Koskii — Master CRO & Conversion Rate Optimization Audit

Prepared: 27 August 2026 · Site: https://www.koskii.com · Audience: internal growth / e-commerce leadership Synthesis of pageaudit/homepage.md, pageaudit/plp.md, pageaudit/pdp.md, pageaudit/cart-checkout.md, pageaudit/support-and-utility.md, plus business-analysis.md, web-quality-audit.md, competitive-analysis.md, ai-seo-audit.md, digital-brand-intelligence.md, ad-intelligence.md, content-strategy.md, traffic-estimation.md.


1. Executive Summary (SCQA Format)

  • Situation: Koskii is a funded (₹61 Cr Series A, 2023), omnichannel women's ethnic / occasion-wear challenger — ~14,227 SKUs, ~30 stores across 8 cities, median price ~₹5,400, and an estimated 937K monthly visits (top of the Medium traffic tier; traffic-estimation.md). The storefront is a technically solid headless Shopify + Next.js build with genuine strengths: a strong predictive search, a working guest checkout, a rich cart-level coupon engine, star ratings on listing cards, named customer testimonials, and best-in-class returns/shipping policy copy.
  • Complication: The conversion funnel has three high-confidence structural leaks that disproportionately hurt a high-consideration, sizing-sensitive, mobile-dominant audience: (1) the PLP faceted-filter engine does not execute — selecting any colour/price leaves a 1,108+ SKU grid unchanged (audited twice; Critical); (2) the PDP shows no structured rating/review block even though Judge.me is installed and the homepage/PLP do show stars (Major trust gap, corroborated by competitive-analysis.md — Koskii is the only brand in its set without on-site reviews); and (3) two real support dead-ends — the Size Chart page renders empty and Contact Us is a 3-button shell with masked numbers (both Major). Layered on top: a mobile experience that fails Core Web Vitals (Performance 47, CLS 0.18), a homepage with zero semantic headings, and an English-only storefront in a multilingual category.
  • Question: How does Koskii convert its well-acquired, well-funded traffic into purchases when the discovery, trust, and sizing layers all leak at the moments that matter most for occasion-wear buyers?
  • Answer: Fix friction before adding traffic. The single highest-leverage win is repairing the PLP filter engine (it silently blocks narrowing on a huge catalog). Pair that with surfacing structured PDP reviews, building a real Size Chart, repairing Contact Us, and a cheap template sprint (homepage H1, zoom-unblock, image dimensions, country-code default). These are mostly low-to-medium effort and address confidence-level-1 friction, not strategic bets.

Overall CRO Score: 5.8 / 10 — a competent, conversion-aware storefront held back by a small number of high-impact functional defects and a trust/sizing layer that underperforms for the category.

Estimated revenue impact range (see §11): fixing the priority set is modelled to lift conversion +8% to +20% relative, i.e. roughly ₹7–20 Cr of incremental annual revenue at current traffic — before any traffic growth.


2. Funnel Analysis

Funnel: Visit → Browse/PLP → PDP → Add to Cart → Checkout → Purchase → Repeat.

StageObserved behaviour (this audit)Benchmark (B2C/fashion)Status
Visit → Browse (Homepage→PLP)Strong occasion taxonomy, search, mega-menu. Discount-led, not brand-led.Browse 40–60%Healthy, with messaging-balance caveat
Browse → PDP (PLP→Product)Filter engine broken — cannot narrow 1,108+ SKUs; no URL/filter state; login-gated wishlistProduct-view 30–50%Blocked artery (Critical)
PDP → Add to CartAdd-to-cart works for guests; cart persists. PDP reviews absent; Size Chart same-tab redirectAdd-to-cart 8–12%Trust/sizing friction (Major)
Cart → CheckoutGuest checkout reachable; rich coupon engine; GoKwik external hand-offCart→checkout 45–55%Payment hand-off opacity (Major)
Checkout → PurchaseFree shipping, transparent tax, ETA shown. Phone code defaults to Afghanistan(+93)Checkout completion 50–65%Minor friction
Repeat30 stores, app, Klan loyalty — but Klan page under-sold; no structured post-purchase review captureMissing-link (Strategic)

Blocked arteries

  • The PLP filter failure is the clearest blocked artery: high-intent browsers on a 14K-SKU catalog cannot slice by colour/price/occasion, so they bounce instead of drilling down. This is the top conversion leak in the funnel.
  • PDP→cart trust: a ₹24,990 lehenga decision with no visible reviews on the product page (while the recommendations rail beside it does show stars) leaves the highest-ticket moments under-supported.

Missing links

  • No post-purchase structured review capture / no UGC loop (reviews exist only as curated testimonials on a separate page).
  • No cart-level upsell/cross-sell, no gifting fields (a notable miss for an occasion-wear gifting category).
  • No genuine scarcity/urgency (no low-stock, no delivery-countdown) — currently offset by pervasive discounts, which can train price-only buying (see §6 / Urgency).

3. Per-Page Reviews

3.1 Homepage

  • What's Working
    • Search is excellent — predictive overlay with trending chips, category suggestions, and inline "Recommended for you" product cards with quick-add (homepage.md §3). For a discovery-led ethnic catalog this is a genuine strength that reduces zero-result frustration.
    • Trust layer is real — the BESTSELLERS carousel renders 4.5–5.0★ star ratings (verified via wait-and-recheck) and named, occasion-specific testimonials (homepage.md §10). This directly counters the "no social proof" assumption.
    • Occasion taxonomy front-and-centre — haldi/mehendi/sangeet/reception grouping matches how Indian shoppers actually shop the wedding cycle; a differentiator vs. peers (business-analysis.md §21).
    • Merchandising is strong — festive edits, app-exclusive, Aaina collection, shop-by-size bands; footer surfaces 30 stores + app + franchise (omnichannel reinforcement).
  • Opportunity Areas
    • Category: Category B — Strategy-Dependent Choice
      • Observation: The above-the-fold opens with a discount ticker ("Flat 40% Off", live coupon codes) rather than a brand/value-prop statement; the "Celebrating The Girl In You" tagline was not surfaced on the homepage hero.
      • Possible Impact: Weakens differentiation scent for first-time brand-discoverers; may train price-only comparison behaviour.
      • Confidence: High Confidence Opportunity (discount density observed directly)
      • Effort to Fix: Medium
      • Realistic for Brand?: Yes — a brand-led hero variant can coexist with promo bands (an A/B cookie HomePageV2 already exists).
    • Category: Category A — Objective Friction
      • Observation: The homepage renders zero semantic H1–H6 headings (confirmed via DOM; web-quality-audit.md §2.1). This harms SEO topical signalling and screen-reader navigation on the most important page.
      • Possible Impact: SEO crawl clarity + accessibility; secondary CRO via clearer page structure.
      • Confidence: High Confidence Opportunity
      • Effort to Fix: Low (Next.js template)
      • Realistic for Brand?: Yes — pure template fix.
    • Category: Category A — Objective Friction
      • Observation: Product-card images largely lack alt text (imgAlt:"" on sampled cards; 344 imgs on page). Footer newsletter email input was not reliably rendered in this environment (flagged for verification).
      • Possible Impact: SEO crawlability + screen-reader exclusion; hurts Meta/ad Quality Score.
      • Confidence: High Confidence Opportunity (alt gap confirmed)
      • Effort to Fix: Low–Medium
      • Realistic for Brand?: Yes.

3.2 PLP (Product Listing Page)

  • What's Working
    • Sort works and is meaningful — Relevance / What's New / Price Low→High / Price High→Low genuinely re-order the grid.
    • Star ratings reliably inject on cards (4.0–5.0★ after a short delay) — a real trust signal at the consideration stage.
    • Infinite scroll + editorial SEO block below the grid (buying-guide copy, "Read More") supports both discovery and organic search.
    • Card click-through is clean (no dead-zones); "Add to bag" quick-add present.
  • Opportunity Areas
    • Category: Category A — Objective Friction
      • Observation: The faceted-filter engine does not execute. Selecting COLOR (e.g. "Red 91") or a PRICE range leaves the product count (1,108 / 241) and the exact same SKU set unchanged; the active-filter chip appears but the grid never re-queries. Reproduced on two independent PLPs with waits up to 13s (plp.md §1, §5).
      • Possible Impact: On a 14,227-SKU catalog this is the single most damaging leak — high-intent shoppers cannot narrow, so they bounce. Likely a front-end wiring bug (filter state → fetch not firing).
      • Confidence: High Confidence Opportunity (Critical severity per source audit)
      • Effort to Fix: Medium (front-end filter→query wiring + regression test)
      • Realistic for Brand?: Yes — requires engineering, not strategy change.
    • Category: Category A — Objective Friction
      • Observation: Filter/sort state is never written to the URL — filtered views cannot be shared, bookmarked, or restored via browser Back; analytics cannot attribute filter usage to a URL (plp.md F2).
      • Possible Impact: Lost shareable discovery, broken back-button context, analytics blindness.
      • Confidence: High Confidence Opportunity
      • Effort to Fix: Low–Medium
      • Realistic for Brand?: Yes.
    • Category: Category A — Objective Friction
      • Observation: Wishlist forces login — clicking the card heart on a logged-out session opens an "empty → LOGIN" drawer; no guest shortlist exists (plp.md F3). For a multi-outfit, compare-many-SKUs occasion category this suppresses shortlisting.
      • Possible Impact: Abandons comparison/shortlist intent at peak interest; lowers return-visit recovery.
      • Confidence: High Confidence Opportunity
      • Effort to Fix: Medium
      • Realistic for Brand?: Yes (localStorage guest wishlist merging on login).
    • Category: Category A — Objective Friction
      • Observation: Duplicate colour taxonomy — Mehandi green vs Mehendi green appear as separate facet buttons, fragmenting results (plp.md F4).
      • Confidence: High Confidence Opportunity · Effort: Low · Realistic: Yes.
    • Category: Category B — Strategy-Dependent Choice
      • Observation: No Quick View / no variant swatches on cards; no product-comparison feature (plp.md F5/F6).
      • Possible Impact: Extra PDP hops to check colour/variant; harder side-by-side compare.
      • Confidence: Moderate Opportunity · Effort: Medium · Realistic: Yes.

3.3 PDP (Product Detail Page)

  • What's Working
    • Gallery + variant switching are solid — 8–9 descriptive-alt images; colour swatch swap correctly updates title/image/code in place (pdp.md §2, §5A).
    • Structured product info is complete — DETAILS (fabric/work/lengths), SIZE & FIT, RETURN POLICY, and OFFERS accordions; "Inclusive of all taxes"; compare-at strikethrough + % OFF on sale SKUs.
    • Return confidence is strong — 7-day easy returns, pickup, courier reimbursement up to ₹300, exchange eligibility spelled out (pdp.md §4). This is a genuine reassurance strength.
    • "YOU MAY ALSO LIKE" recommendations are relevant and do show 4.5–5.0★ (a sharp, useful contrast — see below).
    • Guest cart persists — the earlier PDP-audit reading of "Add to Bag → empty cart" was a pre-hydration false positive; the cart-checkout traversal confirmed items are retained and the "Login to sync your cart" line is a non-blocking cross-device nudge (cart-checkout.md §0). Reclassified to Minor copy-clarity.
  • Opportunity Areas
    • Category: Category A — Objective Friction
      • Observation: No PDP-level rating/review block renders, even though Judge.me is installed (preloader + jdgm classes present in HTML) and the PLP/homepage do show stars. The "Reviews" accordion sits inside a md:hidden (mobile-only) container and surfaces no content on desktop (pdp.md F2; web-quality-audit.md §3.3 confirms no aggregateRating/review schema). Review data clearly exists (rec cards show stars) — it is simply not surfaced on the product page.
      • Possible Impact: For a deliberative, trust-sensitive ₹1,800–25,000 ethnic purchase, the absence of social proof on the product being decided on is a major conversion and SEO gap. Competitive-analysis.md confirms Koskii is the only brand in its peer set without structured on-site reviews.
      • Confidence: High Confidence Opportunity
      • Effort to Fix: Medium (Judge.me theme-integration fix + schema)
      • Realistic for Brand?: Yes — engineering + content.
    • Category: Category A — Objective Friction
      • Observation: On the lehenga PDP the "Size Chart" link navigates away in the same tab (URL becomes /account/size-chart), destroying cart-drawer state and scroll position at the exact sizing-decision moment (pdp.md F3).
      • Possible Impact: Funnel leak at the highest-consideration step.
      • Confidence: High Confidence Opportunity
      • Effort to Fix: Low (modal or target="_blank")
      • Realistic for Brand?: Yes.
    • Category: Category B — Strategy-Dependent Choice
      • Observation: The COD badge reads "COD for orders under ₹11,000" on every PDP, including a ₹24,990 lehenga where COD is correctly withheld. The cart-checkout audit confirmed this is accurate (COD correctly not offered above threshold), so it is a copy/clarity issue, not a defect — the negative framing could be reassuringly rephrased for prepaid buyers (pdp.md F4; cart-checkout.md §4).
      • Confidence: Moderate Opportunity · Effort: Low · Realistic: Yes.
    • Category: Category A — Objective Friction
      • Observation: The lehenga PDP exposes a single size "L" with no S/M/XL grid (pdp.md F5). Either a merchandising limit or a variant-display bug — unconfirmed; Shopify .json returned the Next.js error shell.
      • Confidence: Moderate Opportunity · Effort: Low–Medium · Realistic: Yes (verify variant set).
    • Category: Category B — Strategy-Dependent Choice
      • Observation: No PDP Q&A/FAQ, no product video/360° view, no quantity stepper (pdp.md F6/F7/F9). For premium-bridal tier, video/360 would strengthen craftsmanship proof.
      • Confidence: Moderate Opportunity · Effort: Medium · Realistic: Yes (staggered).

3.4 Cart & Checkout

  • What's Working
    • Guest checkout is genuinely reachable — only an Email required; "Sign in" offered but not mandatory; no forced account creation or phone-OTP wall before shipping/payment (cart-checkout.md §4). This is a meaningful friction win vs. category norms.
    • Rich cart coupon engine — FLASH5 / KOSKIILUV10 with "add ₹X more to unlock" progress nudges; a real AOV driver.
    • Instant AJAX quantity math, transparent tax inline, clear "PROCEED TO BUY" CTA, per-line delivery ETA ("Get it by 29 Aug") and "7 Day Return available" trust badges.
    • Free shipping + specific ETA shown ("Estd. Delivery by Mon, Aug 31"); taxes shown before payment (no surprise).
  • Opportunity Areas
    • Category: Category A — Objective Friction
      • Observation: Payment is delegated to a third-party GoKwik external redirect — native UPI/Card/COD options are not shown inline; the shopper is bounced to an external domain to pay ("You'll be redirected to GoKwik Payments"). Legitimate, but a real context-switch/trust moment (cart-checkout.md §4).
      • Possible Impact: Checkout-exit spike at the payment step; loss of payment-choice visibility.
      • Confidence: High Confidence Opportunity
      • Effort to Fix: Low–Medium (reassurance microcopy + GoKwik trust badge; consider enabling Shopify native express inline)
      • Realistic for Brand?: Yes.
    • Category: Category A — Objective Friction
      • Observation: Phone country-code defaults to Afghanistan (+93), not India (+91) on first paint of the checkout contact field (cart-checkout.md §4).
      • Possible Impact: Domestic shoppers may submit mis-prefixed numbers → OTP/support failures.
      • Confidence: High Confidence Opportunity
      • Effort to Fix: Low
      • Realistic for Brand?: Yes.
    • Category: Category B — Strategy-Dependent Choice
      • Observation: "Login to sync your cart" nudge sits atop a populated guest cart; its wording could imply emptiness (cart-checkout.md F3). Re-copy to "Saved in this browser — log in to access on any device" and move below line items.
      • Confidence: Moderate Opportunity · Effort: Low · Realistic: Yes.
    • Category: Category A — Objective Friction
      • Observation: No in-cart shipping estimator / free-shipping progress bar; no gift message / order notes / "this is a gift" toggle; no cart-level upsell (cart-checkout.md F4/F5/F8). For a gifting-heavy occasion category these are missed AOV levers.
      • Confidence: Moderate Opportunity · Effort: Medium · Realistic: Yes.
    • Category: Category A — Objective Friction (risk)
      • Observation: The empty-cart recovery state was not verified in-session (removal did not empty the drawer); the login nudge renders regardless of contents, raising the classic "empty cart dead-end" risk (cart-checkout.md F7).
      • Confidence: Moderate Opportunity (risk) · Effort: Medium · Realistic: Yes — verify bestsellers render when empty.

3.5 Search & Navigation

  • What's Working: Predictive search with trending chips, category + product suggestions, and inline quick-add is a standout (homepage.md §3). Mega-menu is deep, image-rich, occasion-led. Breadcrumbs present and accurate on PLP.
  • Opportunity Areas
    • Category: Category A — Objective Friction — PLP filter engine broken (see §3.2) is the dominant navigation defect; it negates much of the strong taxonomy investment.
    • Category: Category B — Strategy-Dependent Choice — No result-page shareability / back-restore (no URL state). Mobile nav (mobile-bottom-nav, 2-col grid implied) not emulated; verify on real mobile (highest-priority follow-up for a mobile-dominant audience).

3.6 Policy & Support Pages

  • What's Working
    • FAQ (25 objection-defusing Qs), Returns (7-day + pickup + ₹300 reimbursement + clear refund timing), Shipping (free pan-India, named carriers), Store Locator (8 cities, no-login), Blog (shoppable WordPress with "Visit to Buy" rails), and named customer testimonials are all genuine strengths (support-and-utility.md §2).
  • Opportunity Areas
    • Category: Category A — Objective Friction
      • Observation: Size Chart page (/account/size-chart) renders emptymain.innerText length = 10 (title + back link only); no measurement tables, no PDF, no sizer (support-and-utility.md F2). Footer-linked and reproducible.
      • Possible Impact: Sizing is the #1 apparel-return/hesitancy driver; a broken guide suppresses add-to-cart confidence and inflates returns.
      • Confidence: High Confidence Opportunity
      • Effort to Fix: Medium (build real guide)
      • Realistic for Brand?: Yes — critical for apparel.
    • Category: Category A — Objective Friction
      • Observation: Contact Us is a near-dead-end — only 3 action buttons (WhatsApp/Phone/Email); no readable address, support email, landline, or hours; numbers masked (+918****7474); cart drawer pushes LOGIN (support-and-utility.md F1).
      • Possible Impact: Trust leak for a high-consideration first-time buyer; no "we reply within X" assurance.
      • Confidence: High Confidence Opportunity
      • Effort to Fix: Medium
      • Realistic for Brand?: Yes.
    • Category: Category B — Strategy-Dependent Choice
      • Observation: Koskii Klan loyalty page is FAQ-first — 25 collapsed accordions, no benefits hero, no earning-rate, no tier table, no proactive join CTA, despite loyalty being a named growth pillar (support-and-utility.md F3).
      • Possible Impact: Under-sells a strategic pillar; low opt-in/engagement.
      • Confidence: High Confidence Opportunity
      • Effort to Fix: Medium
      • Realistic for Brand?: Yes.
    • Category: Category A — Objective Friction
      • Observation: About Us & Customer Reviews end with no next-step CTA (dead-ends); 7/13 support links route via /account/* app-proxy (empty raw HTML, SEO/no-JS blind, cart pushes LOGIN); contact-number inconsistency (masked site-wide vs full number on franchise) (support-and-utility.md F4/F5/F6).
      • Confidence: Moderate Opportunity · Effort: Low–Medium · Realistic: Yes.

4. The O/CO Framework (Objection / Counter-Objection)

Visitor Objection (Big 5)Current Counter-ObjectionGapRecommended CO
Trust — "Why believe this brand / will the outfit arrive as shown?"Named testimonials + 30-store footprint + returns copy + card-level starsNo structured PDP reviews; Contact page is a 3-button shell; off-platform complaint cluster (Trustpilot 4.0, MouthShut, IBTimes CS critique — digital-brand-intelligence.md)Surface Judge.me rating+reviews on PDP; add aggregateRating schema; repair Contact Us with readable email/phone/hours + a query form
Price — "Is it worth ₹1,800–25,000?"Compare-at + % OFF; "Inclusive of all taxes"; coupon unlock nudgesDiscount-led opening can frame Koskii as "cheap" not "value"; no ROI/quality proof for premium tierLead with craftsmanship/heritage value-prop hero; add PDP video/360 for bridal tier; keep discounts but balance with brand story
Fit — "Will this suit MY size/body?"DETAILS sizing specs; SIZE & FIT accordion; ONESIZE on sareesSize Chart page empty; Size Chart link same-tab redirect on lehenga; single "L" on lehenga PDPBuild real size guide (cm/inch tables, how-to-measure, printable PDF); open Size Chart in modal; surface all available sizes
Timing — "Why buy now?"Coupon codes + festive edits; free shippingNo genuine scarcity/urgency (no low-stock, no delivery countdown); pervasive discounts may blunt urgencyUse genuine low-stock/"only N left" or delivery-countdown for wedding deadlines — never fabricated urgency (ethical guardrail)
Effort — "How hard is this to buy/return?"Guest checkout, free shipping, clear ETA, 7-day pickup returnsGoKwik external payment hand-off opacity; phone code defaults to +93; no in-cart shipping estimatorReassure payment step (GoKwik badge + microcopy); force +91; add in-cart pincode→shipping estimator

5. Persuasion Assets Audit

  • Credentials & Authority (present): 150-yr textile heritage narrative, 30 stores, Series A (Baring PE), founder story, named carriers in Shipping, warehouse transparency in Returns. Missing: no visible awards/ press logos on-site; no GST/business-reg number in footer (web-quality-audit.md §6.2); no media/celebrity endorsements on home.
  • Social Proof (partial): Named, occasion-specific testimonials + star ratings on cards/recs + "Bestsellers" labels. Missing: no UGC customer-photo reviews, no PDP-level rating block, no influencer/ambassador proof on home, no structured review schema. (This is the headline persuasion gap — see §3.3, competitive-analysis.md.)
  • Risk Reversal (strong): 7-day returns + pickup + ₹300 reimbursement, free pan-India shipping, COD under ₹11k, "100% Purchase Protection". Minor gap: returns/contact phone masked; no "no-questions" framing.
  • Data & Specificity (moderate): Delivery ETAs shown; tax transparent. Missing: no size/fit data viz, no "X bought this" specifics on PDP, no AOV/volume proof.
  • Asset Acquisition Wishlist (action plan):
    1. Launch structured reviews (Judge.me/Loox) with photo uploads + auto-request 7–14 days post-delivery — fixes the #1 trust gap and unlocks star rich-results (web-quality-audit.md §3.3).
    2. Collect UGC — incentivise customer photos for occasion wear (wedding/haldi/reception) to populate PDP + home.
    3. Add authority badges — "As seen / funded by Baring PE", heritage "since 2017 / 150-yr roots", store-count proof in footer.
    4. Publish press/awards if any exist; add GST/reg number for E-E-A-T.

6. Friction Audit (Fogg B = MAP)

Ordered by Fogg diagnosis — most conversion problems live in Ability, not Motivation, so Ability leads.

Friction TypeFindingSeverityQuick Win?
Physical / AbilityPLP faceted-filter engine does not execute (cannot narrow 1,108+ SKUs)High Confidence (Critical)No — needs eng, but highest ROI
Physical / AbilityMobile Performance 47; CLS 0.18 (CWV not passing); 7.1 MB payload; 5.9s main-thread (web-quality-audit.md)High ConfidencePartial — template fixes help
Physical / AbilityHomepage viewport maximum-scale=1.0 blocks zoom (WCAG 1.4.4 fail)High ConfidenceYes — 1 line
Physical / AbilityHomepage zero semantic headings; missing image altHigh ConfidenceYes — template
Physical / AbilityCheckout phone code defaults to +93 (Afghanistan)High ConfidenceYes — low
Physical / AbilityGoKwik external payment hand-off with no inline payment optionsSignificantPartial — microcopy
Physical / AbilityNo in-cart shipping estimator / gifting fields / upsellModerateNo — medium
Cognitive / AbilityNo URL/filter state (no share, no back-restore)SignificantPartial — low/med
Cognitive / AbilityLogin-gated wishlist; no comparison; no Quick ViewModerateNo — medium
Cognitive / AbilityDiscount-density / cognitive load on home (coupon ticker + sale mega-menu + strike prices)ModeratePartial — rebalance
Emotional / MotivationNo PDP structured reviews → trust hesitation on high-ticketHigh ConfidenceNo — medium (but high impact)
Emotional / MotivationEmpty Size Chart + same-tab redirect → fit anxietyHigh ConfidencePartial — low/med
Emotional / MotivationContact Us dead-end → pre-purchase reassurance gapHigh ConfidenceNo — medium
Emotional / MotivationKlan loyalty under-sold → weak repeat-purchase motivationSignificantNo — medium

7. Bugs & Structural Anomalies

  • PLP filter engine non-functional (COLOR/PRICE facets don't re-query; active chip shows but grid unchanged) — Critical functional bug (plp.md F1).
  • Size Chart page empty (/account/size-chart renders title + back link only) — true dead-end, footer-linked (support-and-utility.md F2).
  • PDP Judge.me widget not rendering — installed but no rating/review block on desktop (pdp.md F2).
  • Duplicate Product JSON-LD on PDP (two identical Product blocks) (web-quality-audit.md §3.3).
  • Viewport maximum-scale=1.0 disables pinch-zoom (web-quality-audit.md §2.1).
  • Homepage zero H1–H6 (web-quality-audit.md §2.1).
  • Checkout phone country-code default +93 (Afghanistan) (cart-checkout.md §4).
  • PDP "Size Chart" same-tab redirect destroys cart/scroll state (pdp.md F3).
  • No URL/filter persistence on PLP (plp.md F2).
  • Duplicate colour taxonomy (Mehandi/Mehendi) fragments facets (plp.md F4).
  • Contact Us masked + 3-button shell; cart drawer pushes LOGIN on info pages (support-and-utility.md F1/F5).
  • GoKwik hand-off removes inline payment-choice visibility (cart-checkout.md §4).
  • us.koskii.com is a separate classic Shopify theme (parity-drift risk) (web-quality-audit.md §4).

8. Quick Wins (Ranked by Impact/Effort — implement in 1–3 business days)

  1. Remove maximum-scale=1.0/minimum-scale=1.0 from viewport — unblocks zoom, WCAG + a11y score. (Low)
  2. Force checkout phone country-code default to +91 (India). (Low)
  3. Add a real homepage H1 + heading hierarchy. (Low)
  4. Re-copy the "Login to sync your cart" nudge to "Saved in this browser — log in to access on any device" and move below line items. (Low)
  5. Open PDP "Size Chart" in a modal or target="_blank" instead of same-tab redirect. (Low)
  6. Collapse duplicate Product JSON-LD to one block. (Low)
  7. Trim PDP/collection meta descriptions to ≤160 chars (PDP is 299). (Low)
  8. Add WhatsApp deep-link to footer (brand uses WA for support but footer omits it). (Low)
  9. Add explicit width/height (or aspect-ratio) to all <img> — kills CLS 0.18 → Good. (Medium)
  10. Auto-open drawer / show toast on PLP quick-add (like PDP does) to confirm add. (Low)

9. Top Opportunities (PIE-Prioritized)

PIE = (Potential + Importance + Ease) / 3.

#OpportunityPotentialImportanceEasePIEExpected Impact
1Fix PLP faceted-filter engine (re-wire filter→query; emit Shopify ?filter.*=; add regression test)101068.7Recovers browse→PDP on huge catalog; likely largest single CVR lift
2Surface structured PDP reviews (fix Judge.me render + add aggregateRating/review schema + post-purchase capture)9978.3Trust lift on deliberative purchases; star rich-results CTR
3Build real Size Chart (cm/inch tables, how-to-measure, PDF, interactive sizer; link from PDP size selectors)8867.3Raises add-to-cart confidence; lowers returns
4Brand-led homepage value prop (hero H1 + heritage/occasion story; balance discount density)7777.0Stronger differentiation; reduces price-only framing
5Repair Contact Us (readable email/phone/hours + lightweight query form + "reply within X")6766.3Closes pre-purchase reassurance leak

10. A/B Testing Roadmap

Format: If we [X], then [Y] will improve because [Z]. ICE = (Impact + Confidence + Ease)/3. Priority: P0 = implement without test, P1 = test first, P2 = when capacity allows.

#HypothesisICEPriority
1If we fix the PLP filter engine, then filtered-result sessions and PLP→PDP CTR will rise because shoppers can finally narrow 1,108+ SKUs.8.0P0 (it's a bug fix, not a bet)
2If we surface PDP star ratings + reviews, then PDP→add-to-cart will rise because deliberative buyers get the social proof they currently lack.8.3P1 (high research confidence; still validate lift)
3If we replace the discount-led hero with a brand-led value-prop hero (A/B via existing HomePageV2 cookie), then new-visitor bounce + brand-search will improve because differentiation scent strengthens.6.7P1
4If we add a real Size Chart + modal Size Chart link, then add-to-cart on sized categories will rise and returns fall because fit anxiety drops.7.0P1
5If we add GoKwik trust badge + "Secured by GoKwik, you'll pay on the next step" microcopy at payment, then payment-step completion will rise because the context-switch is de-risked.6.3P1
6If we force +91 country code + add in-cart pincode→shipping estimator, then checkout completion will rise because address friction and shipping surprise drop.6.0P2
7If we add genuine low-stock / delivery-countdown (wedding deadlines) — never fabricated, then urgency-driven conversion will rise for in-stock items.5.7P2 (ethical guardrail: real signals only)
8If we lead the Koskii Klan page with a benefits hero + tier table + join CTA, then loyalty opt-in / repeat-purchase will rise because the value is visible.6.0P2

11. Revenue Impact Estimate

Assumptions (explicit):

  • Monthly visits: 937,400 (SimilarWeb, traffic-estimation.md — top of Medium tier).
  • Current conversion rate: assumed 2.0% (mid of fashion/jewellery benchmark 1.5–2.5%; no on-site CVR was measurable in this audit).
  • Average order value: assumed ₹4,500 (median catalogue price ~₹5,400 softened by multi-unit saree buys; business-analysis.md §11).
  • Baseline monthly orders ≈ 937,400 × 2.0% ≈ 18,748; baseline monthly revenue ≈ ₹84.4M (~₹8.4 Cr); annualised ≈ ₹101 Cr (consistent with the ~₹95 Cr figure in business-analysis.md).

Modelled uplift from the priority set (§9):

ScenarioRelative CVR liftNew CVRIncremental orders/moIncremental revenue/yr
Conservative+8%2.16%~1,500~₹8 Cr
Optimistic+20%2.40%~3,750~₹20 Cr

The conservative case is anchored on the filter fix + PDP reviews + Size Chart alone; the optimistic case adds homepage value-prop, Contact repair, and checkout friction removal. These are directional models, not forecasts — true lift should be measured via the §14 plan and the §10 A/B tests.


12. CRO Score Breakdown

CategoryScore (/10)Notes
Value Proposition Clarity6Strong occasion taxonomy; but home is discount-led, no brand H1, tagline absent from hero
Trust & Credibility5Card stars + testimonials + returns strong; PDP reviews absent, Contact dead-end, off-platform complaint cluster
Navigation & Findability5Excellent search/mega-menu; PLP filters broken (Critical), no URL state
Product Presentation7Solid gallery/variants/details/returns; no video, single lehenga size
Cart & Checkout6.5Guest checkout + coupon engine strong; GoKwik opacity, +93 default, no in-cart estimator
Persuasion Assets5Testimonials + some ratings; no UGC, no PDP ratings, no review schema
Urgency & Scarcity4No genuine scarcity; pervasive discounts risk training price-only buying (Category B — strategic)
Risk Reversal7Best-in-class returns/shipping/COD copy; minor masking
Overall5.8Competent storefront held back by a few high-impact functional + trust/sizing defects

13. Recommendations Summary

Immediate (This Week) — bugs & quick wins

  • Fix PLP faceted-filter engine (P0, §9 #1).
  • Remove viewport zoom block; add homepage H1; force +91; re-copy cart nudge; open Size Chart in modal; collapse duplicate JSON-LD; trim meta descriptions; add footer WhatsApp; add image dimensions (§8).

Short-Term (This Month) — high-impact improvements

  • Surface structured PDP reviews + aggregateRating schema + post-purchase review capture (§9 #2).
  • Build a real Size Chart (§9 #3).
  • Repair Contact Us with readable details + query form (§9 #5).
  • Add GoKwik reassurance microcopy/badge; add in-cart pincode→shipping estimator.

Medium-Term (This Quarter) — structural changes

  • Brand-led homepage value-prop hero (A/B) balancing discount density (§9 #4).
  • Guest wishlist (localStorage) merging on login; add Quick View + variant swatches; product comparison.
  • Re-architect Klan loyalty page (benefits hero + tier table + join CTA).
  • Mobile performance sprint (code-split JS, lazy-load, AVIF/srcset) to pass CWV; true mobile traversal of PDP/cart/GoKwik.
  • Migrate /account/* info pages to server-rendered CMS or add SSR + structured data (SEO/no-JS).

Long-Term (This Year) — strategic initiatives

  • Structured UGC review loop + customer-photo program.
  • Localize storefront (Hindi/Tamil/Telugu) for tier-2 expansion (competitive-analysis.md).
  • Groom/kids "wedding basket" line to capture couple AOV (competitive-analysis.md top-3 learning).
  • Reactivate dormant blog (dormant since Jan 2023 — content-strategy.md) as top-of-funnel E-E-A-T + AI-search asset.
  • Defend branded + category Search; flip ad COD-posture toward prepaid/EMI/regional creative (ad-intelligence.md).

14. Measurement Plan

MetricToolFrequencyBaseline (assumed)Target
Conversion rateShopify Analytics / GA4Weekly2.0% (assumed)+8–20% relative
Cart abandonment rateShopify AnalyticsWeekly↓ via GoKwik reassurance
Add-to-cart rate (PLP→cart, PDP→cart)GA4 / ShopifyWeekly↑ post filter + reviews
Filtered-result sessionsGA4 (post URL-state fix)Weeklyn/a (broken)>0; shareable filtered URLs
PDP review-impression rateJudge.me / GA4Weekly~0>60% of PDP views
Email capture rateKlaviyo / ShopifyWeeklyverify (homepage input unreliable)establish baseline
COD vs prepaid splitPayment dashboardMonthlyCOD-prominentrebalance toward prepaid/EMI
Average order valueShopify AnalyticsWeekly₹4,500 (assumed)↑ via coupons + gifting
Mobile Core Web VitalsPSI / CrUXBi-weeklyCLS 0.18 (failing)LCP/INP/CLS all Good
Traffic by sourceGA4Weeklydefend branded/category

15. Mandatory Self-Check

CheckQuestionResult
FrictionDiagnosed via Fogg (Prompt → Ability → Motivation)?✅ Ability-led (filters, mobile perf, reviews, size chart)
MobileRecommendations tested on mobile?⚠️ With remarks — environment had no mobile emulation; mobile traversal flagged as highest-priority follow-up
EthicsNo dark patterns / misleading urgency?✅ Urgency recommendation limited to genuine signals
DataRecommendations backed by data/benchmarks?✅ Benchmarks cited; traffic/AOV/CVR stated as assumptions
PrioritizationTests PIE-prioritized?✅ §9 PIE table; §10 ICE
LocalLocal payment methods accounted for?✅ COD/UPI/GoKwik noted; +91 fix; prepaid/EMI rebalance advised
AssumptionsSector/traffic/CVR assumptions explicit?✅ §11 assumptions stated

Self-Check: Complete (with remarks on mobile emulation gap).

Assumptions:

  • Current CVR ~2.0%, AOV ~₹4,500, traffic 937,400/mo — modelled, not measured on-site.
  • CrUX/GA device split not available in this environment; mobile-dominance inferred from category norms + storefront signals (business-analysis.md §5a).
  • The PDP-audit "guest cart empties" CRITICAL was treated as refuted per the later cart-checkout traversal (wait-and-recheck evidence); reclassified to Minor.
  • The PDP COD-badge MAJOR was treated as confirmed accurate (COD correctly withheld >₹11k); reclassified to Minor copy issue.
  • Several interactive behaviours (mobile drawers, GoKwik payment page, empty-cart state, Judge.me root cause, Shopify variant set for GCSB0044758) could not be fully root-caused due to sandbox limits.

To validate:

  1. Run a true mobile traversal of PDP, cart drawer, and GoKwik payment hand-off (mobile-dominant audience).
  2. Confirm empty-cart recovery merchandises bestsellers (not just the login nudge).
  3. Engineering root-cause of PLP filter bug (capture the failing XHR) and Judge.me non-render.
  4. Verify Shopify variant set for the lehenga SKU (single "L" — limit or display bug?).
  5. Confirm homepage newsletter email input renders/works (unreliable in captured DOM).
  6. Establish real baseline CVR/AOV/email-capture via GA4/Shopify before projecting revenue.

Citations: pageaudit/homepage.md, pageaudit/plp.md, pageaudit/pdp.md, pageaudit/cart-checkout.md, pageaudit/support-and-utility.md, business-analysis.md, web-quality-audit.md, competitive-analysis.md, ai-seo-audit.md, digital-brand-intelligence.md, ad-intelligence.md, content-strategy.md, traffic-estimation.md — all in /root/cro-reports/koskii/20260827-154233/. Live evidence captured via headless browser DOM inspection + curl + PageSpeed Insights UI on 27 Aug 2026.