Then it starts. Retinal is the fastest-converting form of vitamin A your skin can use — but nothing in this category works on the timeline the ads imply. So we will tell you the real one before you pay.
30 ml · amber glass · fragrance-free
Full refund any time inside 90 days, opened or not. You keep the bottle.
Paid-social traffic arrives with an expectation set by the ad, not by the page. In skincare that expectation is almost always faster than reality. The person who buys expecting week-one results is the same person who refunds in week three and leaves the review that kills the ad account's ROAS.
So the headline spends its position on the objection instead of the benefit. The goal is not maximum conversion on this page — it is maximum conversion that survives 90 days.
My reasoning, not a tested result. I have not run this against a control.
There is no menu bar. Someone who clicked an ad did not come to browse a catalogue — every nav link is an exit that costs you the click you already paid for.
The logo is not a link either, for the same reason.
Week one to two is the single largest refund driver in retinoid products, and it is entirely predictable. A customer who was not warned reads flaking as "this broke my skin". A customer who was warned reads the exact same flaking as "it's working".
Same physical event. Opposite outcome. The only variable is whether the page got there first.
This section also does quiet work for the ad account: it filters out the buyer who wants a one-week transformation, before they cost you a refund, a chargeback, and a one-star review.
Compared to a $12 drugstore serum it looks indefensible. So compare it to what it actually replaces — a night, repeated.
A price objection is rarely about the amount. It is about the comparison the customer reached for on their own — and if you leave them alone, they reach for the cheapest thing in the category.
You cannot win that comparison. You can replace it. Dividing by nights moves $68 out of the "serum shelf" frame and into a frame where it reads as small.
The clinic line is doing something specific: it is not a claim of equivalence, it is an anchor placed above the price so the number the customer sees last is $0.76, not $68.
On a page whose only job is same-session conversion, this section is a mistake. On a page attached to a paid media account, it is the highest-leverage block here.
Each of those four buyers converts, then refunds, then reviews. They inflate the CPA you optimise against, feed the algorithm bad signal, and drag the account average down for months.
There is also a trust effect: a page that talks you out of a purchase is read as more credible on everything else it says. Small admissions buy large credibility.
Worth being clear — this is a design judgement I would want to test, not a proven lift. If it were my call on a live account I would run it against a version with this block removed.
"Week two I nearly returned it. The page told me that would happen, so I didn't. Week nine is when I finally saw it."
— fictional, written for this demo
"Does what it says. It is slow, exactly as slow as they said. Four stars because I wish the dropper were easier."
— fictional, written for this demo
"Fine product, wrong person. I have rosacea and it was too much for me. Refund was quick, no argument."
— fictional, written for this demo
Reviews are placed after the timeline, the price, and the disqualifier — not directly under the buy button. A testimonial only does work once the objection it answers has been raised. Placed too early it reads as decoration and gets scrolled.
The three-star review is deliberate. An all-five-star wall is the strongest signal a page can send that the reviews are curated. One honest negative — from someone the page already told not to buy — makes the other two readable.
On a real build these would be real reviews, pulled and labelled with their source. Fabricated reviews presented as genuine are not something I will build, on any account.
Yes, but that is the wrong framing. Retinal converts to retinoic acid in one step instead of two, which is why it acts faster than retinol at the same concentration. It is also why the first two weeks are rougher. There is no version of this where you get the speed without the adjustment period.
It may help with texture and post-acne marks. It is not an acne treatment and we are not going to pretend it is. If active breakouts are your main concern, a targeted product will serve you better than this will, and we would rather say so than take the order.
Start at two nights a week for a month before going higher. If you have diagnosed rosacea or an active barrier issue, this is probably not for you — see the third sample review above, which is the honest version of what that looks like.
Because a 30-day window on a 12-week product is a trick. You would be asking for a decision during the flaking phase, which is the worst possible moment. Ninety days is the shortest window that lets you judge it fairly.
That question has real search and ad-comment volume, and the tempting answer is a soft yes. A soft yes converts better today and produces a refund in six weeks.
Answering it with a clear no does two things: it removes a buyer who was going to be unhappy, and it makes every other answer on the page more believable — because the reader has now watched the page decline a sale.
The 90-day window is the same logic applied to policy instead of copy: the refund term is set by the product's actual timeline, not by the industry default. A 30-day window on a 12-week product guarantees returns.
You now know the flaking is coming, when it stops, and what it will and will not do. That is everything we know.
Start the 12-week run — $68Demo page. This button does nothing.
The button repeats the price. A person scrolling back up to check the number is a person who might not scroll back down.
The phrasing is "start the 12-week run", not "buy now" or "add to cart" — it names the commitment the page spent the whole scroll setting up. The last line should describe what the customer is agreeing to, not what the button technically does.