Why does our messaging keep sliding back to features after we fix it?

The scene I'm in this week

This week I got on a call with a founder I'd worked with eight months ago. A $19M Series B B2B SaaS company, data infrastructure for engineering teams. Back then we did the real work. We cut through the platform-speak, named who they were actually for, and put the outcome at the top of the homepage. It landed. Demo quality jumped, reps stopped burning the first ten minutes explaining what the company was.

He pulled the homepage up on the call, and I watched his face change. Because it had slid back. The clean line we'd fought for was pushed halfway down the page. Above it now sat a stack of capability blurbs. Real-time sync. A new observability layer. Three integrations they'd shipped since. Each one added by a different person at a different time, each one reasonable on its own, and together they'd quietly turned the page back into a feature list.

"Nobody touched the main copy," he said. And he was right. Nobody sat down one day and decided to undo the work. It crept. A capability here, a launch there, a homepage refresh that folded in the newest thing. He thought the message had gone stale, that it needed another pass. That's a real instinct, and it's the wrong diagnosis.

The message didn't go stale. It fell. There's a gravity to a B2B message, and if you don't build something to hold it up, it drops right back to describing what you built. This is just truth, and it's one of the most common things I see with founders who did the messaging work and can't understand why it didn't hold.

Naming what's actually broken

Here's what's actually happening. Call it Feature Gravity. Your message has a resting state, and that resting state is a feature list. Describing what you built is the path of least resistance, because it's the thing your team knows best, the thing that's easiest to write, and the thing every new person defaults to when they add a section. The clear, buyer-first message you fought for isn't the natural state. It's the exception you were holding in place. The moment you stop holding it, it falls.

This is Solution-Centric Marketing reasserting itself. Not because anyone chose it, but because it's the ground state. A feature is concrete and easy to describe. The outcome a buyer actually cares about is abstract and takes discipline to keep in front. Every new capability ships with a natural gravitational pull toward its own line on the homepage. Ship enough of them without a rule, and the outcome story ends up buried under the spec sheet you spent real money to escape.

And it isn't only features that add mass. Every new hire who wasn't in the room when you found the message writes from the default. Every AI draft with nothing specific to work from reverts to the category average. Every well-meaning refresh that bolts on the latest launch pulls a little harder. None of it looks like sabotage. It looks like progress. That's exactly why founders miss it until the page reads like everyone else's again.

Why this is worse now than ever

Adding used to be expensive, and that expense quietly held Feature Gravity in check. A homepage rewrite meant a designer and a week. New copy meant someone sitting down to write it. The friction meant you couldn't pile on features faster than you could think about them. That brake is gone.

AI dropped the cost of adding to zero. You can generate a capability section in twenty seconds, a fresh homepage draft over lunch, a month of content by Friday. The mass that pulls your message back toward a feature list now accumulates faster than any founder can watch. Worse, the machine has its own gravity toward the average. Rival.tips ran a similarity index across 178 AI models and found only 25% of them showed any unique stylistic traits. The default of a language model, like the default of an un-anchored message, is to sound like everyone else. Feed your AI tools a message with no documented core and they'll accelerate the fall, not slow it.

There's a second turn here that matters more every quarter. The machine is now the thing briefing your buyer before a human ever reads your page. When your message keeps drifting back to an interchangeable feature list, the AI reads an interchangeable company and repeats it. Clarity that won't hold still can't be cited. That's the mechanism behind strategic positioning being the only moat AI can't copy: a position holds because it's a decision you locked, and a feature list drifts because it's just a description of the moment.

The diagnostic: run this on your last messaging project

You don't need to hire anyone to see whether Feature Gravity has pulled your message back down. Run these three tests this week.

1. 1The Six-Month Diff. Find the version of your homepage from right after your last messaging project, in your CMS history or the Wayback Machine. Put it next to today's page. Count the sentences about the buyer's problem and outcome versus the sentences about your features. Has the ratio slid back toward features? If the outcome line that used to open the page is now halfway down under a stack of capability blurbs, gravity won and you didn't notice.
2. 2The Last-Feature Test. Look at the most recent thing you shipped and find where it lives on the homepage. Did it get folded into the existing story as evidence of the outcome you already promised? Or did it get its own new line, its own little section, competing for attention with everything else? A new bullet every time you ship is the single clearest fingerprint of Feature Gravity at work.
3. 3The Cold-Draft Test. Ask your newest hire to describe what the company does in three sentences, with no help. Then ask your AI tool to do the same, cold. If either one hands you a feature list, that's the resting state of your message with nobody holding it up, because that's all the signal they had to work from. The gap between what you wish they'd said and what they actually said is the size of your gravity problem.

What I see across 100+ B2B companies

Across more than a hundred growth-stage B2B companies, the pattern is remarkably consistent. Messaging doesn't decay slowly, like paint fading. It snaps back, fast, the moment the founder stops personally guarding it. I've watched a message hold beautifully for one quarter and be half gone by the third, not because anyone disagreed with it, but because the person keeping it alive got pulled into a raise, a hire, a product push, and stopped being the human gravity holding the story up.

That's the real finding, and it stings a little. If your clear message only exists in your head and on one page you personally defend, it has a half-life. The Indie Hackers community put the underlying truth well: "Positioning is the first irreversible decision a startup makes, and everything that follows quietly obeys it." The problem is that most founders never make it a decision. They make it a nice paragraph. A paragraph gets overwritten. A decision, written down and shared, is what the next feature has to obey.

This is a different problem from outgrowing your message, though the two get confused. When your messaging still describes your old company, the story is real but frozen at an earlier, smaller stage. Feature Gravity is the opposite motion: the story was current and correct, and it fell back to a feature list because nothing held it in place. One is a message that never moved. The other is a message that couldn't stay put.

A real example

Here's a composite from the pattern, with details changed. A $23M Series B cybersecurity company came to me a second time, frustrated. We'd rebuilt their message a year earlier, and it had worked. Clear category, a named old way they were fighting, an outcome the buyer could feel. For a quarter, the homepage did its job and win rates ticked up.

Then they shipped. Three features over two quarters, each a real advance, each added to the homepage by whoever owned that launch. A new detection module got its own section. A compliance report feature got a block. An integration got a line. None of it was wrong. All of it was heavy. By the time they called me, the outcome story that used to open the page was three scrolls down, and the top of the homepage read like a product-comparison chart. Their own head of sales had gone back to explaining the company from scratch on calls, the exact problem we'd solved a year before.

We didn't rewrite the copy again. That would've just reset the clock on the same drift. We documented the core as a framework, then trained their team and their AI writing tools on it. The category, the villain, the old-way-to-new-way shift, the one outcome only they delivered. After that, every new feature had a rule to obey: it had to prove the outcome, not claim its own spot. The next launch folded in as evidence instead of another bullet. Two quarters later the homepage was still opening on the buyer, not the spec sheet, and nobody had to stand guard over it. The message held because the system held it, not because one person did.

What this means for you

If you fixed your messaging and it slid back, stop blaming the copy and stop reaching for another rewrite. The rewrite isn't the fix, it's the thing you'll be doing again next year on the same schedule. Feature Gravity is a system problem, and it needs a system answer: a place where the decision lives, that every feature, every hire, and every AI draft has to obey. Without that, clarity depends on one person's attention, and attention is the one resource a growing founder runs out of first.

That place is the Magnetic Messaging Framework (MMF), the documented brand bible that holds your positioning still while the company keeps moving. It writes down the decisions that drift when they're only in your head: the category you own, the villain you fight, the old-way-to-new-way shift, and the one outcome only you deliver. Then it becomes the context your whole team and your AI tools write from, so a new feature folds into the story instead of replacing it. It's the difference between a message you personally guard and a message the system produces on its own. Greg Rosner, founder of PitchKitchen and author of Story Craft for Disruptors, builds the MMF for founder-led B2B companies in the $5M-$75M range, so the clarity you found once stops being something you have to keep re-finding. Why it matters: until the message is written down as a living framework, every quarter of growth quietly costs you the clarity you already paid for, and you rediscover the same truth over and over instead of compounding it.

1. 1Run the Six-Month Diff this week. Put your post-project homepage next to today's and measure how far the outcome slid under the features. Name the drift before you fix anything.
2. 2Write the message down as a decision, not a paragraph. The category you own, the villain you fight, the one outcome only you deliver, in the buyer's actual words. A decision is what the next feature has to obey. A paragraph is what it overwrites.
3. 3Give your team and your AI tools that written core to work from. If a new hire or a cold AI draft still hands you a feature list, the message isn't documented anywhere they can reach, and gravity will keep winning.

You already found the clear message once. That was the hard part, and you did it. Now give it something to hold onto, because clarity you don't write down is clarity you rent, and Feature Gravity comes to collect every quarter.