← prelander teardowns

Top 5 Fatty Liver Supplements of 2026 – Clarity MD

claritymd.co · family top5 · open the live page ↗ · kept on the prelander swipe · found on the store (not yet seen behind an ad)
ArchitectureF4 — Auto tag F4, confirmed: the spine is five ranked cards under a scoring frame. One first-person narrator (the named reviewer) where the house F4 has none. She has no scene and no turning point, only a byline and a verdict voice, so this is a ranking with a narrator attached, not a story. From entry 11 the same voice speaks of the winner's customers as its own, so reviewer and seller are one. Narratora named woman reviewer with no title, outlet or credential, who says she tested the products herself (editor, 1) Big ideaLiver supplements fail on amount and form, not on the idea: judge them on three standards and only one product passes all three. Hook devicefirst-person tester with a count VillainUnder-dosed, crowded liver formulas: a dozen or more ingredients spread thin, amounts hidden in blends, plain capsules the body takes up poorly. What they call itNone named. The working idea is a clinically dosed antioxidant the liver can use: 12 mg of natural astaxanthin in an oil softgel. The three-part rubric (dose, source, delivery) does the job a named mechanism would do. Proof stackthe reviewer's own testing (50 looked at, 15 tried, 5 ranked): asserted, never shown · 12 mg described as what most clinical studies used: no study named or counted · a poll of 243 of the brand's own customers at 8 weeks or more, self-reported, with a footnote admitting only 34% had follow-up blood work · a customer count above 142,000 plus a 4.8-star rating: no source · doctor recommended: no doctor named · third-party tested, CGMP facility in the US, clean label: stated, no report shown · rivals' own evidence graded in the fine print (own survey, sponsored study, informal challenge, testimonials) · letter grades A+ to C-: no method Offer$29.99 stated as today's price; a first-order discount of as much as 70% with no base price given; free US shipping; a 90-day money-back guarantee tied to feeling better; one softgel daily in pouches of 30; sold only on the brand's own site. No bundle prices and no bonus on this page. Reader enters / leavessolution-aware → most-aware Product first named16% down the page · 3 buttons, the first at 44%

The two chiefs

House chief · 7/10

Top-5 ranking advertorial on the seller's own store, dressed as a review site · solution shoppers from search, native and retargeting (found on the store's page list, not yet seen behind an ad) · adults with a fatty liver diagnosis or raised liver enzymes, likely 45 and over · solution-aware in, most-aware out · no named mechanism: a three-part rubric built from the winner's spec

Lead: Method lead (how the ranking was made), followed by a second-person problem and a pivot. Grade B+: the replayed blood test is strong, but it arrives after the method, and the tester is never shown testing.
Offer: A clear price, free shipping and a long guarantee tied to feeling better. The discount has no base and no reason, and there is no bundle or bonus on the page. Grade B-.
Close: Three recaps and three links, only the last a hard ask. Risk reversal is strong; urgency is limited to one unsupported remark about stock. Grade C+.
Throughline: Disappointment, vindication, a rule, a fair-looking list, relief. It holds to the end of the winner's card, survives the rival cards through specific faults, and sags in the recap.

Strongest

  • entries 6 to 14: criteria before candidates, so the reader agrees to the rule before meeting the winner
  • entries 15 to 20: proof placed ahead of the list, with a candid footnote
  • entries 34 to 57: rivals praised first and marked down only on checkable facts, the two higher ones compared with the winner by name
  • entries 39, 45, 51, 57: fine print that grades each rival's evidence, weaker card by card
  • entries 33 and 66: a soft link after the winner, the hard ask held for the end

Weakest

  • entry 2: the central claim of personal testing has no support on the page
  • entries 11 and 15: reviewer and seller speak with one voice
  • entries 21 and 22: grades with no method
  • entry 26: almost no explanation of how the product works
  • entries 58 to 65: a close that repeats itself

Craft chief · REVISE

For two thirds of its length it sounds like a careful friend who has done the homework; then it says its conclusion three times and stops sounding like a person.

Strongest

  • entry 9: the reader's failed purchase replayed in two sentences with one concrete object
  • entry 10: the pivot that clears the idea and blames the product, which makes a second purchase reasonable
  • entries 12 to 14: a rubric cut from the winner's own spec and taught before any product is named
  • entry 20: a footnote that prints the unflattering denominator and buys belief for the whole block
  • entries 40 to 43: a rival's best proof turned into its markdown by naming what the study measured

Weakest

  • entry 2: a testing claim the body never shows
  • entries 11 and 15: the reviewer speaks of the winner's customers as her own, which ends the pose of independence
  • entry 24: two cons that are benefits in disguise
  • entries 58 to 65: three recaps in a row with no new information
  • entry 26: the working principle is one clause long

001 PASS·noted · entries 9 and 10 have a spoken rhythm and the lists scan; the close (entries 58 to 65) repeats one cadence three times · 002 PASS·noted · all four jobs present: attention 2 to 3, belief 6 to 21, desire 22 to 31, action 33, 64, 66. The links never say what happens after the click · 003 PASS · plain words throughout; the technical names on rival cards are explained where they appear (entry 35) · 004 PASS·noted · each section hands to the next as far as entry 33; entries 58 to 65 add nothing new and the slide goes flat · 005 REVISE · entry 2 promises a personal test of fifteen products and the body never shows a test, a protocol or the other ten; the second half of the promise is paid in entries 12 to 14. Cost: trust, for any reader who looks for the testing · 006 PASS · specifics carry the page (12 mg, 243, 34%, 571 mg, $77, 14 people, 50 people over 30 days). Noted: the discount in entries 23, 30 and 61 has no reason and no base price · 007 PASS·noted · spacing left in the rows suggests inline bold on the dose, the customer count and the no-fillers phrase (entries 26, 28, 31): emphasis sits on the numbers · 008 PASS·noted · one intensifier adverb recurs eleven times across the page; rival verdicts and bottom lines repeat their own cons (entries 37, 38) · 009 N/A · a ranking, not a story lead; the only scene is the second-person blood test in entry 9 · 010 PASS · contractions, first person, asides in brackets (entry 59) · 011 PASS·noted · price, claims and links are in text; the hero has empty alt text (entry 8) and no rival product image appears in the rows · 012 PASS·noted · footer is a bare copyright line (entry 67); the disclosure label is generic (entry 1); the reading time is decoration (entry 4) · 013 PASS · entry 9 names the reader's real frustration in the reader's own terms · 014 PASS·noted · disappointment, vindication, a rule, relief: proportional to a ranking. The close never returns to the blood test, so the feeling is not paid off · 015 REVISE (tier 1, flagged not rewritten) · a stock setup phrase opens entry 10; a quotable contrast line opens entry 56; a polished closing pair in entry 65. No long dashes on this page · 016 PASS·noted · entry 12 points at an antioxidant one section before it is named in entry 13; no house names elsewhere · 017 PASS · emotion (money wasted, entry 3), curiosity (what decides, entry 2), stakes (the liver, the lab number); the loop closes in entries 12 to 14 · 018 PASS · color named: the footnote that prints the denominator (entry 20), the rival study that measured something else (entries 40 to 43), the checkout that enrols a membership (entry 48). Gray: the winner's pros and the close, which any astaxanthin seller could paste · 019 N/A · no founder. Noted: a first-person verdict signed by a reviewer who cannot be verified (entry 25) · 020 REVISE · the card template, link labels, fine print and footer are shared across the store's ranking pages, and the three standards are stated six times on this page (entries 7, 12 to 14, 22, 23, 30, 59). Cost: by the second sighting it reads as a template · 021 N/A · a page, not a sequence · 022 PASS · nothing flagged for boldness; the blunt claim in entry 10 and the winner superlatives stand as written · 023 PASS·noted · the first screen names what the reader leaves with, a way to judge; the counts in entry 6 come after it · 024 PASS · the product supports and helps manage; outcomes are reported as customers' own accounts (entries 17, 20) · 025 PASS·noted · 243, 34% and 72% agree with each other (entry 20). For the guard: customer count and rating (entries 23, 28), reading time (entry 4), the base of the 70% discount (entry 30)

Take

  • Sell the rubric before the winner: the headline and the first fifth of the page teach how to judge, so the winner arrives as the result of a test the reader has already agreed to (entries 2, 6, 7, 12 to 14).
  • Cut the criteria from the winner's own spec sheet, then generalise them into three chips that every card is scored on (entries 7, 22, 34).
  • Replay the reader's own failed purchase in the second person, then clear the idea and blame the product: two short moves that make buying again reasonable (entries 9, 10).
  • Put the proof block between the criteria and the list, so results reach the reader before the winner's card does (entries 15 to 19).
  • Print the denominator: a footnote that admits the subset and the self-report buys belief for every figure above it (entry 20).
  • Rank the winner first, then let four rival cards finish the reader's comparison shopping on the page (entries 22, 34).
  • Praise every rival first and mark it down only on facts a reader can check; compare the higher rivals with the winner by name and judge the lower ones against the rubric; let pros fall and cons rise with rank; call the runner-up the best of the rest (entries 34 to 38, 52 to 56).
  • Turn a rival's best proof into its markdown by naming what that proof measured (entries 40 to 43).
  • End each rival card on who should still buy it (entries 38, 44).
  • Use the fine print under each card to name the basis of that card's results (entries 39, 45, 51, 57).
  • A soft, review-register link under the winner's verdict and the hard, guarantee-led ask only at the very end (entries 33, 66).
  • Restate the winner as a short checklist straight after the last rival, while the contrast is fresh (entries 58, 59).
  • Proportion: about a fifth of the words before the list, a fifth on the winner, half on the rivals, a tenth on the close.

Leave

  • Every liver, fatty liver and liver enzyme claim: CellScience holds none. The nearest gate claims are kidney numbers, brain fog, energy, eyes and joints.
  • The reviewer who cannot be verified, the claim of hands-on testing of fifteen, and the funnel of fifty to five (fabrication risk: asserted, never shown).
  • The reviewer's voice speaking as the brand (entries 11, 15).
  • Their survey figures, their customer count, their rating and a doctors' recommendation with no doctor.
  • Delivery-form superiority (oil softgel against powder): it sits beside the banned carrier-oil claim.
  • Guarantee length as a ranking weapon: ours is 30 days, shorter than several rivals'.
  • Cons that are benefits in disguise, and the unsupported sell-out remark (entry 24).
  • Letter grades with no method, and a second layer of unweighted editorial factors (entries 21, 22).
  • A percentage discount with no base price and an implied deadline with no date (entry 30).
  • The triple recap and the late name line (entries 58 to 65).
  • A publication name that appears only in the footer and does not match the domain (entry 67).
  • Their sentences. Nothing on this page is to be reused as wording.

Line by line · 67 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.

#BeatWhat the line doesLever · readerWhy hereCraftOn our page
1
0.1%
disclosure
cta · 2 w
A two-word control above the headline labels the page as advertising. It is a link to a notice, not the notice itself, and it does not say the top-ranked product belongs to the publisher.candor
solution-aware
Above the headline it is seen before the reader cares about anything, so it costs nothing. After the ranking it would land as a confession. Nothing has to be believed yet, and every first-person claim below can stand without a later reveal.Generic label in small type. It does the legal job and no selling job. The words name the practice and hide the relationship: the winner is the store's own product.page-furniture
CS-P1 carries its ad disclosure only in the footer fine print. A short disclosure line under the masthead costs nothing and protects the page with ad reviewers. CS-P1 lacks it at the top. No claim needed.
2
0.6%
hook-headline
h1 · 11 w
In the first person, the reviewer says she personally tried fifteen well-known products for a fatty liver, then promises the short list of things that decide whether one is worth buying.earned-authorityspecificitycuriosity-gap
solution-aware
It is the one line every visitor reads, so it sets the contract: this is a test report. The reader must already believe a supplement can help and distrust the shelf. Because it promises criteria and names no winner, the first fifth of the page can teach a rubric without reading as a pitch.Eleven words, one colon, a count odd enough to sound counted. The condition is named plainly, so only that reader stops. The weak point is the verb: no test, protocol or list of the other ten products ever appears (the testing is asserted, never shown).rewrite-claim
No liver word. CS-P1's H1 is a plain category ranking naming kidneys, brain, eyes and joints, and its subhead counts five softgels and three tests. It has no first-person tester. We cannot claim hands-on testing of fifteen: we read five rivals' own product pages on a dated day. The honest twin sells the three label checks (the full 12 mg in one softgel, third-party lab tested, cost per day).
b7c1c3
3
1.5%
deck
p · 20 w
One sentence names two sources behind the list, the published science and a poll of buyers, then splits the category into products that truly help the liver and costly pills that do nothing.borrowed-authoritysocial-proofloss-aversioncontrast
solution-aware
Directly under the headline it answers the first doubt (tested how?) before the byline can raise it. The reader must accept that most of the shelf is worthless, which a burned buyer half believes already. It announces both sources in advance, so the standards and the survey arrive as promised.Twenty words, one breath. The closing two-word insult is the only colored phrase on the first screen, and it is aimed at products, never at the reader. It leaves out that the buyers polled are the winner's own.rewrite-claim
No liver claim. Same shape with our sources: the labels as read and the lab results. Our enemy is in the gate: about 64% of unverified astaxanthin products fail third-party potency tests. CS-P1's subhead counts softgels and tests and names no enemy; the 64% sits lower, in the method box.
c1r14
4
2.3%
authority-stamp
2× p · 12 w
A woman's name with no title or credential, then an update stamp from late July 2026 and a reading time of four minutes.earned-authorityease
solution-aware
It follows the deck because the narrator of the headline needs a person at once. The date has to come before the list so the year in the title reads as fresh. The short reading time lowers the cost of starting.Bare: no role, no years, no outlet beside the name, the weakest byline a review can carry (an unverifiable reviewer; treat as a persona). The reading time is decoration: about 2,260 words here, while the store's shorter liver page claims nine minutes.keep-structure
CS-P1 already has this and does more: a named health editor with a role, years of experience, a headshot, a date, a read time and a press strip. Nothing to take. One lesson: keep the read time true to the word count.
5
2.9%
page-furniture
3× p · 5 w
Three short tab labels that jump to the benefits, to the dosing section and to the ranked list.ease
solution-aware
Under the byline and above the first section, it is the last thing seen before reading starts, so it previews the order of the page. Nothing needs to be believed. A list-seeker can reach card one without reading the method.One or two words each. The middle label uses a medical-sounding adjective before any evidence has been shown, which sets the register early.page-furniture
CS-P1 has a breadcrumb and no jump tabs. For a reader in their 60s and 70s a jump to the list helps if the tap targets are large. Optional. No claim.
6
3.4%
ranking-criteria
h3 · p · 49 w
Under a how-we-ranked heading the reviewer gives a funnel: more than fifty products looked at, fifteen tried in person, five ranked by one method built on three factors tied to whether an antioxidant gets into liver cells.earned-authorityspecificityreason-whyopen-loop
solution-aware
Method comes before the problem, which is unusual: a ranking is only as good as its judge, so the judge is qualified first. The reader must already want a list. Once the funnel is accepted the five survivors feel selected, and the three factors can be taught as science, not as a sales argument.Three numbers in two sentences do the persuading. Personal testing is never defined for a pill. The last clause quietly assumes the answer is an antioxidant before any reason has been given.rewrite-claim
CS-P1 has a how-we-ranked section, placed after a two-paragraph intro, and says it bought and read five. It has no funnel. Only true counts: five rivals, each read on its own product page on a dated day. The liver-cell clause becomes the cell membrane, which is what The Cell Shield protects.
m2b7
7
5.0%
ranking-criteria
4× p · 18 w
Three short labels name the factors: an amount that matches the studies, an active from a natural source, and a form the body can take up. A closing line promises the reasoning further down.specificityopen-loopcommitmentease
solution-aware
Straight after the funnel, so the method has a shape the eye can hold. The reader has to agree that three factors are enough, a small yes that every later card collects on. The same three return as chips on each card: this is where the scorecard is installed.Two or three words per label, parallel, easy to remember. The promise of detail is an honest open loop, paid within the next two screens.keep-structure
CS-P1 already has three scoring axes: the full 12 mg in one softgel, third-party lab tested, cost per day. Two of theirs match claims we hold (dose, natural source). Their third, delivery form, is not ours: it sits beside the banned carrier-oil claim. Keep our three.
b7c4c3
8
7.4%
image
img
A wide image with empty alt text sits between the method and the problem section.none
solution-aware
After the method and before the first long paragraphs, where a phone reader needs a break. Nothing in the argument depends on it.Empty alt text: the image says nothing to a screen reader and nothing in the extracted rows. Whatever it shows, it carries no labelled claim.keep-structure
CS-P1's hero is our own bag, captioned as the top pick, so our winner is known from the first screen; their page holds the name back until the survey. A lineup of all five products would keep the comparison frame. On our pages alt text is copy.
9
8.7%
failed-solutions
h3 · p · 45 w
A heading says most buyers see no result. Then the reader's own story in the second person: a doctor's flag or a diagnosis, a supplement bought in hope, and months later a blood test that had hardly changed.mirroringidentificationspecificity
solution-aware
It comes after the method, so the voice describing the disappointment is already the expert's. This audience has tried something, which the line depends on. Naming the flat lab result raises the question of why, and the next line answers it.The best writing on the page. Two sentences, a time marker and one concrete object, the number on the next lab report. No adjectives about feelings: the event carries the feeling.rewrite-claim
We make no liver or enzyme claim. The move transfers as the cabinet of vitamins that changed nothing; the nearest lab number the gate holds is kidney numbers. CS-P1's intro states the problem for the category (most vitamins never reach the cell) but does not replay the reader's own purchase in the second person. It lacks this.
r3r5r10p7
10
10.1%
villain-reveal
p · 25 w
A two-sentence pivot clears the idea and blames the shelf: antioxidant help for the liver is called sound, and the failure is put on products that carry too little or carry it in a form the organ cannot take up.vindicationhopecommon-enemyreason-why
solution-aware
Placed straight after the replayed disappointment, while the reader is asking why. Before it there is no reason to buy again. After it a new purchase is reasonable, provided it passes a test of amount and form, which is exactly what the standards then supply.Two sentences, the first very short. It opens on a stock setup phrase, then lands a plain claim. A directional quantifier keeps it honest.rewrite-claim
Same pivot, our content: antioxidants are the right idea, but ordinary ones guard only part of the cell and vitamins stay in the bloodstream, while astaxanthin spans the whole membrane. That is The Cell Shield. CS-P1's intro already makes this turn in its first paragraph. No liver.
p3r5m2r3
11
10.6%
ranking-criteria
p · 35 w
The reviewer sums up her inputs (the products, the research and a poll of 243 buyers she calls her own) and says three patterns kept dividing what works from what does not.earned-authorityspecificityopen-loop
solution-aware
After the pivot the reader needs to know what the test of amount and form is. This line says it has three parts, a loop closed in the next section. It also gives the sample size early, so the later percentages have a base.One long sentence that works. The possessive gives the game away: an outside reviewer has no customers of her own. From here the tester and the seller speak with one voice.keep-structure
CS-P1 has the bridge (it says it ranked on three things). Keep ours. Do not borrow a customer poll we have not run, and never let the editor's voice speak as the brand's.
r17
12
11.5%
ranking-criteria
h2 · h3 · h4 · p · 51 w
A heading announces three scientific standards. The first is dose: 12 mg a day is presented as what most trials used, and crowded formulas are said to give far less of each ingredient, the key one included.specificityborrowed-authorityanchoringcontrast
solution-aware
Dose goes first because it is a number, and a number is the easiest test for a reader to apply alone. The reader must believe trials exist. It marks down every multi-ingredient rival in advance, before a single card is shown.The number is the argument. The studies are neither named nor counted. The line points to an antioxidant that has not yet been named on the page, so a careful reader has to guess. It ends on an expander link.keep-structure
CS-P1 already leads its method with this test: is the full 12 mg in one softgel. The gate holds the dose claim and the study count. Their liver framing is dropped. Name the ingredient before pointing at it.
b7r14
13
13.4%
ranking-criteria
h3 · cta · p · 29 w
The second standard is source. Here the ingredient is named for the first time: natural astaxanthin from a named microalga, described as the form most research on oxidative stress in the liver has used.borrowed-authorityspecificitynew-mechanism
solution-aware
The ingredient arrives as a criterion, not as a product, so it carries the authority of the method. The reader has already accepted dose. After this line no milk thistle formula can score full marks, and the reader has not noticed the field being narrowed.A Latin name for weight, a cautious quantifier for cover. The label shouts its second half in capitals and mentions absorption, which belongs to the third standard: the labels were not proofread.rewrite-claim
We hold the source claim without the liver: natural microalgae astaxanthin is the form the positive human studies used, and synthetic astaxanthin comes from petrochemicals. CS-P1 has it as a method item and says candidly that all five rivals pass it, so it did not decide the order.
c4r16c5
14
14.5%
ranking-criteria
h3 · cta · p · 25 w
The third standard is delivery: because the ingredient dissolves in fat, an oil-filled softgel is said to get more of it into the blood and the liver than a dry powder does.reason-whycontrastspecificity
solution-aware
Last of the three because it depends on the other two: what it is, how much, then whether it arrives. It gives the pivot line its proof. With it, the three chips on every card are defined.One sentence with a cause in it. The adjective of importance is weaker than the two standards before it, and no figure says how much more is absorbed.rewrite-claim
Delivery-form superiority is not in our gate and sits beside the banned carrier-oil claim. What we hold: astaxanthin is fat-soluble, settles into cell membranes, and is taken with a meal that has some fat (an instruction, not a superiority claim). CS-P1's third axis is cost per day. Keep it.
r31r26r5
15
15.7%
proof-social
h2 · p · 31 w
A new section turns from lab criteria to buyers. Speaking in the plural, the page says it polled 243 verified customers of the winning brand who had taken the product for eight weeks or more.social-proofspecificitycandor
product-aware
It sits between the standards and the list, so results reach the reader before the winner's card does. The rubric must be accepted first or the numbers would read as advertising. It is also the first time the brand is named: the reveal happens inside evidence, not inside a pitch.Three qualifiers in one sentence. The heading says plainly whose customers they are, which is honest and also ends the pose of an outside reviewer.swap-proof
CS-P1 has no customer panel block. The gate allows one only from our own panel with signed-off numbers, and none exist yet. Until then the slot holds gate proof: 60+ clinical studies and the human trial results. Never publish invented percentages.
r17r14b1b3
16
16.7%
proof-social
img · h1 · p · 8 w
An illustration, then the first figure: 61% said they bought the product for a fatty liver in particular.social-proofidentificationspecificity
product-aware
First of four because it is about the buyer, not the result: it tells the reader they are in the right room. Outcome figures only matter once the respondents are seen to share the reader's problem.A big numeral and seven words. It is a reason-for-purchase figure given the visual weight of a result, which flatters it.rewrite-claim
No liver claim. If our panel runs, the first figure should also be a relevance figure: the share who bought for kidney numbers, eyes, joints or energy. CS-P1 lacks it.
r17i2
17
18.9%
proof-social
img · h1 · p · 12 w
The second figure, 72%, covers only the respondents who went back for blood work: that share said their liver enzyme readings had got better.social-proofhopespecificityloop-closed
product-aware
Second, once relevance is set. It answers the replayed disappointment directly: the number that would not move, moving. Everything after it on the page leans on this figure.The limit is inside the line, so the claim is a reported experience of a subset. Read fast it still sounds like seven in ten of everyone. The strongest and riskiest line in the block: a disease marker from a self-report.rewrite-claim
Drop liver enzymes. The nearest the gate holds is kidney numbers and better labs, and any figure must come from our own panel. CS-P1 handles labs in the FAQ and the timeline with a supplement-not-treatment line. It has no figure.
r11r10r17
18
21.1%
proof-social
img · h1 · p · 11 w
The third figure: 64% said a doctor or other clinician had responded well when their liver health came up.borrowed-authoritysocial-proofrelief
product-aware
After the lab figure the reader asks whether their own doctor would agree. This answers through other patients' doctors. It comes before the recommend figure because approval from authority outranks approval from peers.Vague on purpose: what the clinician approved of is left open. Ten words that imply a clinical endorsement and claim none.rewrite-claim
No liver. The gate holds a direct route to this: longevity doctor recommended, with a quote. A figure about what customers' doctors said needs our own panel. CS-P1's ranking has neither.
r19r17
19
23.4%
proof-social
img · h1 · p · 12 w
The fourth figure: 91% would tell a friend or relative to take it.consensussocial-proof
product-aware
Last because it is the biggest and the least specific: it sends the reader into the list with a general feeling of agreement. It needs the three figures before it to mean anything.The standard satisfaction figure, the grayest of the four. Its work is rhythm: the block ends high.swap-proof
Only from our own panel. CS-P1's sidebar carries a star rating and a customer count; the brand file notes two different counts in use, to be settled before either is repeated in a panel block.
r17
20
25.5%
disclosure
p · 51 w
A footnote restates the sample, then admits that only about a third of those polled went back for blood work, that the enzyme figure covers that third alone, and that nothing was checked against lab records.candor
product-aware
Directly under the four figures, where a skeptic looks for the catch. Finding the catch stated by the page itself disarms the skeptic. The ranking that follows inherits that credit.Fifty-one words of plain admission, including the unflattering subset share. The arithmetic: about 83 people had blood work and about 60 reported improvement. This is real color; few competitor pages print the denominator.keep-structure
Take the habit, not the numbers: when our panel exists, print the sample, the subset and the words self-reported under the figures. CS-P1 has a general results-vary line in the footer only.
r17
21
26.8%
ranking-criteria
h2 · p · 46 w
The list heading, then a paragraph: all five were held to the three standards, and the final order also used softer editorial factors, among them outside testing, label openness, value and buyer protections.reason-whyearned-authorityobjection-preempt
product-aware
Right before card one, because it must explain in advance why a rival that passes a standard can still rank lower. The reader has accepted three standards; this adds four more without defining or weighting any of them.It reads as thoroughness. In practice it is a licence: the grades that follow have no stated method (scores with no method). The four factors are the winner's trust points listed as neutral categories.keep-structure
CS-P1's method box lists five deciding items and scores three of them out of ten on every card, which is more checkable than a letter. Keep ours. Do not add guarantee length as a factor: ours is 30 days, shorter than some rivals'.
c3c1
22
28.6%
ranked-entry
h2 · h3 · 2× p · 3× fine · 20 w
Card one: the store's own astaxanthin softgel, its maker, the top grade, and three chips that restate the standards in liver words: the dose, the microalgae source, the oil base.loop-closedconsensusspecificity
product-aware
At about 29% of the page, after method, problem, standards and survey. The winner is first in the list, not last: shoppers read the top card and skim the rest. Every belief the card needs was installed above it.The chips are the criteria turned into badges, so the card grades itself on its own spec sheet. The grade has no reasoning behind it. Nothing in the header admits a weakness.keep-structure
CS-P1 has this card first, with three numeric score bars, and gives itself 7.5 on cost per day: a visible weakness that makes the two tens believable. Keep ours. The noun on the card is The Cell Shield; the organs named are the gate's, never the liver.
b7c3c4
23
31.0%
ranked-entry
h2 · 9× p · 126 w
Nine pros, each a label and a reason. The three standards come first and in order, then testing, a clean label, US manufacture, the 90-day guarantee, a doctors-and-customers line (142,000 buyers, 4.8 stars) and the $29.99 price.specificitysocial-proofrisk-reversalanchoringborrowed-authority
most-aware
Pros come before any prose so a skimmer gets the whole case in one screen. The order mirrors the page so far: the rubric, then the editorial factors, then the offer. Price arrives last, after eight reasons.Label, colon, reason: quick to scan. Two of the reasons end on the organ to keep the claim in view. The doctor line names no doctor, and the customer count and rating have no source (unverifiable).keep-structure
CS-P1 lists five pros under 'why readers are switching'. Ours rest on the gate: the full 12 mg, lab tested every batch, natural microalgae from closed bioreactors, one softgel a day, plus the 30-Day Money-Back Guarantee and free shipping from the locked offer. Trust badges need supplier facts first. No liver.
b7c3c4c5b9r18
24
34.9%
ranked-entry
h2 · 2× p · 20 w
Two cons only: it is sold nowhere but the brand's own site, and the multi-pack deals run out when a large sale is on.exclusivityscarcity
most-aware
Directly under nine pros the reader expects a catch and is given two that are not. It keeps the review form intact and plants the only scarcity on the page.Both are benefits in disguise, and a practised reader sees it. The second has no evidence: no stock count, no date. One real limit would have made the box believable.keep-structure
CS-P1 has three under 'worth knowing', and the first is a true limit: a daily supplement, not a treatment, keep your doctor and your labs. That is better candor than theirs. Official site only is in our gate. Drop the sell-out line.
r29
25
36.1%
ranked-entry
h3 · p · 23 w
The verdict opens in the first person: this was the single pick that asked for no trade-off, and it has everything the reviewer wants in an antioxidant for the liver.earned-authorityconsensus
most-aware
After two lists, prose: the skimmer has the facts, the reader now gets the voice. The personal line has to come before the spec, or the spec would read as a label.Two short sentences in a spoken register. It claims completeness and names no single deciding fact, so it leans entirely on the lists above.keep-structure
CS-P1 has the equivalent in its 'why it took the top spot' box, and states the deciding fact: the only product that passes all three tests at once. Ours is the stronger sentence. Already has.
b7c3
26
37.2%
mechanism
p · 50 w
One paragraph turns the spec into a reason: the full 12 mg of natural astaxanthin as an antioxidant the liver draws on against daily oxidative stress, an oil softgel that reaches the blood, and Hawaiian microalgae set against synthetic material.new-mechanismreason-whycontrast
most-aware
It is the only place the page says what the ingredient does for the organ, and it is kept inside the winner's card, where the reader asks why this one. It relies on the standards having made dose, source and form feel like science.The explanation is one clause long and stops at a general phrase about oxidative stress: the thinnest part of the argument. The sharp word is the one for what cheaper brands are made from.swap-mechanism
This is the slot for The Cell Shield: astaxanthin spans the whole cell membrane, where vitamin C and E guard only part of it, with about 6,000 times the power of vitamin C. CS-P1's card explains fat-solubility and the kidneys but never says in one sentence what The Cell Shield is. Add that. No liver.
m2m3p3r16c4
27
38.5%
differentiation
p · 31 w
A paragraph on openness: outside lab testing for purity and strength, a certified plant in the US, and a label with nothing added.borrowed-authorityspecificity
most-aware
Trust follows the working principle because a reader who now believes in the ingredient asks whether this maker really supplies it. It cashes the editorial factors promised above the list.A first-person clincher opens it and keeps the reviewer's voice alive. The facts are standard for the category; no report, lot or lab is named.swap-proof
Our strongest ground: every batch third-party lab tested with results verified, in a category where about 64% of unverified products fail potency tests, from microalgae grown in closed bioreactors with no heavy metals. CS-P1's card already says this. Already has.
c3c1c5
28
39.2%
proof-social
p · 43 w
The crowd paragraph: more than 142,000 buyers, a 4.8 rating and doctors' backing, then the routine (one softgel a day, thirty to a pouch) and ninety days to make up one's mind.social-proofborrowed-authorityeaserisk-reversal
most-aware
Numbers come after testing so they read as a result of quality, not a substitute for it. The routine and the window are attached so the reader moves from others' experience to their own first month.A list dressed as prose: six facts joined by commas. The count, the rating and the doctors have no source (unverifiable).swap-proof
Ours: 60+ clinical studies and the longevity-doctor line with a quote, then the routine: a single softgel daily, thirty per bag. CS-P1's card has the routine; it carries no doctor line and no customer figure inside the card (the sidebar has a rating and a count).
r14r19b9
29
40.0%
ranked-entry
h3 · p · 37 w
The bottom line names three ways other liver brands cut corners (low doses, hidden blends, plain capsules that pass through unused) and says who this one is built for.common-enemyidentificationcontrast
most-aware
It closes the winner's argument by pointing forward: the three faults are what the next four cards will be marked down for. The reader now reads the rivals looking for them.A triad that carries content, each item one standard turned into a fault. The clause about capsules passing through is the most spoken phrase on the page.keep-structure
CS-P1 has the equivalent above the card body. Our three faults are checkable on rivals' own pages: more than one softgel to reach 12 mg, a tested badge with no report behind it, a high cost per day. Already has.
b7c1c3
30
41.3%
offer
p · 34 w
One sentence restates the three standards as what the buyer receives, then gives the price of $29.99, a first-order discount of as much as 70%, and free shipping in the US.anchoringspecificityease
most-aware
Price appears in prose only after the verdict, the working principle and the trust paragraphs. Earlier it would have been judged against cheaper bottles; here it is judged against a rubric the cheaper bottles fail.A two-word connective turns the price into the sum of everything listed. A time word implies a deadline that is never stated, and the discount has no base price and no reason: the reader cannot work out what the percentage is taken from.swap-offer
Ours: $29.99 a bag, 3 bags $59.97, 5 bags $89.95 ($17.99 a bag), free shipping on every order, against a stated $59.99 anchor. CS-P1 keeps the price out of the card body on purpose and puts it in the block under the card. Keep that, and state the anchor so the saving can be checked.
x4
31
42.1%
guarantee
p · 40 w
No fillers, so each softgel counts; then the guarantee in its plainest form: ninety days, money back if the buyer does not feel better, framed as enough time to judge.risk-reversalreason-whyrelief
most-aware
Last line of the winner's argument, straight after the price, because the reader's remaining fear is the one the page opened with: another bottle that does nothing. A long window fits a result that shows on a blood test months later.The guarantee is tied to a feeling, not to a lab result, which is generous and also safe for the seller. A reason is given for the length, which makes it believable.swap-offer
Ours is the 30-Day Money-Back Guarantee, keep the bag (the brand's locked offer; it supersedes the 90-day wording in the gate's second round). CS-P1 states it in the card's pros, the block under the card, the FAQ and the close. Do not argue window length: some rivals offer longer.
32
43.0%
disclosure
fine · 25 w
Small print under the card: what is described comes from users' own accounts, is not typical, and the product's site has more.none
most-aware
Under the last persuasive line and above the link, where a regulator looks and a reader does not.Boilerplate. The same line type closes every card with only the basis changed; that variation is the clever part, and this first one is the plain one.page-furniture
CS-P1 carries the FDA supplement disclaimer and a results-vary line in the footer, and nothing under the cards. A one-line note under our card is cheap cover for the kidney and eye language.
33
43.7%
cta
p · 5 w
The first link out, at 44% of the page: a text link with an arrow that offers more information about the winner, not a purchase.commitmentease
most-aware
Immediately after the winner's card and before the rivals: the earliest point at which the reader has the whole case. A softer ask than buying keeps the page's neutral pose for the readers who go on to the rivals.Five words, no benefit, no price. It asks for a small yes. It is also the page's only ask for the next half of its length, which is thin.keep-structure
CS-P1 has a full block here: a headline with the guarantee, the per-bag price, a button and a security line. Ours is the harder ask. A plain review-style text link above that block is a cheap test. Button words name The Cell Shield.
34
44.6%
ranked-entry
2× h2 · 3× p · 3× fine · 49 w
Card two: LiverMD from 1MD Nutrition, graded B+, with the same three chips. A one-sentence summary grants truly absorbable ingredients, then lists three drawbacks: two capsules a day, amounts missing from the sales page, a higher full price.candorcontrastspecificity
most-aware
The rival cards let the reader finish comparison shopping without leaving the page. Second place must look strong or the ranking looks rigged. The summary sits above the pros, so a skimmer takes away the markdown even if they read nothing else in the card.Concession first, then the turn, then three short faults. The grade drops a full letter from the winner with no arithmetic shown (scores with no method).keep-structure
CS-P1's rival cards open the same way: a tagline that grants the strength and names the catch, plus three numeric bars. Our rivals are astaxanthin softgels at the 12 mg serving, never liver formulas. Already has.
b7b9
35
47.5%
ranked-entry
h2 · 5× p · 46 w
Five pros, generously worded: formulated by a specialist doctor, an absorbable milk thistle, a complete vitamin E, the same 90-day guarantee as the winner, a certified US plant.candorborrowed-authority
most-aware
Pros come before cons on every card: the reader sees fairness before judgment. Naming the rival's branded ingredients shows the reviewer read the label.The technical names are explained in the same breath, so they read as expertise, not jargon. Five pros is the most any rival gets; the count falls with rank.keep-structure
CS-P1 gives each rival three true pros read from its own product page on a dated day. Same move, already there. Keep the accuracy rule: no invented defects, no praise we cannot verify.
36
49.9%
ranked-entry
h2 · 4× p · 51 w
Four cons, each a checkable fact: twice the daily capsules of the winner, $49 or more a bottle, exact amounts printed on the label but not the sales page, and the best price tied to a recurring order.contrastspecificityloss-aversion
most-aware
After the praise, so each fault reads as a fair observation. Two of the four name the winner in the comparison, which keeps the top card in mind while the reader is inside another product's card.Every con can be verified on the rival's site. The third is the weakest: the amounts are on the label, so the fault is only where they are printed. The subscription point is the one that stings.keep-structure
CS-P1 marks rivals down the same way: softgels per day to reach 12 mg, no batch report to open, cost per day. It rarely names CellScience inside a rival's card. Naming ours in one con per card is the piece to add. One softgel a day and a one-time purchase are ours.
b9b7c3
37
52.1%
ranked-entry
h3 · 2× p · 84 w
The verdict praises first (a specialist's formula pairing two ingredients made for better uptake, placing it in front of nearly every rival), then lists where it trails the winner: capsule count, amounts missing from the sales page, price.candorcontrast
most-aware
Prose follows lists on every card. The second paragraph opens by naming the winner, so the comparison is explicit at the moment the reader might be tempted by the runner-up.The second paragraph is one long sentence of three clauses, each a repeat of a con above. Nothing new is said; the repetition is the method.keep-structure
CS-P1's rival cards carry two paragraphs in this order: what it gets right, then the catch. Already has. Ours stay on facts read from the rival's own page.
38
54.6%
ranked-entry
h3 · p · 47 w
The bottom line calls it the best of the four alternatives for someone who wants a doctor's formula, then repeats the three costs of choosing it.candoridentificationloss-aversion
most-aware
Ending each card on who the product suits keeps the reviewer's neutral pose to the last line. For the runner-up it matters most: a fair second place makes first place believable.It closes on a spoken warning formula, friend to friend. It is the third pass over the same three faults inside one card, and by now it drags.keep-structure
CS-P1 ends each rival card with a price tier and a short verdict tag. It has no sentence on who should still buy the rival. Worth adding, one line per card: it costs nothing and reads as fairness.
39
56.0%
disclosure
fine · 23 w
Small print: this card's results rest on a customer survey the rival published about itself.candorcontrast
most-aware
Under the rival's card, the slot that protected the winner now grades the rival's proof. It starts a ladder that gets weaker card by card.One line, legal in tone, persuasive in effect. The winner's own proof is also a self-run survey, which the line does not mention.keep-structure
CS-P1 has no per-card source line. Ours could state where each rival's facts were read and on what date, which is true and protects us. The same standard must apply to our own card.
40
57.0%
ranked-entry
2× h2 · h3 · 2× p · 3× fine · 41 w
Card three: Hepagard from Nutreance, graded B-. The summary grants a simple, open formula, then says its own study measured bloating and energy instead of liver enzymes and that its extracts are not standardised.contrastspecificitycandor
most-aware
Mid-list is where a reader starts to skim, so the most interesting markdown on the page goes into the summary line. It depends on the survey block: the reader has been taught that enzymes are the measure that counts.Real color: a fault found by reading the rival's study, not its label. The sentence turns on a single conjunction.keep-structure
The move (their proof answers a different question) transfers only where it is true and read on the rival's own page. CS-P1's closest is the tested badge with no batch report behind it. Our own trial results must be named with the same care.
c1c3b1b3
41
59.8%
ranked-entry
h2 · 4× p · 35 w
Four pros: five ingredients that are easy to follow, a real placebo-controlled study of 50 people over 30 days, a price of $39.95 a month, and a clean allergen profile.candorspecificity
most-aware
The study is praised here so that the con which follows cannot be called unfair. Pros drop from five to four as rank falls.Numbers on three of the four lines. The study's size and length are given exactly, which makes the later criticism of what it measured read as informed.keep-structure
CS-P1 already credits rivals with their true strengths: the lowest cost per day, the full dose in one softgel. Already has.
42
61.6%
ranked-entry
h2 · 5× p · 60 w
Five cons: the study tracked digestive symptoms and energy, not enzyme levels; the plant extracts are not standardised; the capsule is plain; the guarantee is the shortest, 30 days; and distribution is narrow, mostly the brand's site plus Amazon.contrastspecificityloss-aversion
most-aware
Cons grow to five as pros shrink to four. The third repeats the delivery standard from the top of the page, so the reader sees the rubric at work.The last con is odd: narrow distribution is a fault here, and the winner's own single-site sale was listed as a mild con too. The guarantee point only works because the winner's window is long.rewrite-claim
Do not take the short-guarantee markdown: our window is 30 days, the length they criticise. CS-P1 rightly scores no guarantee axis and lists rivals' longer windows as their pros. Our markdowns stay on dose per softgel, lab proof and cost per day.
b7c3
43
64.4%
ranked-entry
h3 · 2× p · 75 w
The verdict names the five ingredients and credits the study, then states the catch: thirty days of symptom tracking, no enzyme data, no standardised extracts. It ends by setting this formula against one built on a single dosed active.contrastreason-whycandor
most-aware
The explanation comes in prose after the list has made the point in short. The final clause names the winner, as the runner-up's verdict did. The two lower cards never name it: they are judged against the rubric alone.The pivot is a short plain sentence about what was measured. It calls the rival honest and affordable in the same breath as marking it down, which keeps the tone fair.keep-structure
Already in CS-P1's two-paragraph rival shape. If we ever criticise a rival's evidence, the same paragraph must say what ours measured: wrinkle depth and skin elasticity, retinal blood flow and eye strain.
b1b3
44
66.8%
ranked-entry
h3 · p · 39 w
The bottom line calls it a reasonable choice for a simple, low-cost formula with a study, and warns that the study is about digestion, not the enzyme change people search for.candorloss-aversion
most-aware
Card end: a recommendation with a caveat. The caveat speaks to the reader's search intent, the enzyme number from the top of the page.The same warning formula as the card above: the template shows on the second sighting.keep-structure
CS-P1's verdict tags do this in five or six words per rival. Keep ours short and vary the wording card to card.
45
68.0%
disclosure
fine · 21 w
Small print: results here rest on a placebo-controlled study that the rival's maker paid for.candorcontrast
most-aware
Second rung of the evidence ladder under the cards. It comes after the praise of the study, so the word about sponsorship lands last.One adjective does the work. It is a fair point, and still a better class of evidence than the winner's own, which the page leaves unsaid.keep-structure
Same as the note on card two: a dated source line per rival. CS-P1 lacks it. No claim.
46
69.0%
ranked-entry
h2 · h3 · 2× p · 3× fine · 35 w
Card four: Liver Nurture from American Vitality. The grade letter is missing from the rows (neighbours: B- and C-). The summary says the product trades on image and a lengthy return window, not on stated amounts or research.contrastcandor
most-aware
Fourth, where the reader expects weak products. The summary drops the concession-first pattern of the cards above: by this rank the reviewer no longer needs to look generous.One sentence with no turn in it. A single verb of dependence carries the judgment.keep-structure
CS-P1's fourth and fifth cards still open on a concession, so ours stay fair to the last card. Keep ours.
47
71.4%
ranked-entry
h2 · 3× p · 29 w
Three pros: a 365-day return window, the longest on the list; nine familiar herbs; two free e-books with an order.candor
most-aware
The longest guarantee on the page belongs to a rival and the page admits it, which protects the credibility of the guarantee comparisons made elsewhere.Short lines, and a different separator from the other cards (a dot, not a colon): this card was built at a different time or by a different hand.keep-structure
CS-P1 credits Rovina with a 90-day guarantee, longer than ours: same candor, already there. The gate also allows three free digital guides on every order, which CS-P1 does not mention.
r20
48
72.8%
ranked-entry
h2 · 3× p · 41 w
Three cons: no amounts shown for any of the nine ingredients; the only results cited come from a 14-person challenge run by the company; and checkout can enrol the buyer in a separate $77-a-month membership unless they opt out.loss-aversionspecificitycontrast
most-aware
The checkout warning is kept for the lower half of the list, where it paints the whole field as risky and the winner as the safe place to buy. It depends on the reader having seen the winner's plain price first.The third con is the most alarming line on the page and the most specific: a price, a period, a default. These cons are full sentences, unlike the labelled fragments in the other cards.keep-structure
Buyer-protection markdowns transfer when true and read on the rival's own checkout. Ours to state plainly: one-time purchase, no subscription, price shown before payment. CS-P1 says this in the fine print under its button.
49
74.7%
ranked-entry
h3 · p · 28 w
The verdict is one paragraph: the brand trades on its year-long guarantee and a list of well-known herbs, and the reviewer grants that this looks good at first sight.candoropen-loop
most-aware
It ends on a qualifier that hands the reader to the bottom line for the other half of the thought.Short. The last sentence is five words and sets up a turn: the only card that uses verdict and bottom line as the two halves of one thought.keep-structure
Nothing new for CS-P1. The two-part shape (looks good, then what is missing) is already our rival paragraph order.
50
75.9%
ranked-entry
h3 · p · 47 w
The bottom line names what is absent (amounts per ingredient, research on the finished product), repeats the small company-run challenge, grants that the long guarantee reassures, and tells the reader to read the checkout terms.loss-aversioncandorobjection-preempt
most-aware
The warning comes last in the card so it is the memory the reader keeps. It also neutralises the rival's best feature: a guarantee is worth less if the checkout cannot be trusted.A run-on: two sentences joined by a comma. The advice about the checkout is practical and sounds like a friend.keep-structure
Ours: say what our checkout does. CS-P1 already prints one-time purchase and free shipping beside the button.
51
77.4%
disclosure
fine · 23 w
Small print: results rest on an informal customer challenge run by the brand.candorcontrast
most-aware
Same slot, weaker basis than the card above it: a survey, then a sponsored study, then a casual challenge.One loaded adjective carries the line.keep-structure
As above: a dated source line per rival, with no adjective we cannot support. CS-P1 lacks it.
52
78.4%
ranked-entry
h2 · h3 · 3× p · 3× fine · 46 w
Card five: Vital Liver Support from Vital Nutritive, graded C-. The summary names the gap: with more than twenty ingredients most sit far below research amounts, and part of the formula is hidden in a blend.contrastspecificityloop-closed
most-aware
Last place goes to the product that is the opposite of the winner: many ingredients, small amounts. It closes the loop opened in the dose standard, which warned about exactly this.It opens by naming the gap outright. The summary carries no concession at all.keep-structure
For us the opposite of the winner is a softgel that needs several a day to reach 12 mg, or a 4 to 6 mg product. The gate holds that comparison. CS-P1 puts that rival second, not last; the order could be tested.
b7r23
53
81.1%
ranked-entry
h2 · 3× p · 33 w
Three pros: a wide range of plant ingredients, a 60-day guarantee, and several pack sizes with a lower price per bottle on the larger ones.candor
most-aware
Even the bottom product gets three pros, so the template never breaks and the list never looks like an attack.The praise is faint on purpose: breadth is the very thing the card then criticises.keep-structure
Already in CS-P1: every rival has three pros.
54
82.6%
ranked-entry
h2 · 7× p · 105 w
Seven cons, the page's longest list: tiny amounts of most ingredients, a 571 mg blend with no breakdown, a plain capsule, no clinical study, a price that moves with pack size, reviews mentioning serious illness, advice to double the serving.contrastspecificityloss-aversion
most-aware
The cons count peaks on the last card: the reader leaves the list with the longest fault list fresh and meets the winner's recap straight after.Exact figures make the faults feel measured. Two of the seven (price varies by pack, reviews mention illness) are a stretch. The point about testimonials is a compliance criticism most readers will not follow.keep-structure
CS-P1 holds each rival to three cons, all read from the rival's page. Keep three. A longer list on the last card reads as piling on to an older reader.
55
86.4%
ranked-entry
h3 · 2× p · 84 w
The verdict grants that twenty-plus ingredients look impressive, then checks the amounts: tiny individual doses and a hidden six-part blend. A second paragraph calls the formula spread thin, unstudied, and reliant on testimonials that mention serious liver illness.contrastreason-whyspecificity
most-aware
It restates the first standard as a story of reading a label: looks good, then the numbers. By the last card the reader can predict the judgment, which means the rubric has been learned.The picture is of ingredients watered down. The turn is the spoken kind: the fault appears when the amounts are checked.keep-structure
Our version of reading the amounts is reading the directions: how many softgels make 12 mg. CS-P1 uses it on the Rovina card. Already has.
b7
56
88.8%
ranked-entry
h3 · p · 40 w
The bottom line opens on a maxim: a long ingredient list and a well-dosed one are different things. Then a fair line for readers who prefer breadth, ending on the point that nothing in it matches research amounts.contrastcandor
most-aware
Last words of the last rival card: the takeaway is phrased as a general rule, so it applies to every product the reader may meet after leaving the page.It opens on a quotable contrast line, the kind that reads as written. The two-noun opposition that follows is tidy but abstract.keep-structure
CS-P1's final rival card ends on a verdict tag. A one-sentence rule of thumb after the last rival would help: count the softgels it takes to reach 12 mg, then ask to see the lab result. CS-P1 lacks it.
b7c3
57
89.9%
disclosure
fine · 19 w
Small print: results rest on customer testimonials.candorcontrast
most-aware
The ladder ends at its weakest basis on the lowest-ranked card: own survey, sponsored study, informal challenge, testimonials.Plain. The order of the four lines is the argument, not any one of them.keep-structure
As above: a dated source line per rival. CS-P1 lacks it.
58
90.7%
close
h2 · p · 24 w
A heading declares the winner by name. One sentence repeats the headline's claim of fifteen products tried in person and says a single brand had the best all-round offer.loop-closedearned-authorityconsensus
most-aware
Directly after the last rival's fault list, where the contrast is sharpest. It returns to the opening claim so the page feels like a finished test, from fifteen down to one.The heading gives the name and the sentence then withholds it: a leftover from a version in which the heading did not name it.keep-structure
CS-P1 goes from the fifth card to the timeline, the comments and the FAQ, and restates the winner only in the final block. It lacks a recap straight after the last rival. Add a short one there.
b7c3
59
91.8%
differentiation
4× p · 30 w
A short label, then the three standards as a checklist, each with a bracketed reason: the dose as the number that counts, the microalgae source, the once-daily oil softgel the body takes up.loop-closedspecificityease
most-aware
The sixth time the three standards appear (labels, explanations, chips, pros, the price sentence, now the recap), here as the reason for the result. A reader who skipped to the end gets the whole method in three lines.The brackets talk like an aside to a friend. The second bracket credits the natural source with better absorption, which mixes two standards.keep-structure
CS-P1's proof bar has three items (30 days, 12 mg in one softgel, lab tested every batch) and sits after the timeline. Same content, later position, no reasons attached. A three-tick version under the last rival is the change.
b7c3c4b9
60
92.9%
differentiation
3× p · 17 w
A second label for extras, then two of them: outside lab testing the buyer is told can be checked, and a clean label free of GMO, gluten, dairy and fillers.borrowed-authorityspecificity
most-aware
These are the editorial factors from just above the list, now ticked for the winner. They follow the three standards, so the order of proof is: passes the science, then passes the trust test.The promise that the buyer can check the testing is strong, and the page backs it with no link and no report.swap-proof
Our edge is strongest here: every batch third-party lab tested with results verified, in a category where about 64% of unverified products fail potency tests. CS-P1 says it in the card and the proof bar. Allergen badges need supplier facts before use.
c3c1r18c5
61
93.8%
offer
2× p · 17 w
Two more ticks: the 90-day money-back window, described as enough time to judge, and the $29.99 price with its discount of as much as 70%, labelled the best value.risk-reversalanchoring
most-aware
The offer comes last in the checklist, as it did in the winner's pros: reasons first, money after.The price appears for the third time and the guarantee for the fourth. The value claim is made against rivals whose prices were given as ranges.swap-offer
Ours: 30-Day Money-Back Guarantee, keep the bag; $29.99 a bag, $59.97 for three, $89.95 for five; free shipping. CS-P1 states these in its blocks. The lowest-price-ever line is approved for the five-bag tier only.
x4
62
94.3%
close
p · 4 w
A four-word sentence on its own line names the brand, as if revealing it.loop-closed
most-aware
After the ticks, the name. It would work if the heading above had not already given it.Template residue: a reveal with nothing left to reveal. Harmless, slightly silly.drop
Nothing to take. CS-P1 names its winner from the hero image down; a late reveal line has no job on our page.
63
94.7%
differentiation
h2 · 3× p · 68 w
Why the winner differs: every product claims to help the liver, few meet the same three-part bar, and this one adds the best all-round offer (outside testing, an uncluttered label, a genuine refund promise, its customers' own results).contrastconsensussocial-proof
most-aware
It sits between the checklist and the link to catch the reader who wants a sentence, not ticks. It adds nothing new; its job is to hold the reader at the decision point for one more screen.Three short paragraphs, the middle one a single line. Gray: any brand could paste it. The same two-word phrase for the offer is used twice within ten lines.keep-structure
CS-P1's final block says it in two sentences: the full 12 mg in one softgel, third-party lab tested, ships free from the maker. Tighter than theirs. Keep ours. No liver.
b7c3
64
96.9%
cta
p · 5 w
The second link out, at 97%: the same soft text link as under the winner's card.commitmentease
most-aware
It follows the recap sections, more than half a page after the first link. Readers who finished the rivals get the same low-pressure ask.Identical wording to the first. No benefit named.keep-structure
CS-P1 has a sticky bar and a sidebar button across the page, so our reader is never half a page from a link. Already stronger.
65
97.3%
close
h2 · p · 28 w
A last two-sentence verdict: passing the science only gets a product considered; testing, openness, a genuine refund promise and what customers reported are the things that put this one first.contrastloop-closed
most-aware
Final prose before the final link. It echoes the two-layer scoring set up before the list, so the page ends where its method began.A polished closing pair that tries to be quotable. It is the third recap in a row: the close is where the page sags.keep-structure
CS-P1 closes on a question headline and one short paragraph, then the button. One close is enough for our reader. Already has.
66
98.4%
cta
p · 7 w
The last link, at 98%, is the only hard ask: try the product for ninety days at no risk.risk-reversalcommitment
most-aware
The page holds its strongest ask for the end, after two soft ones. The guarantee is the promise on the button because the reader's last fear is a repeat of the wasted purchase.Seven words: verb, brand, risk, window. The best-written link on the page.swap-offer
CS-P1's button already offers a risk-free trial of The Cell Shield, with the 30-day money-back line beside it. Keep ours: the noun in the button, the window beside it.
67
99.7%
page-furniture
fine · 8 w
A copyright line gives the publisher's name as a health news title that appears nowhere else on the page.none
most-aware
Footer. It is the only place the reader learns the review claims to come from a publication, while the address bar shows the seller's store.Eight words. No links, policy or contact in the extracted rows. The publication name and the store domain do not match, which a careful reader will notice.page-furniture
CS-P1 has a full masthead and footer under its own review title, with an ad disclosure. Already has more.