last month. A new rep is the closest human stand-in for a cold reader, and they can only repeat what's been written down.
  4. 4The forward test. Send the page, and nothing else, to someone who matches your buyer and has never heard you talk. Ask them to say back who it's for and what problem it solves. Answer none of their questions. Your silence is the entire test.

Founders usually fail the fourth one and want to explain why the reader misunderstood. That explanation is the missing copy. Write it down and you're most of the way there.

What actually changes on the page?

Five edits, and they're the same five whether you sell clinical software or cybersecurity.

  1. 1Name the buyer by role, not by segment. "Health systems" is a market. "The VP of perioperative services who signs off on documentation tooling" is a person who can recognize themselves.
  2. 2Lead with the problem and take a side on the status quo. A page that opens with what your product is has already chosen Solution-Centric Marketing, and buyers tune it out.
  3. 3Write one liftable sentence. One clean, standalone claim a champion can paste into Slack and an AI engine can quote without editing. Most homepages have zero.
  4. 4Put the number next to the claim. Not on the case-studies page. Next to the sentence it proves, where a skimmer will hit it.
  5. 5Pick one path. When the page offers a demo, a whitepaper, a newsletter and a pricing call with equal weight, you've handed the reader a decision you should have made for them.

Do those five and the page starts doing what you do in the room. It won't do it as well. It'll do it at three in the morning, to a hundred people at once, without you.

What if the message only lives in our heads?

That's the common case, and it's the reason the writing stalls. You can't document a position you've never actually taken. What most founders have is a set of live moves that work, assembled over three hundred conversations and stored as instinct.

Getting it out is truth extraction: sitting with the founder and the people who sell, pulling out what you actually believe about the buyer's situation, what you refuse to do, who you're wrong for, and what changes for a customer that wouldn't have changed otherwise. That's your narrative identity, and it's the layer underneath the words. Without it, a rewrite produces a nicer version of the same fog, which is how a website ends up underselling what you built and why so many B2B sites sound identical.

Worth noting the failure runs the other direction too. A story tuned for one audience can land beautifully and still be the wrong story, which is why a pitch that wins investors doesn't move customers. Working in the room proves the message works on that room. Nothing more.

What this means for you

Take the version of the pitch you'd give if you were standing at a booth and someone asked what you do. Write it down verbatim, including the part where you'd glance at their badge first. Now put it beside your homepage.

Everything in the spoken version that isn't on the page is the Room Subsidy, and it's the exact list of what your website owes a buyer who'll never meet you. Most of your pipeline is now made of those buyers. AI brought the cost of content to zero, so volume stopped being the moat. Perspective is the moat, and perspective only counts once it's written down where a stranger, a champion and a machine can all find it.

Questions People Ask

FAQ

Why does our pitch work in person but not on our website?

Because in a live conversation you supply four things the page can't: you target (you read who they are and choose an entry point), you sequence (you skip what they already know), you select proof (you reach for the number that matches their job), and you handle objections before they harden. None of that is written down. It runs from your reflexes. The page has to make all four decisions in advance for a stranger who never asks a follow-up, so what's left on it is your pitch with you subtracted.

How do we tell whether the problem is our message or our website design?

Give the page to someone who matches your buyer and has never heard you talk about the company. Ask them two questions: who is this for, and what problem does it solve. Do not answer anything they ask you. If they can't tell you inside a minute, no redesign fixes it, because the words a designer would lay out don't exist yet. If they can tell you and the page still doesn't convert, then you have a design and structure problem worth spending on.

What's the fastest test that our message travels without us?

The Cover-the-Logo Test takes thirty seconds. Cover your logo, show your homepage to someone outside the company, and ask them to name who it's for and why this vendor over the others. A page that reads the same with any competitor's logo on it has no position on it. Follow it with the 60-Second Test on a rep who joined in the last month, since a new rep is the closest human stand-in for a cold reader.

Does this mean conferences and demos are a waste?

No. They're the highest-signal rooms you get. The trap is treating them as evidence that your message works, because you're present in every one of them, so every one of them returns a false positive. Use live channels to learn what buyers actually say and which objection keeps coming up, then move that language onto the page and test it in a channel where you're absent.

How long does it take to fix a message that only works in person?

Writing takes days. Deciding what to write is the part that takes weeks, because the position has to come out of the founder's head and survive contact with the sales team, the buyers, and the competitive set before it's safe to put on a homepage. Our engagements run 90 days for that reason: most of the calendar goes to extraction and pressure-testing, not to copy.

Want this kind of thinking shipping for you?

If your pitch converts when you're standing there and stalls when you're not, the gap isn't your writing. It's a position that lives in your reflexes instead of on the page. That's the work of the 90-Day Magnetic Messaging Sprint: truth extraction to get the position out of your head, then documentation so your website, your reps, your champion's forwarded deck and the AI engines all carry the same message when you're nowhere near the room.

That's the 90-Day Magnetic Messaging Sprint. One quarter, one fixed price: we extract your story, build the Magnetic Messaging Framework and your AI Brand Twin, then ship the website and sales enablement that run on it. $25K–$45K fixed for the quarter, and you own all of it at the end.

About the Author

Greg Rosner

Greg Rosner

Founder, PitchKitchen · Author of StoryCraft for Disruptors · Creator of the Magnetic Messaging Framework™

Greg is a B2B messaging therapist for growth-stage CEOs ($5M-$75M). He helps founders extract the truth they've been hiding from themselves, name the villain in their industry, and build the messaging infrastructure that scales their voice through AI. PitchKitchen has worked with 100+ B2B companies across SaaS, healthtech, fintech, cybersecurity, and AI-driven solutions.