Feedback Roadmap

Ecommerce teams receive customer feedback across a dozen touchpoints — product reviews, post-purchase surveys, chat transcripts, return reason codes, and social DMs — but rarely have a structured process to turn that raw signal into an actionable product improvement plan. This skill synthesizes feedback clusters into a prioritized roadmap with clear rationale, urgency scores, and suggested owner assignments so product and operations teams know exactly what to fix next and why.

Use when

• You have collected 20 or more customer reviews or survey responses and want to identify which product complaints appear most frequently so you can prioritize fixing them before next month's restock decision.
• A new product has launched on TikTok Shop or Amazon and you are seeing 1-star reviews pile up around a specific feature or packaging issue and need to escalate a fix recommendation to your sourcing team within 48 hours.
• You are preparing a quarterly product review and want to present stakeholders with a data-grounded list of the top 5–10 improvement opportunities ranked by customer impact, complaint volume, and estimated revenue risk.
• Your customer service team has logged a batch of recurring complaint tags (e.g., "size runs small", "lid leaks", "charging cable too short") and you want to convert those tags into a formal product backlog with severity tiers.

What this skill does

This skill ingests raw feedback text, structured complaint logs, or theme summaries and applies a systematic clustering and prioritization process. It groups related complaints into named issue categories, counts occurrence frequency, estimates customer sentiment severity on a 1–5 scale, and cross-references the issue against revenue risk factors such as return rate impact and repeat-purchase suppression. The output is a ranked improvement roadmap table with an executive summary, urgency tiers (Critical / High / Medium / Low), suggested resolution owner (Product, Ops, Content, Sourcing), and a one-sentence rationale for each priority decision.

Inputs required

• Feedback data (required): Raw review text, complaint log exports, survey open-ends, or pre-summarized theme lists. Paste directly or describe the top themes if you do not have raw text. The more specific the input, the more actionable the output.
• Product context (required): Product name, category, and selling platform (e.g., "silicone baby bottle set, TikTok Shop US"). This helps calibrate urgency based on platform-specific return and review policies.
• Business constraints (optional): Known limitations such as "supplier lead time is 90 days" or "we cannot change packaging before Q3" so the roadmap can mark certain fixes as deferred rather than immediate.

Output format

The output contains three sections. First, an Executive Summary (3–5 sentences) describing the dominant feedback themes and the single highest-priority fix. Second, a Prioritized Issue Table listing each issue cluster with columns for Issue Name, Complaint Volume (approximate), Severity Score (1–5), Urgency Tier, Suggested Owner, and a one-line Rationale. Third, a Next Steps section with three to five concrete action items formatted as owner-tagged tasks (e.g., "[Sourcing] Request updated sample with reinforced seam by April 15") to move the top-priority items forward immediately.

Scope

• Designed for: ecommerce operators, DTC brand teams, TikTok Shop sellers, Amazon private label sellers
• Platform context: TikTok Shop, Amazon, Shopee, Shopify, platform-agnostic
• Language: English

Limitations

• This skill does not connect to live review APIs or fetch data from Amazon, TikTok Shop, or Trustpilot in real time — all feedback must be pasted or summarized by the user.
• Urgency scores are qualitative estimates based on the input provided and should be validated against your actual return rate and refund data before making major sourcing decisions.
• This skill does not replace a dedicated product management process or customer research function — it is a structured first-pass analysis to accelerate prioritization, not a substitute for user interviews or VOC programs.