How do we know if our messaging actually works, or just sounds good in the room?

Last week I got on a call with the CEO of a $22M cybersecurity company. Two months back, his exec team flew to an offsite and rewrote the company's whole message. New homepage headline. New deck. A one-liner everybody could finally agree on. He told me the room actually stood up and clapped. Real applause.

Then he said the quiet part. 'Greg, everybody loved it. And the pipeline hasn't moved an inch.'

I asked him one question. Who was in that room when everyone clapped? Twelve people. His co-founder, his VP Sales, his product lead, a couple of board folks. And there it was. He'd tested the company's new message on the only twelve people on Earth who already know exactly what the company does. Of course they loved it. They wrote half of it. They've lived it for six years. They can't read that headline cold even if they tried.

That's the trap. The message passed the room and never once met a buyer. And a message that's only ever been graded by insiders isn't a working message. It's an agreeable one. Those are not the same thing, and the gap between them is where pipeline goes to die quietly.

What's actually broken here?

Here's the thing almost nobody tells you. Internal applause is the most dangerous signal in marketing.

The people in your conference room can't un-know what you do. Every time your message leaves something out, they fill the gap automatically with years of context your buyer will never have. A headline that reads as crisp and obvious to them can read as foggy and generic to a stranger. They're not grading the message. They're grading their own memory of the company.

There's a second problem hiding inside the applause. Rooms reward safety. A message gets unanimous agreement precisely when it offends no one and names no one. Everybody nods because there's nothing sharp enough to argue with. But a buyer who's drowning in fifteen competitors that all sound the same doesn't lean in at safe. Safe is exactly what makes you invisible. The message that wins the room is usually the one that's been sanded down until there's nothing left to grab.

The only verdict that counts comes from someone who's never heard of you, has the problem you solve, and decides in about five seconds whether to lean in or close the tab. That's the Three Questions Test and the Cover-the-Logo Test, and it's the opposite of an offsite. Nobody in that test already loves you.

Why is this worse now than it's ever been?

Because AI made sounding good free. A message that reads smooth used to take a copywriter a week. Now you can generate fifty smooth options before your coffee's cold. Smooth is the floor now. It's the thing the machine does on autopilot, for everyone, at once.

If sounding good is free, sounding good is worth nothing. The only thing that's scarce now is a message true and specific enough that one particular buyer reads it and feels like it was written for them and only them. That feeling can't be averaged into existence. It comes from a real, lived, slightly uncomfortable truth about who you're for and what you're against. This is just truth.

And there's a machine reading over your shoulder now. ChatGPT, Claude, Perplexity and the rest read every surface you publish and average them into one picture of who you are. A message engineered to please a room is almost always the de-risked, lowest-common-denominator version. That's the exact version the machine can't tell apart from your competitor, so it recommends neither of you with any confidence. Garbage in, garbage out, at machine scale.

What's the diagnostic I can run on my message this week?

No agency. No budget. No second offsite. Three tests, and you can finish all of them before Friday.

1. 1The Stranger Test. Take your homepage hero, cover the logo, and send it to five people who fit your buyer but have never heard of you. Ask one question: who is this for, and what problem does it end? If they hedge, stall, or guess wrong, your message failed in the only room that matters.
2. 2The 'Not You' Test. Read your message out loud and ask: could our nearest competitor slap their logo on this, unchanged, and have it be just as true? If yes, you didn't write a message. You wrote a category brochure. A real message names a villain, an old way, or a buyer your competitor won't claim.
3. 3The No-Decision Test. Pull your last ten lost deals. How many went to a competitor, and how many just stalled, went dark, or 'decided to wait'? If most of them stalled, that's not a sales execution problem. That's a message that never made the cost of staying put feel real.

What do I see when I run this across 200+ B2B companies?

The same loop, over and over. A founder fixes the message in the room, ships it, watches the silence that follows, and reads that silence as a traffic problem or a sales problem. They buy more traffic and push the sales team harder, pouring fuel on a message that was never tested where it counts.

The numbers back up where the leak actually is. RCKT Marketing found that 57% of SaaS marketers say their poor-fit leads come straight from unclear or overly broad messaging. Not bad targeting. The words themselves. And April Dunford's research puts 40 to 60 percent of B2B deals ending in no decision at all, which means roughly half your lost pipeline never went to a competitor. It went to confusion.

“As AI makes it trivial to build and launch products, the biggest challenge for product teams is quickly becoming distribution: getting people to pay attention to your product in the increasing cacophony of launches.”

... April Dunford, positioning expert and author of Obviously Awesome

The room can't fix that, because the room isn't the cacophony. The buyer is. And the buyer is the one person who was never invited to the offsite.

What does this look like in a real company?

Here's one, details changed to protect the company. A roughly $30M finance-infrastructure company came to me right after their own messaging refresh. Their new offsite-approved line was 'the modern platform for intelligent finance operations.' The room had loved it. I covered the logo and ran the Stranger Test with five of their actual buyer profiles. Not one could tell me who it was for or what it fixed. Two of them guessed it was an accounting tool. It is not an accounting tool.

We threw the polished line out and rebuilt the message around the one true, specific thing they'd been burying because it felt too narrow to say out loud. It wasn't narrow. It was the whole reason their best customers picked them. The new message named the old way they were replacing and the exact operator they were for. Over the next quarter, the change in the sales motion was the tell: demo conversations got shorter and warmer, and the reps mostly stopped having to re-explain what the company even did. The product never changed. The message finally met a buyer.

What does this mean for you?

If your message has only ever been graded by people who already know what you do, you don't actually know whether it works. You know it's agreeable. Stop confusing the two. The applause at the offsite is the least reliable data you'll ever collect about your own message.

1. 1Get your message in front of five strangers who fit your buyer this week. Cover the logo. Ask who it's for and what it ends. Believe their confusion over your team's applause.
2. 2Find the one true, specific thing you've been smoothing out because it felt too narrow or too pointed. That sharp, slightly risky truth is almost always the actual message.
3. 3Name the old way you're against. A message without a villain is a brochure, and the machine and the buyer both skip brochures.

This is the work that sits underneath everything I do. It's why I built the Magnetic Messaging Framework (MMF), the strategic narrative system that gives a company one clear, true, repeatable message built around category design, villain framing, an old-way / new-way contrast, and a promised-land outcome. Get that right and a stranger feels seen, your sales team stops re-explaining, and the machine finally has something specific enough to recommend. PitchKitchen builds Magnetic Messaging Frameworks for founder-led B2B companies in the $5M-$75M range, because the message is the part that actually scales. If you want to see how I diagnose a homepage on the spot, read 'How Brand Signal Score Reveals What's Actually Killing Your B2B Homepage,' and if your sales team keeps calling marketing's leads garbage, 'Our Marketing Team Gave Us 59 Leads Last Month. Our Sales Team Says They're All Garbage. Who's Right?' is the same disease from a different seat.