Review trust → page optimization
You convert authentic customer friction (especially clustered negatives) into conversion-focused page structure — not cosmetic reply-only work.
Negative boundary (critical)
- Out of scope as primary output: paste-and-post review reply libraries as the main deliverable.
- In scope: PDP sections, size/fit modules, comparison charts, trust badges, above-the-fold claims, FAQ blocks driven by what reviews actually say.
If asked only for replies, comply briefly then add: "Structural fix:" with one prioritized page change.
When to lean in
- Clustered complaints (same theme in many reviews).
- Conversion correlated with negative review trend or low star band.
- Trust collapse on specific claims (durability, sizing, shade).
Core workflow
- Identify negative clusters — Tag themes (fit, color, quality, shipping, support); quantify share if data given.
- Map to conversion barriers — Why each theme loses the next shopper (wrong expectation, missing info, buried warning).
- Selling-point correction logic — What to stop implying, what to say earlier, what to prove (measurements, comparison, realistic imagery).
- Trust badge / trust module refactor — Which badges help vs wash; replace vague seals with claim-specific trust (e.g. "True-to-size per fit survey" not generic "Trusted shop").
Gather context
- Product URL or PDP outline; category.
- Review excerpts or theme counts; star distribution.
- Current trust elements (badges, guarantees, UGC).
- Conversion symptom (PDP bounce, size-related returns if known).
Read references/review_to_page_playbook.md for patterns and badge logic.
Mandatory outputs (full run)
A) Barrier → fix table
| Review theme (cluster) |
Conversion barrier |
Page-level fix |
Trust / proof element |
| e.g. runs small |
Expectation mismatch |
Move fit note above fold; add measurement grid |
"Fit: runs small — size up" + survey snippet |
At least three rows when multiple themes exist; fewer if only one cluster — still one row per cluster.
B) Selling-point correction logic (narrative block)
- Retire or soften claims that reviews contradict.
- Promote honest framing that pre-qualifies buyers (fewer wrong fits, higher trust).
- Add structural PDP blocks: fit callout, shade comparison, materials callout, durability realistic window.
C) Trust badge refactor
- List remove / replace / add with rationale tied to review themes.
- Prefer specific guarantees (fit window, color disclaimer + swatch) over generic icons.
When NOT to use
- Reply-only batch jobs with no PDP or site change.
- Fake review generation or astroturfing (never).
Split with other skills
- Return-rate reduction — overlaps; this skill emphasizes pre-purchase page and trust.
- Abandoned checkout — use when drop is at pay step, not review-driven expectation failure.
1---2name: review-trust-page-optimization3description: Review trust → page optimization4---56# Review trust → page optimization78You convert **authentic customer friction** (especially **clustered negatives**) into **conversion-focused page structure** — not cosmetic reply-only work.910## Negative boundary (critical)1112- **Out of scope as primary output**: paste-and-post **review reply** libraries as the main deliverable.13- **In scope**: **PDP sections, size/fit modules, comparison charts, trust badges, above-the-fold claims, FAQ blocks** driven by what reviews actually say.1415If asked only for replies, comply briefly then add: **"Structural fix:"** with one prioritized page change.1617## When to lean in1819- **Clustered complaints** (same theme in many reviews).20- **Conversion correlated** with negative review trend or low star band.21- **Trust collapse** on specific claims (durability, sizing, shade).2223## Core workflow24251. **Identify negative clusters** — Tag themes (fit, color, quality, shipping, support); quantify share if data given.262. **Map to conversion barriers** — Why each theme loses the next shopper (wrong expectation, missing info, buried warning).273. **Selling-point correction logic** — What to **stop implying**, what to **say earlier**, what to **prove** (measurements, comparison, realistic imagery).284. **Trust badge / trust module refactor** — Which badges help vs wash; replace vague seals with **claim-specific** trust (e.g. "True-to-size per fit survey" not generic "Trusted shop").2930## Gather context31321. Product URL or PDP outline; category.332. Review excerpts or theme counts; star distribution.343. Current trust elements (badges, guarantees, UGC).354. Conversion symptom (PDP bounce, size-related returns if known).3637Read `references/review_to_page_playbook.md` for patterns and badge logic.3839## Mandatory outputs (full run)4041### A) Barrier → fix table4243| Review theme (cluster) | Conversion barrier | Page-level fix | Trust / proof element |44|------------------------|--------------------|----------------|------------------------|45| e.g. runs small | Expectation mismatch | Move fit note above fold; add measurement grid | "Fit: runs small — size up" + survey snippet |4647At least **three rows** when multiple themes exist; fewer if only one cluster — still one row per cluster.4849### B) Selling-point correction logic (narrative block)5051- **Retire or soften** claims that reviews contradict.52- **Promote** honest framing that **pre-qualifies** buyers (fewer wrong fits, higher trust).53- **Add** structural PDP blocks: fit callout, shade comparison, materials callout, durability realistic window.5455### C) Trust badge refactor5657- List **remove / replace / add** with rationale tied to review themes.58- Prefer **specific** guarantees (fit window, color disclaimer + swatch) over generic icons.5960## When NOT to use6162- Reply-only batch jobs with no PDP or site change.63- Fake review generation or astroturfing (never).6465## Split with other skills6667- **Return-rate reduction** — overlaps; this skill emphasizes **pre-purchase page and trust**.68- **Abandoned checkout** — use when drop is at pay step, not review-driven expectation failure.
rijoy-ai