Skip to content

Aetheria-Labs1/storefront-skills / skills/ab-test

ab-test

Verified

Analyze an editable Lexsis landing-page URL, plan one controlled change, build verified challenger pages, and create or evaluate a draft A/B test. Use for URL-first storefront experimentation, not ordinary page editing.

FileKindSize
SKILL.mdskill md2.5 KB
agents/openai.yamlconfig471 B
references/ab-testing.mdmarkdown6.3 KB
references/ad-to-page.mdmarkdown1.3 KB
references/animation-system.mdmarkdown11.1 KB
references/anti-patterns/copy-anti-patterns.mdmarkdown12.2 KB
references/anti-patterns/cro-anti-patterns.mdmarkdown17.8 KB
references/anti-patterns/dark-patterns.mdmarkdown23.9 KB
references/anti-patterns/design-anti-patterns.mdmarkdown9.3 KB
references/anti-patterns/mobile-anti-patterns.mdmarkdown9.7 KB
references/asset-prep.mdmarkdown6.9 KB
references/assets/asset-sourcing-sequence.mdmarkdown12.2 KB
references/assets/generation-policy.mdmarkdown25.9 KB
references/assets/image-jobs-by-page-type.mdmarkdown31.8 KB
references/assets/imagery-by-vertical.mdmarkdown24.8 KB
references/assets/slot-spec.mdmarkdown20.2 KB
references/assets/video-rules.mdmarkdown17.6 KB
references/authoring/css-and-styling.mdmarkdown4.0 KB
references/authoring/source-authoring.mdmarkdown9.0 KB
references/blob-shapes.mdmarkdown1018 B
references/cart-composition.mdmarkdown7.2 KB
references/cart-profile-management.mdmarkdown17.1 KB
references/consumer-behavior-cro.mdmarkdown12.7 KB
references/copy/copy-frameworks.mdmarkdown20.5 KB
references/copy/headline-and-cta-rules.mdmarkdown12.9 KB
references/copy/message-match.mdmarkdown9.2 KB
references/copy/voice-of-customer-mining.mdmarkdown10.9 KB
references/design-assets.mdmarkdown366 B
references/design-enrichment.mdmarkdown334 B
references/design-rules.mdmarkdown15.5 KB
references/generate-collection.mdmarkdown555 B
references/generation-protocol.mdmarkdown3.2 KB
references/islands/collection-grid/index.mdmarkdown5.4 KB
references/islands/collection-swipe/index.mdmarkdown4.1 KB
references/lexsis-mcp-contract.mdmarkdown14.2 KB
references/merchant-templates.mdmarkdown2.6 KB
references/module-authoring.mdmarkdown9.0 KB
references/offers/campaign-calendar.mdmarkdown21.4 KB
references/offers/funnel-stages.mdmarkdown17.6 KB
references/offers/offer-ledger.mdmarkdown5.1 KB
references/offers/offer-strategy.mdmarkdown9.6 KB
references/offers/offer-types.mdmarkdown36.6 KB
references/offers/price-presentation.mdmarkdown18.0 KB
references/offers/urgency-scarcity.mdmarkdown17.5 KB
references/page-editing.mdmarkdown9.7 KB
references/page-files.mdmarkdown2.7 KB
references/page-types/_checklist-format.mdmarkdown11.1 KB
references/page-types/_index.mdmarkdown21.9 KB
references/page-types/collection-landing.mdmarkdown21.9 KB
references/product-grid.mdmarkdown2.0 KB
references/proof/before-after-and-claims.mdmarkdown19.2 KB
references/proof/numbers-and-counts.mdmarkdown16.0 KB
references/proof/press-and-media-mentions.mdmarkdown12.5 KB
references/proof/proof-ledger.mdmarkdown8.3 KB
references/proof/reviews-sourcing.mdmarkdown24.8 KB
references/proof/trust-badges-certifications.mdmarkdown21.8 KB
references/proof/ugc-rights-and-display.mdmarkdown16.0 KB
references/publishing.mdmarkdown2.4 KB
references/qa-recipe.mdmarkdown4.3 KB
references/quiz-authoring.mdmarkdown14.5 KB
references/resource-templates.mdmarkdown7.7 KB
references/scroll-patterns.mdmarkdown1.2 KB
references/source-artifact-workflow.mdmarkdown3.7 KB
references/styling-hooks.mdmarkdown7.8 KB
references/traffic-source-google.mdmarkdown1.6 KB
references/traffic-source-meta.mdmarkdown1.5 KB
references/traffic-source-tiktok.mdmarkdown1.5 KB
references/vertical-beauty.mdmarkdown1.8 KB
references/vertical-fashion.mdmarkdown1.7 KB
references/vertical-food.mdmarkdown1.6 KB
references/vertical-home.mdmarkdown1.7 KB
references/vertical-luxury.mdmarkdown1.6 KB
references/vertical-supplements.mdmarkdown1.7 KB
references/workflow-intent.mdmarkdown2.0 KB
references/workflows/_how-to-read.mdmarkdown2.7 KB
references/workflows/copy-workflow.mdmarkdown1.5 KB
references/workflows/island-selection-workflow.mdmarkdown4.8 KB
references/workflows/section-asset-workflow.mdmarkdown5.8 KB

references/anti-patterns/cro-anti-patterns.md

View on GitHub
# CRO anti-patterns

Evidence table of page patterns that measurably lose conversions or trust.
`/design-page` reads it before Compose and during the hosted review;
`/optimize` uses it as the first diagnosis pass. Deceptive practices live in
`references/anti-patterns/dark-patterns.md` (DP ids); mobile-specific
failures in `references/anti-patterns/mobile-anti-patterns.md` (MA ids);
copy in `references/anti-patterns/copy-anti-patterns.md` (CP ids). House
rules N1 to N14 and A1 to A12 in `references/design-rules.md` override this
file where they overlap.

Severity: BLOCK, FAIL, WARN. Tag: LAW, RESEARCH, OPERATOR, HEURISTIC.
Vendor benchmark data (popup tools, sticky-CTA vendors) describes those
vendors' customers; rows built on it are tagged RESEARCH (vendor).

Browser checks run in
the hosted draft at 390 x 844 and 1280 x 800.

## 1. Attention and layout

| Id | Anti-pattern | Evidence | Rule | Sev | Tag |
| --- | --- | --- | --- | --- | --- |
| CA1 | Hero carousel or auto-rotating slider | Notre Dame (Runyon), 3.76M visits: 1.07 percent of visitors clicked any slide; 89.1 percent of those clicks were slide 1. NN/g: users read moving panels as ads; do not auto-forward on mobile. Baymard recommends against auto-rotation on all touch devices. https://erikrunyon.com/2013/07/carousel-interaction-stats/ ; https://www.nngroup.com/articles/auto-forwarding/ ; https://baymard.com/blog/mobile-ecommerce-search-and-navigation | One static hero with one message. Stack extra messages as sections. Manual carousels only for product galleries, with arrows and dots. | FAIL | RESEARCH |
| CA2 | Autoplay video with sound | NN/g 2004: 79 percent rate auto-playing sound negatively; 2017 survey (n=452): autoplay video ads among the most hated. WCAG 1.4.2 (A): auto audio over 3 s needs a stop or volume control. https://www.nngroup.com/articles/most-hated-advertising-techniques/ ; https://www.w3.org/TR/WCAG22/#audio-control | Autoplay only muted, `playsinline`, poster, captions, visible unmute and pause. See `references/assets/video-rules.md`. | FAIL | RESEARCH, LAW |
| CA3 | Entry popup inside 10 s | NN/g: modals are the most-hated format on desktop and mobile. Google: intrusive interstitials on mobile entry from search may rank lower. Wisepops (vendor, 1B displays): immediate popups 4.16 percent vs 6.45 percent at 11 to 15 s; Omnisend (vendor): 0 to 1 s 1.9 percent vs 6 to 10 s 2.4 percent. https://www.nngroup.com/articles/most-hated-advertising-techniques/ ; https://developers.google.com/search/docs/appearance/avoid-intrusive-interstitials ; https://wisepops.com/blog/popup-stats ; https://www.omnisend.com/blog/email-popup-statistics/ | No marketing popup before 10 s and 30 percent scroll. Never while the hero is in view on paid-traffic types. Once per session, capped 7 days (DP14). | FAIL | RESEARCH (vendor), OPERATOR |
| CA4 | Competing CTAs | Unbounce attention ratio: homepages run about 40 links per CTA; a focused page is 1:1. Baymard: competing buttons distract from add-to-cart. https://unbounce.com/ppc/poor-message-match/ ; https://baymard.com/learn/ecommerce-ux-best-practices | One primary action per page on single-goal types. Secondary actions are text links. Never two equal-weight buttons side by side in the hero (see DA5). | FAIL | RESEARCH |
| CA5 | Site navigation on a paid landing page | VWO/Yuppiechef: removing nav doubled registry sign-ups (3 to 6 percent). Unbounce case: nav removed, 3.12 to 13.64 percent (small sample). https://vwo.com/blog/a-b-testing-case-study-navigation-menu/ ; https://unbounce.com/a-b-testing/how-a-single-a-b-test-increased-conversions/ | Types with `nav: none` render the logo only, not a link; slim legal footer. Enforced by type review T5. | FAIL | RESEARCH |
| CA6 | Generic stock imagery, AI-rendered people | NN/g eyetracking: users ignore decorative feel-good stock photos and scrutinise product photos and real people. https://www.nngroup.com/articles/photos-as-web-content/ | Hero shows the product or a real customer using it. People shown as customers are never generated (`references/assets/generation-policy.md` NEVER list). | FAIL | RESEARCH |
| CA7 | Wall of text | NN/g: users read at most 28 percent of words on an average visit; F-pattern scanning dominates unformatted text. https://www.nngroup.com/articles/how-little-do-users-read/ ; https://www.nngroup.com/articles/f-shaped-pattern-reading-web-content/ | Paragraph at most 45 words (3 lines at 375). Every section headline carries the message alone. Lists for 3 or more parallel items. | FAIL | RESEARCH |
| CA8 | Primary CTA below the first screen (single-goal types) | Baymard PDP research and design-rules A8: price and add-to-cart within the first viewport on desktop and 1.5 viewports at 390. Sticky-CTA evidence is mixed: GrowthRock +5.2 percent orders with a sticky drawer, no lift from scroll-to; CleanCommit +8.7 percent ATC but -7.7 percent CR in one test; RevenueFlows (47 stores): average +0.31 percent, only meaningful on pages over 1,200 words. https://growthrock.co/sticky-add-to-cart-button-example/ ; https://cleancommit.io/ab-tests/sticky-mobile-add-to-cart-button/ ; https://revenueflows.ai/blog/shopify-sticky-add-to-cart-button-myth | Follow the type checklist `cta.first_after_section` and `cta.sticky`. Where sticky is `optional`, add it only on pages longer than about two screens, showing price plus button, and treat it as a test. Never stack with other fixed bars (MA5). | WARN | RESEARCH (mixed) |
| CA9 | Dead-end page | The type checklists make `closing-cta` mandatory on selling types; conversion-psychology.md lists dead ends as a conversion killer. | The last content section before the footer is offer plus CTA plus guarantee or returns line. No trailing "related reads" after it on selling types. | FAIL | OPERATOR |
| CA10 | Essentials hidden in accordions | NN/g: accordions reduce discoverability; avoid hiding crucial information in collapsed panels. https://www.nngroup.com/articles/accordions-on-desktop/ | Price, shipping, guarantee, ingredients or materials, and the top three objections are visible. FAQ accordions hold the long tail only. | FAIL | RESEARCH |
| CA11 | Hamburger-only navigation on full-nav types | NN/g (n=179): hidden nav used 27 percent vs 48 to 50 percent visible on desktop; tasks 39 percent slower. https://www.nngroup.com/articles/hamburger-menus/ | On `nav: full` types, expose 3 to 5 top links at 1280; hamburger holds the rest. | WARN | RESEARCH |
| CA12 | Infinite scroll on a landing page | EU Digital Fairness Act fitness check lists infinite scroll and autoplay as addictive design likely to be off by default. https://www.dentons.com/en/insights/articles/2026/june/9/the-digital-fairness-act-dark-patterns-addictive-designs-and-influencer-marketing | Landing pages are finite. Product grids on `collection-landing` and `sale-clearance-flash` paginate or "Load more". | WARN | LAW (pending), OPERATOR |

## 2. Price and commerce

| Id | Anti-pattern | Evidence | Rule | Sev | Tag |
| --- | --- | --- | --- | --- | --- |
| CA13 | Hidden price ("See price", "Contact us") on selling types | Baymard: 12 to 14 percent abandon because total cost was unclear. Drip pricing is DP12. https://baymard.com/lists/cart-abandonment-rate | Price visible where the checklist says `price_above_fold: required` and beside every CTA on selling types. No gates. | BLOCK | LAW, RESEARCH |
| CA14 | Hidden shipping cost or delivery time | Baymard: 64 percent of shoppers look for shipping cost on the product page; 43 percent of sites omit it; "delivery too slow" is the second abandonment reason at 20 to 21 percent. https://baymard.com/blog/show-shipping-costs-on-product-pages ; https://baymard.com/lists/cart-abandonment-rate | Shipping cost or threshold plus a delivery window within one viewport of the primary CTA, from the offer ledger. Cutoff timers only when real (DP1). | FAIL | RESEARCH |
| CA15 | Forced account creation | Baymard: 18 to 19 percent abandonment reason; 60 percent of mobile test subjects overlooked guest checkout when not primary. https://baymard.com/blog/mobile-ecommerce-checkout-forms | Guest checkout primary; account offered after purchase. DP7. | BLOCK | LAW, RESEARCH |
| CA16 | Returns policy absent near the CTA | Baymard: 13 to 15 percent abandon over an unsatisfactory returns policy; 44 percent of sites do not link the policy from the product page. https://baymard.com/blog/current-state-ecommerce-product-page-ux | One-line returns or guarantee statement under the CTA, linked to the policy URL in the offer ledger. | FAIL | RESEARCH |
| CA17 | Free-shipping bar with wrong or negative math | Baymard: a free-shipping offer only in a site-wide banner is missed through banner blindness (32 percent of sites). Threshold bars that read "Add $-4 more" or exceed 100 percent break trust. https://baymard.com/blog/show-shipping-costs-on-product-pages | The threshold bar reads the live cart total, clamps at zero, states the threshold and region from the offer ledger, and repeats the threshold beside the price. | FAIL | OPERATOR |
| CA18 | Discount pills and ribbons | Design-rules N9; Baymard: show price and compare-at clearly, do not shout. | Compare-at is struck-through text only. No "31% OFF", "BEST VALUE", "MOST POPULAR". | FAIL | OPERATOR |
| CA19 | Urgency devices on health, supplement or medical-claim pages | India ASCI and CCPA misleading-ads guidelines require substantiation for health claims; Meta downgrades sensationalised health ads (https://www.facebook.com/business/news/reducing-low-quality-ads-on-facebook ). Timers beside "clinically proven" invite regulatory review of both. | Pages whose copy carries a health, efficacy or safety claim carry no countdown, stock indicator or "last chance" copy. Offer terms are stated plainly. | FAIL | LAW, OPERATOR |
| CA20 | "Buy now" as the primary CTA on tof or mof types | Offers research: "Buy now" assumes a most-aware reader; Baymard and NN/g: "Add to cart" is the familiar wording. `references/copy/headline-and-cta-rules.md` HC14. | Use the checklist `cta.copy_pattern`. "Buy now" only on `offer-page`, `restock`, `sale-clearance-flash`. | WARN | OPERATOR |
| CA21 | Faux progress bars, fake processing delays | FTC bucket I; DP22. | Progress reflects real steps; no decorative delays. | BLOCK | LAW |

## 3. Forms and overlays

| Id | Anti-pattern | Evidence | Rule | Sev | Tag |
| --- | --- | --- | --- | --- | --- |
| CA22 | Long forms | Baymard: ideal checkout is about 8 fields vs an average of 11.3; 17 to 18 percent abandon for "too long or complicated". Popupsmart (vendor): 1 to 2 field popups convert best. https://baymard.com/blog/checkout-flow-average-form-fields ; https://popupsmart.com/blog/popup-conversion-benchmark-report | Lead capture asks for one field (email or phone); a second field goes on step two. No coupon field above the fold; hide address line 2. | FAIL | RESEARCH |
| CA23 | Chat launcher covering the CTA or sticky bar | CleanCommit: moving the chat bubble off the CTAs, shrinking it and auto-minimising on mobile lifted conversion across devices. https://cleancommit.io/ab-tests/repositioning-the-chat-pop-up-significantly-improves-conversion-rates-across-devices/ | Launcher at most 56 px, opposite corner from the sticky CTA, hidden until the hero scrolls out, never auto-opens. Counts against the fixed-height budget (MA5). | FAIL | RESEARCH |
| CA24 | Cookie banner covering the hero | Google: legal interstitials are exempt from the penalty but should overlay, not replace, content. DSA 25(3)(a) and GDPR: no false hierarchy in consent. https://developers.google.com/search/docs/appearance/avoid-intrusive-interstitials | Where consent UI is legally required: bottom sheet at most 25 percent of viewport height, equal accept and reject buttons (DP10), no scroll lock, no marketing popup until dismissed. Where no non-essential cookies exist: no banner. | FAIL | LAW, OPERATOR |
| CA25 | Non-tappable phone, WhatsApp or email | Google Ads landing-page guidance: make it easy to contact you. https://support.google.com/google-ads/answer/7636512 | Every phone number is `tel:`, WhatsApp is `https://wa.me/`, email is `mailto:`. | WARN | OPERATOR |

## 3b. Popups and interstitials

| Id | Anti-pattern | Evidence | Rule | Sev | Tag |
| --- | --- | --- | --- | --- | --- |
| CA34 | Full-page interstitial on mobile entry | Google penalises intrusive interstitials on mobile entry from search (since 10 Jan 2017; legal, age and login gates exempt). https://developers.google.com/search/docs/appearance/avoid-intrusive-interstitials | No overlay taller than 60 percent of the viewport in the first 10 s; never a separate consent page. | BLOCK | LAW (platform) |
| CA35 | Popup without a real dismiss | Close meets A11's 48 x 48 px floor, 3:1, Esc and backdrop close, neutral decline copy (DP6, DP10). | As stated. | BLOCK | LAW |
| CA36 | Popup asking for two fields at once | Popupsmart (vendor): one or two fields convert best; Baymard: every extra field costs mobile completion. https://popupsmart.com/blog/popup-conversion-benchmark-report | One field per step; phone on step two if wanted. Consent boxes unchecked and separate for email and SMS. | FAIL | RESEARCH (vendor) |
| CA37 | Exit intent faked on mobile ("Wait!" on tab switch) | No mouse on touch; back-button or scroll-up intent only. | Use click-triggered ("Get my code") or scroll-triggered popups on mobile; Wisepops (vendor) reports click-triggered at about 54 percent conversion vs 3.9 percent exit intent. https://wisepops.com/blog/popup-stats | WARN | RESEARCH (vendor) |
| CA38 | Gamified popup with fake odds or fake delay | Spin-to-win is allowed only with real odds and every prize honoured; no "you almost won"; no fixed spinning delay (DP22). | As stated; odds recorded in the offer ledger. | BLOCK | LAW |

## 4. Proof and copy

| Id | Anti-pattern | Evidence | Rule | Sev | Tag |
| --- | --- | --- | --- | --- | --- |
| CA26 | Unattributed or invented testimonials | FTC Endorsement Guides: endorsements must reflect real experience; results need typicality. Baymard: 95 percent of shoppers rely on reviews; 53 percent seek negative ones. https://www.govinfo.gov/content/pkg/CFR-2023-title16-vol1/pdf/CFR-2023-title16-vol1-part255.pdf ; https://baymard.com/blog/respond-to-negative-user-reviews | Every quote is a `verified` proof-ledger row with attribution as stored, a date and a source. Include at least one non-five-star review when 20 or more exist. DP17. | BLOCK | LAW |
| CA27 | Trust badges nobody recognises | Baymard surveys: Norton about 36 percent recognition, McAfee about 23 percent, SSL vendor seals about 3 percent; CXL: unaided awareness of most seals "very low"; payment brands most trusted. https://baymard.com/blog/site-seal-trust ; https://cxl.com/research-study/trust-seals/ | Show the store's real payment methods (UPI, Visa, Mastercard, PayPal, Shop Pay, Apple or Google Pay) near the CTA. No generic "100% secure" shields; no seal the merchant does not hold (`references/proof/trust-badges-certifications.md`). | WARN | RESEARCH |
| CA28 | Conflicting proof numbers on one page | Teardowns: Jones Road showed 4.88 stars in the hero and 4.61 in the module; Create showed 22,000+ and 5,000+ five-star reviews on the same URL. https://www.jonesroadbeauty.com/pages/foundation-stick-performance ; https://trycreate.co/pages/5-reasons-try-create-sub | One average, one count, one customer count per page, each from a single ledger row, repeated verbatim. | FAIL | OPERATOR |
| CA29 | Year-stamped or dated headlines | Teardowns: AG1 "5 Health Benefits of Taking AG1 in 2025" and Stars + Honey "...in 2026" need maintenance and read stale after the year turns. https://drinkag1.com/5-reasons-why-variant-a | No year in a headline unless the page is a dated campaign with an offer-ledger end date. | WARN | OPERATOR |
| CA30 | Stock CTA labels ("Learn more", "Submit", "Click here", "Get started") | ContentVerve/Unbounce: "Get my..." vs "Get your..." +90 percent CTR in one test (magnitude unreliable); Google Ads: mirror the ad's call to action. https://unbounce.com/a-b-testing/failed-ab-test-results/ ; https://support.google.com/google-ads/answer/7636512 | CTA is verb plus object plus outcome or price (`references/copy/headline-and-cta-rules.md` HC12). Design-rules A12 and lint C4 already fail the stock strings. | FAIL | RESEARCH, OPERATOR |

## 5. Performance and mobile

| Id | Anti-pattern | Evidence | Rule | Sev | Tag |
| --- | --- | --- | --- | --- | --- |
| CA31 | Tap targets under the house floor | A11 requires 48 x 48 CSS px for every authored tap target, regardless of lower external minima. External context: https://www.w3.org/WAI/WCAG22/Understanding/target-size-minimum.html ; https://shopify.dev/docs/storefronts/themes/store/requirements | MA11. | BLOCK | LAW |
| CA32 | Slow first render | Google: a one-second mobile delay can cut retail conversions by up to 20 percent; Google Ads landing-page experience scores speed. https://support.google.com/google-ads/answer/7543502 | LCP under 2.5 s on simulated 4G at 390; hero image preloaded with `fetchpriority="high"`, everything below the fold lazy. | FAIL | RESEARCH |
| CA33 | Stacked fixed bars | MA4 and MA5 budgets. | At most one top and one bottom fixed region; total fixed height at most 30 percent of the viewport. | FAIL | RESEARCH |

## 6. Diagnostic frame before shipping

Run every selling page through Fogg's B = MAP as Sarah Levinger applies it
(https://www.linkedin.com/posts/sarahlevinger_sometimes-customers-buy-sometimes-they-activity-7457081960512000000-kmqK ):

| Axis | Question | Typical failure on this page | Fix source |
|---|---|---|---|
| Motivation | Does the first screen make the shopper feel the outcome, not just read the spec? | Informative hero, no before-state, no proof beside the claim | `references/copy/copy-frameworks.md`, proof proximity rule |
| Ability | Can they buy in under a minute? | Too many SKUs unexplained, price needing justification, returns unclear, forced account | CA13 to CA16, MA rules |
| Prompt | Is the next step obvious at every scroll position? | Buried CTA, vague "Shop now", no closing CTA | CA4, CA8, CA9, CA30 |

Answer each in one line in `QA record`.
finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the workStars: 322kUpdated Claude CodeCodexVerifiedInstall
systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixesStars: 322kUpdated Claude CodeCodexVerifiedInstall
test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation codeStars: 322kUpdated Claude CodeCodexVerifiedInstall
skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.Stars: 198kUpdated Claude CodeCodexVerifiedInstall