Review Radar: Solo E-Commerce Reputation Assistant

What it solves

Solo e-commerce sellers spend time reading reviews, negative feedback, support chats, and after-sales notes, but the harder part is deciding which issue is urgent and what to change next. ReviewRadar works on reputation data that sellers provide directly, turning scattered feedback into concrete operational actions instead of vague summaries like 'customers are unhappy.'

How it works

• Input scope: supports product reviews, negative reviews, support chat logs, after-sales feedback, CSV / Excel text, and mixed text.
• Evidence indexing: assigns Rxx, Cxx, and Axx IDs to reviews, chats, and after-sales notes, so key conclusions can reference specific source fragments.
• Risk scoring: estimates urgency using frequency, emotional intensity, conversion impact, after-sales cost, and diffusion risk, then shows risk level and confidence.
• Two modes: the compact mode outputs an overview, top issues, high-risk alerts, and today’s three actions; the full mode can add reply drafts, support knowledge-base updates, product page rewrites, and after-sales SOPs.
• Action conversion: produces P0 / P1 / P2 items such as following up with affected users, updating listing copy, or adding after-sales FAQs.

Limits to note

ReviewRadar only analyzes user-provided text. It does not call platform APIs, scrape data, or infer true return, complaint, or conversion rates. When input has fewer than three valid items, it falls back to a low-confidence compact mode and avoids major decisions. Refund, compensation, and complaint-related actions should still be confirmed by a human. It does not generate review manipulation, deletion, or misleading marketing advice.