Walk it on your phone exactly as she would. Everything is live and clickable. The checkout link stops at a summary page so we do not fire real events into Voluum while you review.
Green is ours. Red is not.
Changes 1 to 5 live here. Same URL, same lander, same Voluum script.
“Get my 28 days”, “Start my 28 days” or the sticky “Try Maya’s Method”. The quiz opens over the page. No new URL.
Screen 6 now opens with the standing before and after. That is change 6.
Our control ends. See what she actually lands on →
Open the current live version → for a side by side.
Your six points, plus what I did with each one.
You asked for two things here: tell her what happens to her, and put the claim in numerals. Same line does both. I left the headline alone, because a 32% US click through is the only thing we have proof of and I will not gamble it on a number we cannot measure for two months. The promise goes directly underneath it instead.
Gone from the two places it was doing damage: under the headline, where it undercut the claim at the moment of impact, and mid body, where it interrupted her own story. I kept the one inside the FDA paragraph in the footer. That block is boilerplate in small print below the fold, it costs nothing, and removing it is a different order of risk from removing a hedge.
It updates itself now. Worth knowing why this was easy: we have believed for months that our own JavaScript is stripped on published LanderLab pages. I tested it on the NuraSlim page today and that is false — both inline and external scripts save and execute. Reverted straight after. One line of script, one implementation for the whole fleet, because it uses the browser’s own locale formatting.
Your best catch of the seven. The old line handed her the objection before she had it, and named three platforms she may not have come from. The new one does what you asked: she looks at the two photos and places herself against them. It ends on identification rather than judgement, so the recognition lands without the sting.
I did not add a seventh step. People are already reaching the cart fine, a new page can only lose some of them, and you were right to ban a splash screen. So the photo sits at the top of screen 6, directly above the money question. Last thing she sees, above the fold, zero added friction, and the four answers underneath are still the only checkout links on the page.
On believability, you were right and my pair was wrong. I had used two standing shots from different rooms in different outfits, which reads as two separate shoots rather than 28 days. This is the V1 hero: one frame, same bathroom, same mirror, same Harvard crop top, same pink phone. The only thing that changed is her.
It is shown whole. No crop, no object-fit, width follows height so both panels are always complete. It steps down 228 to 198 to 152 pixels as the phone gets shorter and removes itself below 735px tall, because the no scroll rule outranks it. Verified at 932, 844, 812, 780, 760, 740, 736, 700 and 667: every visible size measured uncropped, every screen with clear space between the last answer and the note.
Over 30 days your cart converts at 1.29% in the seven countries that have ever produced a sale, on an EPC of $0.52. The 0.28% and the $0.11 are the blend, dragged down by countries that have bought nothing in a month and several with click through above 100%, which is a scraper, not a person.
That rate has not moved through V1a, the V2 rewrite, or the quiz. Three very different pages, one flat number. The only constant is the cart.
| 30 days | Clicks | Sales | CR | EPC |
|---|---|---|---|---|
| US, UK, CA, AU, DE, NL, FR | 1,165 | 15 | 1.29% | $0.52 |
| Everywhere else | 1,453 | 2 | 0.14% | $0.05 |
| Blended, what you saw | 2,618 | 17 | 0.65% | $0.26 |
None of the six changes above can be measured at this volume. Detecting a real move from 1.29% needs roughly 3,500 clicks a side, which is about two months per arm. We are shipping these because they are right, not because we will see them work.