Extra Strength Astaxanthin
The two chiefs
House chief · 7/10
listicle advertorial in the ingredient's name, product card at two thirds · cold paid social traffic · women of 40-plus, perimenopause, a full routine already · solution-aware entering, most-aware at the card · singlet-oxygen damage met by a membrane-spanning molecule at the trial dose
Lead: Problem-solution lead opened by vindication: A-minus. The deck opens one loop and reason one closes it within about 10% of the page.
Offer: Nearly absent in the copy: a 60-day guarantee stated five ways and no price, bundle or bonus in text. The page sells a try and leaves the deal to the buy form.
Close: None. After the product card the page adds proof (clips, roadmap, thread) and a long FAQ, then stops on a logo lead-in (entry 63).
Throughline: From vindication to trust. It holds to the product card at 64%, then repeats itself.
Strongest
- Reasons ordered as a belief chain, each answering the question the last one raised (entries 6 to 27)
- Evidence doubling as the buying standard, two reasons before the product is named (entries 13, 21)
- Objections promoted to numbered reasons (entries 17, 20)
Weakest
- No price and no close in text (entries 35, 63)
- Social proof that shows its seams (entries 29, 47, 49)
- The FAQ as a second copy of the page, with one answer missing (entries 52 to 56)
Craft chief · REVISE
A confident, well-briefed voice that makes one argument seven ways; by the fourth reason you can hear the sentence shape coming (a short contrast, a dash, a run of fragments).
Strongest
- entry 10: picture, geometry, number, with the number qualified twice
- entry 13: evidence that also sets the buying standard
- entry 30: a review that sells patience
Weakest
- entry 54: a question with a placeholder for an answer
- entry 34: a bullet that contradicts reason one
- entries 47 to 49: a thread with a reused reviewer and timestamps that disagree
001 PASS·noted · the rhythm is deliberate and strong in entries 10 and 24; the short-fragment ending recurs in reasons three, four, five and seven (entries 13, 18, 21, 26) and becomes predictable. · 002 REVISE · attention, belief and desire are built in order; the action job is under-built. Five identical buttons, no price in text, no button after the FAQ and no closing argument (entries 50 to 63). Cost: money, because the reader who reads to the end is given nowhere to go. · 003 PASS·noted · short sentences and plain verbs. The damage type is taught as it is named (entry 7); the Latin species name stands bare in the FAQ (entries 53, 58). · 004 PASS · each reason ends by raising the question the next one answers (entries 7, 10, 13, 18). The best slide on either page. · 005 PASS · the headline's promise is paid by reasons one to three, the deck's open question is closed within about 10% of the page, and there is no deadline device. · 006 PASS·noted · named trials, doses and weeks throughout. Notes for the claims check: trial lengths and study types (entry 13), the origin of the branded ingredient (entry 27), review counts (entries 1, 24). · 007 PASS·noted · emphasis cannot be recovered from the rows beyond the labels on the product card (entry 34). · 008 PASS·noted · tight at sentence level, fat at page level: the FAQ reprints reasons four and five and the roadmap (entries 52, 53, 56). · 009 N/A · a problem-solution lead; no story was selected. · 010 PASS · the list is contracted, second person and sayable; the FAQ answers drop the contractions and read more written than the list. · 011 PASS·noted · the reason-one picture divides the work well with the copy (entry 5). Every reason image carries its heading as alt text, a duplicate and not a description. No price exists in text at all. · 012 REVISE · a live FAQ question answered by the editor's placeholder (entry 54); a logo lead-in printed twice (entries 45, 63); verification tags applied unevenly (entries 29, 30). Cost: trust. · 013 PASS · the page's strongest point: entries 6, 7, 27 and 51 speak to a woman who has tried everything and been burned, in her terms. · 014 PASS · vindication, understanding, trust: a compact movement from objection through proof to confidence, which holds to the product card. · 015 REVISE · dashes in nearly every paragraph; the not-this-but-that pivot three times in entry 7 and again in entries 18 and 53; stacked short endings in entries 13 and 21. Cost: by reason four the pattern reads as generated and the claims are discounted with it. · 016 PASS·noted · the damage type is taught in the sentence that names it (entry 7); the ingredient supplier's name is glossed in a clause (entry 21). · 017 PASS · emotion (regret), curiosity (what they wish they had found, and which damage) and stakes (ten years) are carried by entries 2 and 3; the loop is closed in reason one. · 018 PASS · colour: grey salmon, the molecule's geometry, the placebo group getting worse, her routine drawn in grey with one thing in colour. · 019 N/A · no copy signed by a person. · 020 REVISE · one reviewer's name with two stories (entries 29, 47); the roadmap printed twice (entries 40 to 43 and 56); reasons four and five reprinted as answers; a framing line cloned (entries 28, 46). Cost: trust, on the second sighting. · 021 N/A · not an email sequence. · 022 PASS · sharp lines kept: the triple negative of the deck (entry 3) and the refusals of the reason-seven heading (entry 26) both pass the negation test. · 023 PASS · the first screen shows the object in hand and names what she gets (one softgel a day) before anything else. · 024 N/A · no product-as-decider verbs on a consumer supplement page. · 025 PASS·noted · notes for the claims check: all-five-star against a 4.8 average (entries 1, 24), week counts against the trials (entries 13, 42, 43), a certificate published against on request (entries 27, 58), two carrier oils (entries 21, 55).
Take
- Reasons ordered as a belief chain, each answering the question the last one raised: her routine, the molecule, the trials, why she never heard of it, why it failed when tried, who else, can the seller be trusted.
- Her routine first: reason one is about what she already owns, not about the new ingredient, and the picture for it draws her whole shelf (entries 5 to 7).
- Evidence used twice: the trials prove the outcome and set the buying standard (dose, source) two reasons before any product is named (entries 13, 21).
- The brand held back to 38% and named as the product that passes a test the reader already holds (entry 21).
- The first button only after problem, mechanism and proof (27%), offering a try under guarantee, not a purchase (entries 14, 15).
- Objections promoted to numbered reasons (four and five) instead of being left to the FAQ.
- The mechanism built as picture, then geometry, then number, with the number qualified (entry 10).
- Expectation-setting as a selling tool: a roadmap whose first step promises nothing visible, and a review that nearly gave up (entries 40, 30).
- One idea carried through the social proof: the review block and the thread are framed by the dose argument, and the thread replays the reasons in the page's own order.
- The guarantee's condition written in the page's two signature outcome words, so the promise and the refund test are one sentence (entry 35).
Leave
- The wrong-target framing of reason one as written (entries 3, 6, 7): for us it sits too close to telling her the routine failed, and it lists collagen among the things that fall short. Any collagen-broken-faster line is p2, any stalled-collagen story is p1; both stay banned.
- Claims outside the gate: the only-one absolute (entry 7), the perimenopause speed claim and the HRT statement (entries 7, 61), redness (entry 41), side-effect thresholds (entries 59, 62).
- The carrier-oil absorption claim (entries 21, 55): b8 is banned.
- The citations as printed: trial lengths and study types need re-reading from the papers before any are used (entries 13, 42, 43, 55).
- Source and testing details that contradict each other or look wrong: a Hawaiian origin for the named ingredient, a certificate published in one place and on request in another (entries 27, 58).
- Fabricated or unverifiable proof: a reviewer whose name carries two stories (entries 29, 47), a comment thread dressed as a social feed with timestamps that disagree (entries 47 to 49), reviewer photos nobody can verify, all-five-star against a 4.8 average (entries 1, 24), a logo strip worded to imply coverage (entries 45, 63).
- Lines that turn on her routine: one softgel replacing her stack (entry 34) and the instruction to stop covering damage with creams (entry 36). Ours adds to her routine and replaces nothing.
- The return-window argument (entry 27) and the 60-day window: ours is the 30-Day Money-Back Guarantee, which closes before skin outcomes are due, and it may not be called risk-free.
- A page with no price in text and no close: no button after the FAQ, a placeholder answer left live (entry 54), the roadmap and two reasons reprinted in the FAQ.
- The sentence habit: a dash in nearly every paragraph, the not-this-but-that pivot, and a run of fragments to end a paragraph. Our copy uses none of them.
Line by line · 63 entries
Every line is described in our words; the page's own text stays on the page (open it beside this table). Shaded rows carry the most pull.
| # | Beat | What the line does | Lever · reader | Why here | Craft | On our page |
|---|---|---|---|---|---|---|
| 1 3.1% | proof-social fine · 4 w | A row of five stars and a claim of more than 10,000 five-star customer reviews, set as an eyebrow directly above the headline. | social-proofconsensus solution-aware | It precedes the headline because the headline names an ingredient most of this market has never bought; a rating tells her the unfamiliar thing is already widely used. It also tells her at once that a product stands behind the article. | Ten thousand five-star reviews here becomes a 4.8 average over 10,000 reviews in entries 24 and 28. Both cannot be true: if every review had five stars the average would be five. | swap-proof Our real rating, stated once and identically everywhere, or the study count (r14) as the eyebrow while reviews are still few. r14 |
| 2 3.3% | hook-headline p · 15 w | Headline: seven reasons the named ingredient is the supplement that women over 40 wish they had come across a decade sooner. | curiosity-gaploss-aversionidentificationspecificity solution-aware | The headline names the ingredient and hides the brand. To a reader who has been sold to by every skincare brand, a claim about a category reads as information and a claim about a brand as an ad. The brand is held back to reason five, 38% down, after the dose argument has made it the logical answer. | Fifteen words and no adjective. The regret is borrowed from women slightly older than the reader, which turns ten lost years into a reason to read now. The plus sign after 40 lets a woman of 58 count herself in. | keep-structure A number, the ingredient named, the reader named. Our version carries the CS-A08 idea: what she already takes for her skin holds (m1), and most of her peers have not met astaxanthin yet (p7). Regret is usable only as wishing she had known sooner, never as years of product wasted (p1 and p2 stay banned). m1p7 |
| 3 4.8% | deck p · 33 w | Deck: a daily softgel aimed at a kind of skin damage three named staples (vitamin C, retinol, collagen) are said to miss; three ordeals spared (needles, peeling, a settling-in period); the result: brighter-looking, calmer skin. | open-loopcontrasteasenew-mechanism solution-aware | The deck plants the page's single open question (which damage?) before reason one answers it. The three spared ordeals are chosen from what this reader has actually been through with retinol and procedures, so ease is defined against her own history. | Two sentences, 33 words. The triple negative works because each item is a real memory of hers (vendor point 022, the negation test). The word calmer is unusual and accurate for a skin result, and it recurs through the page as its signature outcome. | rewrite-claim The gap is ours to open from p3: ordinary antioxidants guard part of the cell. But we add, we do not fault: her collagen and serums are doing their job, and The Cell Shield is what lets the result hold (m1, b9). Collagen is never listed among things that fall short. p3m1b9 |
| 4 0.2% | image img | Hero picture, at the very top on screen: a woman in her forties in a cream jumper, in a bathroom, framed like a selfie, holding the pouch beside her face. Its alt text repeats the headline. | identificationsocial-proof solution-aware | In the page's code it follows the deck; on screen it is what she sees first (0.2% down), above the stars. The pouch in her hand shows the brand name from the first screen, although the copy will not name the brand until reason five. | Same pouch design and same selfie device as the Rovina page's hero, with the brand name changed and a younger woman: the two stores appear to sell one white-label product to two ages. The alt text is a second copy of the headline, not a description. The picture has the smoothness of a generated image. | keep-structure A woman of 35 to 65 with the CellScience bag, in her own bathroom with her own products in view and in colour. Shot-list first; never presented as a named customer. Alt text written as copy, not as a repeat of the headline. |
| 5 6.1% | image img | Picture for reason one: a flat-lay of a whole skincare routine (serums, retinol, collagen powder, creams, rollers, a sheet mask, SPF) rendered in grey, with the pouch in full colour at the centre. Alt text repeats the reason's heading. | mirroringcontrastidentification solution-aware | On this page the picture comes before the heading, the reverse of the Rovina order, so she recognises her own bathroom shelf before she is told anything about it. The grey treatment makes reason one's argument without words. | The strongest image on either page: the copy's claim, drawn. Greying her products tells her they are lifeless, though, which the body copy then has to walk back. | keep-structure Close to the frame already written for CS-A08: her shelf of collagen and serums with one red softgel in front. Ours keeps her products in full colour, because our message is that they work and The Cell Shield protects what they do (m1, p6). m1p6 |
| 6 8.5% | failed-solutions p · 13 w | Reason 1 heading: the ingredient repairs something that three named staples of her routine were never built to repair. | vindicationcontrastopen-loop solution-aware | Reason one is about her products, not about astaxanthin, because this reader's first thought on meeting any new skin product is that she has tried things before. The page has to deal with the cupboard before it can add to it. It does so by design intent, not by failure, which spares her from admitting a mistake. | Thirteen words, one idea. The hinge is the phrase about what the products were built for: it moves the blame from her judgement to the products' job description. | rewrite-claim Our first reason keeps the order (her routine first) and changes the claim: what she takes is doing its work, and one softgel makes it hold (m1, b9). We do not say her products fall short of anything, and collagen is never on a list of things being corrected. Her fatigue with the routine is p6. m1b9p6 |
| 7 9.6% | villain-reveal p · 95 w | Three short paragraphs: her products are not failing, only aimed elsewhere; the missed damage is named (singlet oxygen) and said to break collagen fastest around perimenopause; astaxanthin is called the one antioxidant that neutralises it. | vindicationcommon-enemynew-mechanismspecificity solution-aware | Vindication opens and closes the paragraph with the villain in the middle, so the new information arrives wrapped in reassurance. The villain is a named chemical species, which makes the claim sound measured, not sold. Perimenopause is dropped in as a clause and left undeveloped until the FAQ: a marker that tells the reader of 40-plus the page knows her. | Short declaratives, most under thirteen words and the longest seventeen. The paragraph turns three times on one not-this-but-that pivot (vendor point 015); read aloud it persuades and sounds slightly mechanical. The absolute (the only one) is softened to a multiple by the page's own reason two. | rewrite-claim Usable: ordinary antioxidants guard part of the cell (p3), and astaxanthin's singlet-oxygen quenching is about 6,000 times vitamin C's (m3). Not usable: the only-one absolute, the perimenopause speed claim (not in the gate), and any line in which collagen is broken down faster than she can replace it (p2). Ours says what The Cell Shield protects (m1), not what is being lost. p3m3m1 |
| 8 12.7% | image img | Picture for reason two. Its content was not captured; the alt text is the reason's heading, word for word. | curiosity-gap solution-aware | Picture, heading, body again. Given the heading, this is where the salmon would be shown, giving an unfamiliar ingredient a familiar animal. | Alt text is used on every reason image as a duplicate of the heading, so a screen reader hears each heading twice and learns nothing about the picture (vendor point 011). | keep-structure Our picture here shows the softgel and its colour (b5 lives in photography) or the microalgae source (c4). Alt text describes the picture. b5c4 |
| 9 15.1% | mechanism p · 20 w | Reason 2 heading: the ingredient is the molecule that colours wild salmon pink, and it is said to be forming a shield within her skin at this moment. | specificitycuriosity-gapnew-mechanism solution-aware | Second, because reason one left her with a chemical name and a bare assertion. Before she will accept that one molecule does what three products cannot, she has to be able to picture it. Salmon gives the unfamiliar word a colour and an animal she has eaten. | Twenty words with a dash in the middle. The first half is fact; the second is a present-tense claim about her skin that cannot be true of a reader who has not yet bought. The shield image is what the page rests on, and our brand already owns that noun. | swap-mechanism The Cell Shield is named here, after the explanation: natural astaxanthin from microalgae (c4) that spans the cell membrane (m2). The salmon link is usable as a story about where the molecule is found; the present-tense claim about her own skin is not. m2c4 |
| 10 16.5% | mechanism p · 102 w | Three paragraphs: why wild salmon have pink flesh (algae that make a crimson, cell-shielding molecule); its shape, spanning the cell's membrane, with vitamin C outside and E inside; then a lab figure of up to 6,000 times vitamin C. | story-transportnew-mechanismcontrastspecificityreason-why solution-aware | Picture, then geometry, then number: each step needs the one before. The number would mean nothing without knowing what is measured, and the geometry is abstract until the salmon makes the molecule real. The final clause ties the figure back to her collagen, which closes the loop opened by reason one. | The best-built paragraph on either page. The figure is qualified twice (as a laboratory measurement, and for one damage type), which is unusually careful for this market and makes it more believable. The run of four very short sentences about vitamins C and E is the page's rhythm at its best. | swap-mechanism Transfers almost whole: m2 (spans the membrane), p3 (vitamin C outside, vitamin E at the membrane surface; the gate's wording on E differs from theirs), m3 (about 6,000 times vitamin C and about 550 times vitamin E), c4 (natural, from microalgae). Keep both qualifiers on the number. End on what holds (m1), not on collagen breaking. m2p3m3c4m1 |
| 11 19.7% | image img | Picture for reason three. Content not captured; the alt text is again the reason's heading. | borrowed-authority solution-aware | Picture, heading, body. A proof reason is where a reader expects a chart or a journal page. | Not judgeable from the working copy. | swap-proof A screenshot of the PubMed record or the journal page for the skin trial (b1): a house habit on science sections and better than any illustration. b1 |
| 12 22.2% | proof-study p · 11 w | Reason 3 heading: three fragments: a count of three trials in people, a measured fall in wrinkles, and results that can be seen by week eight. | borrowed-authorityspecificity solution-aware | Proof is third, straight after the mechanism, while the explanation is fresh: the trials are presented as the mechanism observed in women. It comes before the first button, so that the earliest exit is taken by a reader who has seen problem, explanation and evidence. | Three fragments with full stops, a staccato that suits a results line. Every fragment carries a number or a measurement. | swap-proof Human trials named by author and year with their real length (b1; the trial our brand file verified ran sixteen weeks in 65 women), the MMP-1 and UV-inflammation findings (b2), and x3 only alongside the real citations. b1b2x3 |
| 13 23.2% | proof-study p · 95 w | Three named trials: a 2017 one on wrinkle depth, moisture and elasticity at 12 mg, where placebo moisture fell; one on age spots and collagen at nine weeks; one on UV resistance. All, it stresses, used 12 mg natural. | borrowed-authorityspecificitycontrastopen-loop solution-aware | The evidence is used twice. First it supports the claim; then its dose and source become the standard every product will be judged by, two reasons before any product is named. By reason five she is holding a ruler that only one kind of product fits. | The placebo detail (the untreated group got worse) is the most persuasive fact on the page. The trials are named as Tominaga 2017, Nishida and Camera, and the citations need checking before anything is borrowed: our brand file records the 2017 trial at sixteen weeks, not eight, and one of the three papers is, to our knowledge, a cell study and not a trial in people. | swap-proof b1 and b2, with citations re-read from the papers themselves and a PubMed screenshot in the block. 12 mg as clinical strength against the 4 to 6 mg brands (b7); natural against synthetic (r16, c4). Every number re-measured from its source. b1b2b7r16c4 |
| 14 26.3% | guarantee 2× p · 34 w | A panel heading inviting her to join women who have already switched, then one line: a softgel a day, three outcomes, and a 60-day money-back guarantee under which she pays nothing if she sees no difference. | consensusrisk-reversalease solution-aware | The first ask comes at 27%, after problem, mechanism and proof and before the objections. The guarantee is the only offer term shown, so she is asked to try, not to buy. The brand is still unnamed in the copy: the panel sells a switch to the ingredient. | The switching verb assumes she is leaving something behind, which fits reason one's framing. No price here, or anywhere in the captured copy. | swap-offer Our first panel after the proof: one softgel a day (b9) and the 30-Day Money-Back Guarantee by name. She adds; she does not switch. Her routine stays. b9 |
| 15 27.9% | cta cta · 3 w | First button: a three-word invitation to try with no risk, today. It jumps to the product card lower on the same page. All five buttons carry this one label. | risk-reversaleaseurgency solution-aware | It comes after three reasons, not before the first as on the Rovina page: this page assumes a colder reader who needs the argument before the exit. | The label sells the absence of risk, not the product; for a reader burned before, that is the right thing to sell. Using one label five times makes the button furniture by the third sighting. | keep-structure A button after the first proof. Our guarantee may not be called risk-free (the house naming gate forbids it); the label names what she gets. |
| 16 28.6% | image img | Picture for reason four. Content not captured; alt text repeats the heading. | curiosity-gap solution-aware | Picture, heading, body. After the button panel it restarts the list for the reader who did not click. | Not judgeable from the working copy. | keep-structure A picture of the microalgae source or the softgel (c4); alt text describes it. c4 |
| 17 31.1% | objection p · 18 w | Reason 4 heading: the ingredient has spent thirty years hidden in fish feed, and a bracketed second sentence says that is why she is only hearing of it now. | curiosity-gapobjection-preemptpattern-interrupt solution-aware | With the case made, the page turns to the first doubt a convinced reader has: if it is this good, why has nobody told me. It is placed after the first button, so only readers who did not click (the doubters) meet it. | Animal feed is a deliberately unglamorous phrase on a beauty page; it interrupts. The bracketed aside reads as the writer leaning in, and it voices her objection for her. | keep-structure The same doubt exists for our reader and the gate answers it as p7: most of this market has never met astaxanthin. The thirty-year feed-supply history is not in the gate; verify it before using it as more than a light story. p7 |
| 18 32.2% | objection p · 75 w | Supply story: farmed salmon, grey by nature, are dyed pink with the ingredient, so it went to fish farms for thirty years; supplement makers sold synthetic low doses; women blamed the molecule. It closes by promising a right product, unnamed. | reason-whycommon-enemyvindicationopen-loop solution-aware | It answers why she has not heard of it and, in its last third, quietly opens the next objection (it did nothing when women tried it), so reason five arrives as a continuation. Its closing words promise a product without naming one: the reveal is set up a full reason ahead. | Grey salmon is the fact a reader repeats to a friend. The closing pair of short sentences is the page's tic (a balanced contrast, then a dash), though here it does real work as a hand-off. | rewrite-claim p7 (undiscovered by this market), r16 (synthetic astaxanthin comes from petrochemicals; ours does not) and c1 (about 64% of unverified products fail potency testing) as the reason earlier tries disappointed. No claim that ours is the first right product. p7r16c1 |
| 19 34.7% | image img | Picture for reason five. Content not captured; alt text repeats the heading. | contrast solution-aware | Picture, heading, body. Given the heading, a comparison of doses or bottles would be expected here. | Not judgeable from the working copy. | keep-structure If a dose comparison is drawn (b7), rival cells carry a reason and no rival brand is named or pictured. b7 |
| 20 37.2% | failed-solutions p · 16 w | Reason 5 heading: addressed to the woman who has already tried the ingredient and felt nothing; she is told the dose she was given was the wrong one. | vindicationobjection-preemptmirroring solution-aware | Fifth, because it speaks to the smallest and most resistant group: the reader who has tried astaxanthin. It needs reason three (the dose in the trials) and reason four (the trade underdosed) already in place. With them her bad experience becomes evidence for the page instead of against it. | A conditional addressed to one reader; everyone else reads it as proof that the page is honest about failures. The passive verb puts the fault on the seller. | keep-structure b7 makes this ours: 12 mg clinical strength against the 4 to 6 mg brands, addressed to the woman who tried a low-dose product. c1 supports it. b7c1 |
| 21 38.3% | product-reveal p · 97 w | Marketplace astaxanthin is described as 4 mg, synthetic, petrochemical and never trialled in people, and low dosing is put down to cost. Then the brand is first named: 12 mg of a branded natural ingredient in an MCT-oil softgel. | contrastcommon-enemyspecificityloop-closedreason-why product-aware | The brand appears at 38%, in the last third of the fifth reason, as the answer to a standard she has held since reason three. She is not introduced to a product; she is shown which product passes her own test. Named earlier, it would have made reasons one to four read as a pitch. | The paragraph closes on three verbless fragments, a staccato ending of the kind already used in reasons three and four, and wearing thin. Naming the ingredient's supplier (AstaReal) borrows a second brand's credibility. The carrier-oil absorption claim is the weakest link in the paragraph. | rewrite-claim CellScience Astaxanthin 12 mg is named here, as what meets the standard: 12 mg (b7), natural from microalgae (c4), not petrochemical (r16), grown in closed bioreactors (c5), third-party lab tested (c3). No carrier-oil absorption claim (b8 is banned); the permitted line is to take it with a meal that has some fat (r31). b7c4r16c5c3r31 |
| 22 41.3% | image img | Picture for reason six. Content not captured; alt text repeats the heading. | social-proof product-aware | Picture, heading, body. A consensus reason is where faces are expected. | Not judgeable from the working copy. | keep-structure Real customers only, once we have their permission; otherwise no faces here. |
| 23 43.8% | proof-social p · 11 w | Reason 6 heading: over 10,000 women are said to have switched already, with a promise to tell what they notice. | social-proofconsensuscuriosity-gap product-aware | Consensus is sixth, after the product is named, so the crowd is a crowd of this product's buyers. Placed before the reveal it would have been a crowd for the ingredient, which sells rivals too. | The second sentence opens a small loop (what do they notice) that the body closes at once. The count matches the eyebrow of entry 1. | swap-proof Our own customer count and what they report, once real and signed off (r17). r17 |
| 24 44.9% | proof-social p · 73 w | Composite timeline: calmer skin by week four, colleagues asking by week eight, the hunt for another product over by month three. Then unexpected extras (joints, tired eyes), a 4.8 rating over 10,000 reviews, and a denial of paid partnerships. | social-prooffuture-pacingspecificityconsensusrelief product-aware | The timeline tells her what to expect and when, which protects the sale from disappointment in week two. Month three's outcome (she stops shopping) is the emotional resolution of reason one's crowded shelf: the page ends her search, not just her wrinkles. | Three parallel sentences, one per step, and each step names a different kind of evidence: her own eye, other people, her own behaviour. It is the strongest rhythm on the page. The paragraph then ends on a stock pair about real women and real results. | rewrite-claim The expectation timeline is worth keeping, written from the trial record (b1) and not as customer fact. Extras from the gate only, benefit-forward: eyes (b3, b6) and joints (b4). Panel numbers only when signed off (r17). The month-three idea suits us exactly: she stops adding jars because what she has holds (m1, p6). b1b3b6b4r17m1p6 |
| 25 47.3% | image img | Picture for reason seven. Content not captured; alt text repeats the heading. | candor product-aware | Picture, heading, body, for the last time. It is the final picture before the review block and the product card. | Not judgeable from the working copy. | keep-structure The softgel, the bag and a lab report in one frame would say third-party lab tested (c3) without a caption. c3 |
| 26 49.8% | differentiation p · 10 w | Reason 7 heading: four fragments: a pure ingredient, then three refusals: synthetics, auto-ship traps, small print. | contrastobjection-preemptcandorcommon-enemy product-aware | Last, because the final barrier for this reader is distrust of supplement sellers, not doubt about the ingredient. The refusals name what she fears from the checkout, not from the product, immediately before the page takes her there. | All three refusals pass the negation test (vendor point 022): each names something a supplement buyer has actually been burned by. | keep-structure Our refusals must be true of our own checkout: one-time bags with no plan behind these tiers, third-party lab tested (c3), not synthetic (r16). No promise about small print if a trial lane is on the page. c3r16 |
| 27 50.6% | guarantee p · 75 w | Sourcing and testing claims (branded natural ingredient, Hawaiian microalgae, no synthetic solvents, every batch independently tested, certificate published); two refusals (forced auto-ship, short return windows); a direct address to the reader burned before; a 60-day guarantee. | risk-reversalcandorvindicationcommon-enemyspecificity product-aware | The list ends where the Rovina list ends, on the guarantee, but reaches it through the seller's conduct, not through her age. The remark about return windows that close before the molecule can work turns the guarantee's length into an argument that follows from the eight-week timeline. | The burned-before sentence is the page's most direct moment of empathy. Two details need checking before anyone borrows them: the branded ingredient named is not, to our knowledge, grown in Hawaii, and the certificate said here to be published is available only on request in entry 58. | swap-offer 30-Day Money-Back Guarantee, by name. c3 worded as third-party lab tested, c5 closed bioreactors, c4 natural source. Our window is shorter than the skin-trial timeline, so we do not borrow their argument about return windows. c3c5c4 |
| 28 53.4% | proof-social fine · 2× p · 30 w | A rating line (4.8 of 5 from over 10,000), a heading calling the clinical 12 mg difference beyond dispute, and a sub-line setting most brands' 4 mg synthetic product against what women see on the trial dose. | social-proofcontrastcommon-enemy product-aware | The review block is framed by the dose argument, not by the brand: the reviews are offered as evidence for the 12 mg thesis. That keeps the page's single idea running through its social proof. | Calling a rival's product filler is the page's only sneer. The heading's adjective is a claim about the reader's own reaction, which readers resist. | swap-proof Frame our reviews by our idea (what she takes holds, m1) and by the dose (b7). Real rating only. m1b7 |
| 29 55.1% | proof-social img · p · 3× fine · 53 w | Review card with a named photo: six weeks in, low expectations, her aesthetician asked whether she had changed her routine, and she had only added the softgel; skin smoother, less tired. No miracle, she says, just clearly better. | social-proofcandorspecificityidentification product-aware | It is the first review after the list and is chosen for its third-party witness: a professional who noticed unprompted. That is stronger proof than her own opinion, and it echoes the week-eight step of entry 24, where other people start to ask. | Short, flat sentences and an admitted limit make it sound like a person. But the same first name and initial returns in entry 47 with a different backstory and a different number of weeks: one reviewer, two stories. Treat as written in-house; the photo cannot be verified either. | drop No invented reviewers. The pattern to keep once reviews are real: a witness who sees her skin professionally, and an admitted limit. |
| 30 55.1% | proof-social img · p · 2× fine · 55 w | Second review card, with photo: she nearly quit at three weeks having seen nothing, stayed, and by week five or six dry patches on her cheeks had settled and her skin looked livelier. Worth it, she says, but slow. | social-proofcandorobjection-preemptcommitment product-aware | It is paired with the first review as its slow counterpart. A buyer who sees nothing in week two now has a named precedent for waiting, which matters commercially: the guarantee runs sixty days and the promised result arrives at eight weeks. | The most useful review on either page, because it sells patience. It carries no verification tag although the first card has one: the template was filled unevenly. | drop No invented reviewers. When ours are real, lead with the slow-result review: with a 30-day guarantee and skin outcomes measured in weeks, setting honest expectations is our job too. |
| 31 59.4% | cta cta · 3 w | Second button, same three-word label, under the two review cards. | risk-reversal product-aware | Reviews, then button: it takes the reader persuaded by other women straight to the product card that begins one screen below. | Identical to the first button. | keep-structure As entry 15. |
| 32 60.1% | image 2× img | Two product pictures back to back, a gallery; both carry the product's name as alt text. | none most-aware | At 60% the page becomes a product page. Every button above jumps to this point. | Alt text is the product title. | keep-structure Our pack shot, the opaque bag, with alt text written as customer copy. The bag is shown and never argued for (c6 stays banned). |
| 33 63.5% | product-reveal fine · h1 · 7 w | The product card opens with a rating line (4.8 of 5, over 10,000 reviews) and the product's name, a plain descriptor of strength with no brand word in it. | social-proofspecificity most-aware | The name is generic on purpose. After seven reasons about dose, the product title is the dose argument compressed into three words. | The only top-level heading on the page is the product name, so to a search engine the page is about the product, not about the seven reasons. | keep-structure Our card: CellScience Astaxanthin 12 mg, with The Cell Shield named beside it and 12 mg stated as clinical strength (b7). b7 |
| 34 64.5% | product-reveal 4× p · 59 w | Four benefit lines, each a label and a dash: smoother, firmer skin with a clinical claim; hydration from inside the cell; a daily shield with the 6,000-times figure; and one softgel said to replace her whole antioxidant stack. | specificityborrowed-authorityease most-aware | The card restates the argument for the reader who arrived by button. It runs outcome, outcome, mechanism, ease, which is the order of a product page, not of the list above it. | The fourth bullet contradicts reason one: there her products were fine and aimed elsewhere; here one capsule replaces them. The 6,000 figure has lost both qualifiers it carried in entry 10. | rewrite-claim Four rows, one line each: wrinkle depth and elasticity measured in trials (b1); hydration (b1); about 6,000 times vitamin C against singlet oxygen, with its qualifier (m3); one softgel a day that replaces nothing (b9). Never a line about replacing her stack. b1m3b9 |
| 35 67.2% | guarantee p · fine · 23 w | The guarantee, named after the outcome as a 60-day glow guarantee: a refund unless she sees brighter skin and feels it calmer; traps and awkward returns are ruled out. | risk-reversalcandorspecificity most-aware | It is the last element of the card, directly under the bullets. No price, bundle or bonus appears in the captured copy around it, so in text the guarantee is the whole offer. | The guarantee's condition is written in the page's two signature outcome words, so the promise and the refund test are one sentence. Naming the guarantee after the result makes it memorable. | swap-offer 30-Day Money-Back Guarantee: house naming is fixed, so no outcome-branded name. Full refund, she keeps the bag. Our price tiers are stated in text beside it. |
| 36 68.3% | proof-social 2× p · 27 w | A heading inviting her to join thousands said to glow from inside, then a line urging her to quit covering damage with creams and pointing to the 10,000-plus women said to rely on the 12 mg formula. | consensuscontrastsocial-proof most-aware | Below the product card the page starts again for the reader who did not buy, and its first move is more faces. The heading introduces two video clips. | It tells her to stop covering damage with creams: the first line on the page to criticise what she does, and a break with the promise of reason one. | rewrite-claim A heading for customer clips in our voice: women adding one softgel to the routine they already have (b9). Nothing that calls her creams a cover-up. b9 |
| 37 69.9% | image 2× video | Two portrait videos side by side, phone-shaped; nothing of their content was captured. By position they are customer clips. | social-proofidentification most-aware | Video sits directly below the product card, where a hesitant reader looks for someone like herself using it. | Not judgeable from the working copy. | keep-structure Real customer clips only, in spoken register; none until we have them. |
| 38 74.4% | cta cta · 3 w | Third button, same label, under the clips. | risk-reversal most-aware | Clips, then button: every proof block below the card ends on the same exit, which sends her back up to the card. | Identical to the first button. | keep-structure As entry 15. |
| 39 75.1% | results-timeline img · 2× p · 16 w | An image and heading that call what follows her path to resilience, and a sub-line: ageing can't be avoided, accumulated damage can, and here is the route. | future-pacinghopecandor most-aware | The roadmap comes after the card and the clips because it answers the hesitant buyer's next question: how long before I know. It expands the three-step timeline of entry 24 into four dated steps. | Conceding that ageing cannot be avoided is a candid note that buys belief for what follows. | keep-structure A dated expectation strip suits us, headed plainly. The concession about ageing is fine; resilience is our word too (m4). m4 |
| 40 78.9% | results-timeline 2× fine · p · 16 w | Week 1, labelled as cellular defence: protection is said to begin inside and oxidative stress to steady across the body. Nothing visible is promised. | future-pacingcandor most-aware | The first step is invisible on purpose. It gives week one a meaning (something has started) while asking her to expect nothing in the mirror, which protects the later steps from an early verdict. | The honest part of the roadmap. Stress levels steadying in a week is asserted, not sourced. | rewrite-claim A first step that admits nothing is visible yet is worth keeping. We describe what the softgel does (m2, m3), not a measured change in her own stress levels. m2m3 |
| 41 79.6% | results-timeline 2× fine · p · 19 w | Week 4, labelled as a calmer complexion: redness and irritation are said to begin easing once the molecule has built up fully. | future-pacingspecificity most-aware | The first visible step is the modest one, calm and not yet youth, and it is dated inside a single bag's supply so the first purchase has a checkpoint. | The build-up clause supplies a reason for the four-week wait, which is what makes the wait tolerable. | rewrite-claim We hold no redness claim. If an early marker is wanted, the gate's is eyes at 14 to 28 days (b6); skin outcomes follow the trial record (b1). b6b1 |
| 42 80.4% | results-timeline 2× fine · p · 27 w | Week 6, labelled as even tone: age spots and lines are said to look softer, with a named trial (Nishida) cited for a reduction seen at week nine; skin is said to feel more hydrated. | future-pacingborrowed-authorityspecificity most-aware | Citations enter the roadmap at the third step, where the claims become ones she can check in a mirror. Evidence is attached exactly where doubt would start. | A week-six claim is supported by a nine-week measurement: the citation does not cover the step it is attached to. | rewrite-claim Steps dated from the trial record only (b1), each step carrying the citation that actually covers it. b1 |
| 43 81.3% | results-timeline 2× fine · p · 41 w | Week 8 on, labelled as a full transformation: visibly better elasticity, described as in line with a named trial (Tominaga) at eight weeks; brighter, more even skin; and the point at which other people start to ask what has changed. | future-pacingborrowed-authoritysocial-proof most-aware | The last step is dated at the trial's claimed length and inside the 60-day guarantee by days, so the full result and the refund deadline arrive together. It ends on other people noticing, the page's recurring picture of success. | The phrase tying the step to the trial is a careful hedge. The label for this step is the one overclaim in the roadmap. The closing social cue is the third appearance of that idea. | rewrite-claim Elasticity and wrinkle depth from b1, at the trial's real length (our brand file records sixteen weeks for the 2017 trial). Our 30-day guarantee closes before skin outcomes are due, so the strip must say plainly what she can expect inside 30 days and what comes later. b1 |
| 44 82.5% | cta cta · 3 w | Fourth button, same label, under the roadmap. | risk-reversal most-aware | Roadmap, then button: once she knows how long it takes, she is invited to start the clock. | Identical to the first button. | keep-structure As entry 15. |
| 45 83.2% | authority-stamp p · 6 w | A one-line lead-in to a strip of publication logos, worded so that it is research on the ingredient, not the product, that was featured. The logos themselves were not captured. | borrowed-authority most-aware | Press credibility is held back to 83%, among the late reassurances. Coming after the roadmap's citations, it turns named trials into a general impression that the ingredient is established. | Carefully worded: it claims nothing for the brand. A reader skimming logos will not make that distinction, which is the point. | swap-proof Name the journals the trials were published in (x3 paired with its citations, r14). No press-logo strip unless the coverage is of us. x3r14 |
| 46 84.1% | proof-social fine · 2× p · 19 w | A trust line (10,000-plus), a heading placing the proof in a glow at eight weeks, and a sub-line inviting her to see how women fare on the dose the research used. | social-proofspecificity most-aware | It frames the comment thread by the dose argument and the eight-week mark, as the review block was framed in entry 28. The sub-line is nearly the same sentence as that block's sub-line. | The same framing sentence in two blocks is a cloned line (vendor point 020). | swap-proof One framing line for social proof, used once. |
| 47 85.5% | comments 5× fine · 51 w | First comment, three days old: a woman who had tried everything (two years of retinol, three vitamin C serums, microneedling) and still looked tired says that after eight weeks on 12 mg her aesthetician asked what had changed. | social-proofidentificationmirroringspecificity most-aware | The three comments replay the page's three arguments in customers' mouths and in the page's own order. This one is reason one: the full routine that was not enough. Its list of what she tried is the reader's own shelf. | The thread is dressed as a social feed, with like and reply links and ages of posts. The name and the aesthetician anecdote are those of the review card in entry 29, with eight weeks in place of six and a new backstory. Treat the thread as written in-house. | drop No fabricated comment threads. If real comments exist they are shown as what they are, with their source. |
| 48 86.9% | comments 5× fine · 46 w | Second comment, a week old: three months on a cheap marketplace version did nothing; she switched for the higher dose and the microalgae origin, and at week six her husband said she looked as if she had been on holiday. | social-proofcontrastvindication most-aware | This is reason five as testimony: the low-dose product that failed, then the switch. It speaks to the reader who has already had one go at astaxanthin and needs permission for a second. | A customer citing dose and source as her reasons for switching speaks in the page's own argument, not in a customer's words. The holiday compliment is the same social cue the Rovina page builds a whole reason on. | drop No fabricated comments. The switching story is only usable from a real customer. |
| 49 88.1% | comments 5× fine · 47 w | Third comment: a woman in perimenopause says everything changed at once, nothing she had relied on for a decade still worked, and she reports the first visible difference since; her skin has structure again. Its two timestamps disagree. | social-proofidentificationmirroringhope most-aware | The last comment carries the perimenopause marker that reason one planted in a clause and the list never developed. It is the page's most specific address to the woman of 40-plus, and it comes last so that she leaves the thread on her own situation. | The word structure is a good, unusual one for how skin feels. The post is dated two days old in its header and five hours old in its footer: a template filled in carelessly. | drop No fabricated comments and no life-stage claim the gate does not hold. The gate's nearest id is r12, which is not a skin claim. |
| 50 90.2% | cta cta · 3 w | Fifth and last button, same label, under the thread. | risk-reversal most-aware | It is the final button, at 90%. Nothing below it asks for the sale again: the FAQ and the closing line run on with no exit. | Identical to the first button. A page that ends without a button under its FAQ leaves the answered reader nowhere to go. | keep-structure As entry 15, and one more button after our FAQ. |
| 51 90.9% | faq 2× p · 20 w | The FAQ heading, phrased as questions the brand encouraged her to ask, and a sub-line: the brand knows she has been burned before and offers to be honest about itself. | candorvindicationmirroring most-aware | The FAQ opens at 91% and holds 39% of the page's words in 7% of its height, so the answers sit in closed accordions. The heading recasts objections as questions the seller welcomes, in keeping with reason seven. | The burned-before line returns from entry 27; a second use weakens the first. | keep-structure An FAQ framed as straight answers. Our PDP section is already named for honest answers. |
| 52 92.3% | faq p · fine · 97 w | Question: why is it new to her. The answer retells reason four at length (fish farms took the supply for thirty years), adds Japanese researchers publishing human trials from the early 2000s, and hints that the silence was deliberate. | objection-preemptreason-whycommon-enemy most-aware | The first two questions are reasons four and five again, for the reader who jumped past the list to the card. The FAQ is a second, compressed copy of the argument and begins with the two objections. | It repeats reason four almost sentence for sentence (vendor point 020). The closing hint that the gap was deliberate is an insinuation with nothing behind it. | rewrite-claim Answer from p7 in two sentences. No supply-conspiracy hint. p7 |
| 53 92.8% | faq p · fine · 92 w | Question: an earlier try that did nothing. The answer blames dose and source and clears the molecule: most products are 4 mg synthetic from petroleum, the trials used 12 mg from microalgae, and an underdosed bottle tested an imitation. | vindicationobjection-preemptcontrast most-aware | It is reason five again, placed second, so the two objections that opened the second half of the list also open the FAQ. | The last sentence, separating the molecule from a cheap copy of it, is the sharpest line in the FAQ. The remainder duplicates the list. | rewrite-claim b7 (12 mg against 4 to 6 mg), r16 (synthetic is petrochemical; ours is not), c1 (the category's failure rate). Said once on our page, not twice. b7r16c1 |
| 54 93.3% | faq p · fine · 18 w | Question: why dose matters and how this differs from marketplace products. The answer is the page editor's two-word placeholder; it was never written. | none most-aware | The question restates the page's central argument, which makes the missing answer the most exposed gap on the page. It has presumably survived because the accordion is closed by default. | A software default left live (vendor point 012). Any reader who opens it learns how the page was assembled. | rewrite-claim If we ask the dose question we answer it: b7, with the trial dose cited from the paper (b1). b7b1 |
| 55 93.9% | faq p · fine · 91 w | Question: softgel against cream. Answer: creams work at the surface; the damage sits within the membrane of the cell; a fat-soluble molecule swallowed with fat reaches skin cells through the blood. A 2012 paper (Tominaga) is cited for oral use. | reason-whycontrastborrowed-authority most-aware | This reader's habit is topical, and the format itself needs defending. The question is held back to the FAQ because the list made the same point by geometry in reason two; here it is restated for the literal-minded. | Clear cause and effect in four steps. The carrier named here is coconut oil; in entry 21 it is MCT oil. A second paper from the same author appears, dated 2012, which the list never mentioned. | rewrite-claim From within: it spans the membrane (m2) and restores the small-vessel flow that carries nutrients to skin (m5). Her creams stay; the softgel works where they do not reach (p3). No carrier-oil claim (b8 is banned); with a meal that has some fat (r31). m2m5p3r31 |
| 56 94.4% | faq p · fine · 117 w | Question: how soon. The answer reprints the four-step roadmap (weeks 1, 4, 6 and 8-plus) nearly unchanged, adds that results vary, and asks her to hold her verdict until she has taken the full dose for the full time. | future-pacingcandorcommitment most-aware | It repeats the roadmap of entries 40 to 43 for the reader who met the FAQ first. Its new final sentence is the real message: judge only after the full dose for the full time, which carries her to the edge of the guarantee window. | At 111 words the longest answer on the page, and four fifths of it is a reprint (vendor point 020). The results-vary clause is the only hedge of its kind on the page. | rewrite-claim One timeline, stated once on the page, from the trial record (b1); early sign from b6 if used. b1b6 |
| 57 94.9% | faq p · fine · 78 w | Question: what if nothing happens. The answer: a refund, unconditional, with no forms and nothing to post back; then a jab at brands that hide behind hard returns and do not believe in their product. | risk-reversalcandorcommon-enemycontrast most-aware | The refund question comes straight after the timeline because a reader told to wait eight weeks next asks what happens if the wait was for nothing. The guarantee is argued as evidence of belief, not only as insurance. | The answer ends on the page's shortest sentence, two words, which lands. The three refusals are concrete and checkable. | swap-offer 30-Day Money-Back Guarantee: full refund, she keeps the bag. We state our terms and make no remark about other brands' returns. |
| 58 95.3% | faq p · fine · 60 w | Question: is it tested by a third-party lab. The answer: yes, each batch, independently, for purity, potency, contaminants and heavy metals; a certificate can be requested; dose and source are said to match the published trials. | objection-preemptspecificityborrowed-authority most-aware | Testing comes after the refund question: once the money is safe, she asks whether the product is. It repeats reason seven's claims for the reader who skipped the list. | Here the certificate can be requested; in entry 27 it is published with every product. One of the two is wrong, and a reader who opens this answer after reading reason seven will notice. | swap-proof c3 worded as third-party lab tested, with results published per batch once the first certificate exists; c5 for purity (no heavy metals, no environmental contaminants); c4 for the source. c3c5c4 |
| 59 95.8% | faq p · fine · 77 w | Question: is daily use safe. The answer: yes, studied at 6 to 12 mg without notable side effects; above 16 mg some report orange-tinted urine, called harmless; pregnant or nursing women and anyone on prescriptions should consult a doctor. | objection-preemptcandorspecificityborrowed-authority most-aware | Safety is the eighth question, among the late, careful ones. A reader this far down is reading the label, and the answer treats her as someone who does. | Volunteering a side effect at a higher dose is candour that buys belief for the dose being sold. The doctor caveat is standard and well placed at the end. | keep-structure r27 (allergen and age safety) and r28 (medication, with the blood-thinner caution). We do not quote side-effect thresholds the gate does not hold. r27r28 |
| 60 96.3% | faq p · fine · 70 w | Question: what if she has health concerns. The answer: see her doctor, above all if she takes anticoagulants, immune-suppressing drugs or hormones; the safety record is called strong; she is invited to email and promised candour, not selling. | candorobjection-preemptrisk-reversal most-aware | It overlaps the safety question before it and exists for its last sentence: an invitation to write in. For a wary buyer an address to write to is itself proof that someone is there. | The deference to her doctor's fuller knowledge of her is respectful without being timid. Nearly a duplicate of the previous question. | keep-structure r28, including the blood-thinner caution, and our support address. One safety question, not two. r28 |
| 61 96.8% | faq p · fine · 91 w | Question: perimenopause or HRT. The answer: hormone shifts are said to speed oxidative stress and collagen loss; the product was built for that window; it is no hormone and is said not to interfere with HRT; ask the doctor. | identificationmirroringobjection-preemptcandor most-aware | Perimenopause was planted as a clause in reason one and voiced in the last comment; only here, second to last in the FAQ, is it addressed head on. The page's most specific audience marker is kept in the least-read place, where it cannot narrow the headline's reach. | The claim that the product was built for this window is positioning with nothing on the page to support it. The no-interference statement is a medical assertion made flatly. | rewrite-claim The gate holds no perimenopause or HRT claim. The nearest id is r12, a pillar claim and not a skin one. Our answer is r28: talk to her doctor. If the life stage is to be addressed it needs its own gate entry first. r28 |
| 62 97.3% | faq p · fine · 65 w | Question: will it stain teeth or turn skin orange. The answer: no; above 16 mg a faint orange colouring of urine is reported, rarely at 12 mg; the colour stays inside the body. | objection-preemptcandorspecificity most-aware | The last question is the smallest worry, and an odd one to end on. It also shows the distance between the two pages: here skin colour from the softgel is a fear to be denied, while the Rovina page sells it as the headline benefit. | Repeats the urine detail of entry 59. A flat denial first, then the detail. | keep-structure Fair to answer if asked; r27 covers safety. We promise no colour change to skin in either direction; the glow (b5) is shown in pictures. r27b5 |
| 63 98.0% | authority-stamp p · 6 w | The lead-in to the logo strip again, the same line as entry 45, as the page's last words. Logos not captured. | borrowed-authority most-aware | The page ends on the logo strip, with no closing argument and no button after the FAQ. Whatever she decided, she decided above. | A repeated line as the last line (vendor point 020). The page has no close. | drop Our page ends on a short close and a button, not on a logo strip. |