---
name: ecommerce-review-analysis
description_zh: "分析跨平台商品评论、评分、追评及客服反馈，保留 SKU、样本和证据，输出购买场景、好评驱动、痛点根因假设和改进优先级；用于 VOC 和竞品反馈比较，不负责选品或完整文案制作。"
description_en: "Analyze cross-platform product reviews, ratings, follow-ups and support feedback with SKU-level evidence, purchase scenarios, positive drivers, pain themes and action priorities. Use for VOC and competitor feedback, not product selection or full copy production."
category: ecommerce
---

# Ecommerce Review Analysis

1. Identify the corpus and intended decision. Read [data preparation](references/review-data-privacy.md) to map fields, deduplicate and preserve platform/product/variant/time boundaries.
2. Use [VOC taxonomy](references/voc-taxonomy.md) to label purchase scenarios, positive drivers and complaints. Retain a link from each theme to actual anonymized source records.
3. Use [root causes and actions](references/root-cause-and-actions.md) to separate symptoms from hypotheses and prioritize corrective work.
4. Deliver [the evidence-led summary](references/voc-output-templates.md) at the requested depth. Report what the sample supports, what it excludes and the next useful validation.

Supplied CSV, spreadsheet, JSON or document exports require no external data provider. Public reputation samples are supplementary, not complete seller analytics. This Skill ships no review scraper. Do not treat rating, text sentiment, verified purchase or fraud suspicion as interchangeable facts.