What the founder believed walking in
He had a real product and a beautiful site. A dental-developed kit for people who grind and clench at night. The founder is a brand person, so his instinct was the one most founders have. Better marketing. A sexier page, stronger creative, more traffic, and a faster answer on which audience to chase.
The instinct was reasonable. The traffic was there. About 3,000 visits a month were coming in, and only about 300 of them turned into sales. When the numbers look like that, “fix the marketing” is the honest first guess, and he had already tried it. An outside ad team had bolted on hard checkout tactics he hated, and he could not tell if any of it had helped.
So he came to us the way a smart owner should. He named the felt problem plainly: we get the traffic, and we lose them at the finish line.
What we saw
We put his own numbers on the screen. The leak was not upstream, where he wanted to spend. It was downstream, in the checkout and the follow-up, in the exact places he could not see. His store ran on a rented platform built for the seller, not the buyer, and it had removed the parts that would have shown him where people abandoned. So he was optimizing in the dark.
Then we looked at the follow-up sequence. About 160 leads would come in, and only one or two would buy. That is not a traffic problem. With a product this good and a strong money-back promise already in place, a number that low points at the sequence, not the market.
And there was a third thing, the one that made this business different. It is a health product, so the most persuasive, most truthful thing he could say was exactly what the ad platform transcribes, reads, and flags. His truest claim was his riskiest claim. That is not a creative problem you can write your way out of. It is a structure problem.
What we decided, and what we chose not to do
We did not try to make the claim louder or sneak it past anyone. That road ends in a shut-down account, and it is not who we are. The honest concession up front: on the public part of the funnel, the part the platform reads, we would say less, not more.
We also did not touch the deeper platform rebuild yet or chase a new audience on a hunch. Those were real, but they were not the fastest way to stop the bleeding. The bleeding was in the checkout, the guarantee wording, and the follow-up. So we fixed what we could see and measure first, and left the bigger bets for after the economics were proven.
What we built
We built a bridge, not a firewall. The public ad and landing page needed to stay simple and policy-clean. The fuller, clinician-voiced education needed a different mechanism than an ad or a landing page ever allows, and the email sequence became that mechanism: a channel the buyer opts into, with the room to say things properly and the same standard of honesty everywhere. As the founder learned to say it: the deeper education lives in the inbox because that is the channel built to carry it, not because it says anything the ad could not stand behind.
Here is what that meant, step by step.
The ad and the landing page stay policy-clean. We changed the grammar, not the truth. Instead of a second-person call-out, “Are you suffering from jaw pain?”, we wrote it in the third person: “People who suffer from jaw pain.” Same recognition, phrased as a description rather than a call-out aimed at one reader. The proof stack does the persuading a first-party claim cannot: a real testimonial, the device’s cleared status, and the guarantee.
The sign-up page redirects the moment someone submits, so there is no detour to an inbox. The real education, the clinician’s voice, lives behind that email gate, where it belongs.
Then we re-engineered the page and the guarantee. The old page sold a “healing kit” in clinical language. But people who grind their teeth are not shopping for a healing kit. They are shopping for a night guard. So the hero became three lines in the buyer’s own words: the promise, the guarantee, and what you get.
The guarantee itself was the biggest single lever, and the wording was deliberate. “Wear it every night for 30 days, and if you are unhappy, we will buy it back.” Not “if you do not notice an improvement,” because whether you improved is subjective and it edges toward a health claim. Whether you are happy is not, and asking people to wear it every night for 30 days quietly builds the habit that gives the product its best chance to do its job.
We reordered the page to show the parts first, took the escape hatches out of the checkout, and rewrote the 160-lead follow-up sequence with fresh testimonial video.
What changed
The results here are real, and they are specific to this business. We are not promising them to anyone else.
The refund rate fell from about 6% to under 1.5%. About 150 units sold in six to seven weeks. Part of that is the honest guarantee wording setting the right expectation before the sale, and part is the follow-up finally doing its job.
We are not going to dress this up. These are early, and the business had further to go. But they point the right way: when the page tells the truth in the buyer’s language and the follow-up stops leaking, the refunds and the sales both move in the direction you want, without a single claim the platform would flag.
The lesson
When you sell in a locked-down category, the reflex is to believe the rules are the enemy of the sale. They are not. The thing to fix is rarely the claim. It is where the claim lives and how it is worded.
Say the plain version where the format demands brevity, and give the fuller version the channel built to carry it. Let a testimonial, a cleared status, and an honest guarantee carry the weight your own words are not allowed to. Word the guarantee around whether someone is happy, not whether they improved, and it does double duty: it stays defensible, and it builds the habit that gives the product its best chance to do its job.
The rules did not stop this business from selling. They just told us where to put the truth.