Redesign above the fold

Before

After

Hypothesis

HYPOTHESIS We believe that redesigning the PDP to lead with social proof, add a Best Seller badge, simplify benefit scanning with icons, and use card-style bundle selectors will increase conversion rate because the new layout makes trust signals more prominent and the purchase decision easier to process. When visitors see 62,000 customers and a 4.7/5 rating BEFORE the product name, they receive immediate social validation that reduces skepticism. The Best Seller badge leverages bandwagon psychology (if everyone's buying it, it must work). Icon-based benefits with a scannable grid layout allow faster information processing than text-heavy checkmarks. Card-style bundle selectors create clearer visual hierarchy, making the "Best Value" 2-pack option more prominent and easier to select. This trust-first information architecture prioritizes answering the critical question problem-aware visitors need answered: "Will this product actually solve MY pain?" SPECIFIC CHANGES BEING TESTED 1. Social Proof Above Product Name • Control: Product name → then rating • Variation: Rating (4.7/5, 62,000 customers) → then product name • Why: Answer "Does this work?" before telling them what it is 2. Best Seller Badge • Control: No badge • Variation: Green "Best Seller" badge top-left • Why: Psychological trigger (if others bought it, it must be good) 3. Benefits: Icons vs Checkmarks • Control: 5 benefits with checkmarks (pain-point focused copy) ◦ "Ultra comfortable & supportive" ◦ "Prevents back & tailbone pain" ◦ "Durable & built for frequent use" ◦ "Takes seconds to set up & pack away" ◦ "Lightweight, compact & travel-friendly" • Variation: 4 benefits with icons (feature-focused copy) • Why: Icons are faster to scan, grid layout is more visually organized • Trade-off: Losing pain-point specificity ("Prevents back & tailbone pain") for scannable brevity ("Back & Neck Support") 4. Bundle Selector: Radio Buttons → Card Style • Control: Traditional radio button selectors • Variation: Card-style boxes with visual hierarchy • Why: Cards create clear visual separation, "Best Value" badge stands out more, feels more premium 5. Delivery Date Positioning • Control: Below trust badges • Variation: Above trust badges (higher visibility) • Why: Creates urgency earlier, "Get it between Mar 15-18" shown before they click ATC 6. Trust Badge Reordering • Control: "Free Tracked Shipping | 60 Day Risk Free Returns | 62,000+ Happy Customers" • Variation: "60 Day Risk Free Returns | 62,000+ Happy Customers | Free Tracked Shipping" • Why: Lead with risk reversal (60-day guarantee) vs shipping, keep social proof (62k customers) in trust badges since it's also at top 7. Expandable Sections Pushed Lower • Control: Description/How to Use/Shipping & Returns positioned mid-page • Variation: Same sections pushed further down page • Why: Prioritize conversion elements (bundle, ATC, trust badges) over educational content; engaged buyers will scroll for details WHY THIS REDESIGN WILL WORK • Trust-first hierarchy: Social proof (62k customers) answers "Does this work?" before product name answers "What is this?" • Visual clarity: Icon grid + card-style bundles are easier to scan than text lists + radio buttons • Psychological validation: Best Seller badge + leading with reviews leverages bandwagon effect • Urgency earlier: Delivery dates shown before ATC button plants timeline in mind • Reduced friction: Clearer bundle selection and streamlined information hierarchy

What we learned

What happened (top-line) - Variation lost: Conversion rate 9.11% → 8.42% (≈ -7.6%) - Revenue per visitor $11.01 → $10.32 (≈ -6.3%) - AOV $120.86 → $122.56 (≈ +1.4%), not enough to offset lower conversion - Primary leak appears early: ATC rate down ≈ -6.1% (biggest directional mover). Downstream rates are mixed/smaller (begin checkout -6.2%, abandoned checkout +3.7%, abandoned cart +1.5%). Likely reasons the redesign underperformed (ranked by plausibility) 1) Offer comprehension / bundle selector friction (BOGO + gifts) - The variation introduced a new card-style bundle module + "Best Value" hierarchy + potentially different price placement. With BOGO + multiple gifts, users need to quickly understand "what do I get, what do I pay, and what do I select?". If the module adds even slight confusion, it shows up as lower ATC. - In your own notes: price placement changed (right side → under BOGO), "Bundle & Save" wording, and a new physical recommendation element (not in control). All of these can increase cognitive load at the exact decision point. 2) Trade-off: scannability vs persuasion (loss of pain-specific benefit copy) - Control had 5 longer, pain-point benefits; variation shortened to 4 icon benefits and used different copy. For a higher effective price point (BOGO) buyers may need more reassurance, not less. Icons can scan faster, but may reduce perceived specificity/credibility. 3) Trust signals moved up, but "why this works for me" became less clear - Leading with social proof + Best Seller can reduce skepticism, but if it pushes the product story/unique value lower or makes the header feel like "generic ecommerce", it can weaken the fit narrative ("this solves my back/tailbone pain on flights"). That would also primarily hit ATC rather than later checkout steps. 4) Quality/accuracy issues can materially hurt trust - You called out a mismatch: rating should be 4.6/5 (not 4.7/5). Even small inconsistencies create "something feels off" friction, especially when the entire redesign is anchored on social proof as the first thing users see. 5) Segment mismatch: winners in Google, losers elsewhere (creative/traffic intent interaction) - Channel cut suggests Google improved (CR +12%, RPV +12%) while Facebook/Instagram/Klaviyo declined. That pattern often happens when higher-intent traffic tolerates changes (or benefits from clearer trust cues), while lower-intent / colder traffic needs stronger pain-story + less friction to select the offer. - This indicates the redesign might be situationally good, but not a universal PDP default without segmenting by traffic source or intent level. Notes on "% orders by quantity" chart - Since the 3 free gifts count as units, unit-quantity distributions are not a reliable read of bundle selection behavior here. Focus on offer selection events (bundle click / variant selection), ATC, and checkout step funnels instead of units-per-order from analytics. What to check next (diagnostics to confirm the hypotheses) - Funnel by step: PDP view → bundle selection interaction → ATC click (not just ATC rate). Look for lower interaction or higher "hesitation" in the offer module. - Clickmaps/recordings: do users scroll past the bundle module, mis-click, open FAQs, or hesitate at price/BOGO language? - Device split: card-style selectors sometimes degrade more on mobile (spacing, fold, thumb reach). - New vs returning: returning visitors may be more sensitive to layout/offer changes; new visitors may need pain-based persuasion. Recommended next iteration (fast follow test) A) Keep the layout improvements but remove the main risk variables: - Use the same benefit copy length as control (pain-point bullets), even if you keep icons. - Revert bundle module to control structure/price placement (price on right) while keeping the social proof/top trust stack. - Remove "Bundle & Save" wording and the physical recommendation element for a cleaner decision path. - Fix the rating/review values to be accurate (4.6/5). B) If you want to test "social proof first", isolate it: change only the header ordering + Best Seller badge, keep offer module identical to control. Core learning - This redesign likely failed not because "trust-first is wrong", but because the variation bundled multiple high-impact changes at the decision point (offer module + copy + new elements). The net effect increased friction and reduced ATC, and the small AOV lift could not compensate.