Skip to content
NEWMarketplace and your team: sell what you build, and get plans and changes that are tested before you see them. See what is new

Library / Reports, Analytics & Dashboards

Reports, Analytics & Dashboards

Summarize Etsy review themes and rank a fix queue in Sheets with Gemini

Run manually to classify low-star Etsy reviews into complaint themes using keyword rules or Gemini, then build a ranked fix queue in Sheets or CSV without drafting replies.

What it does

## Review Radar for Etsy Shops v0.1 Find complaint themes and a ranked Fix Queue from shop reviews. This workflow does not draft or send review replies. Default Config: themeEngine: "rules", mode: "mock", output: "files". No Etsy API and no Gemini unless you opt in. ### How it works 1. Load reviews from mock data, a CSV folder, or a Google Sheet Reviews tab. 2. Classify each review into themes (shipping, sizing, quality, description, packaging, other) with keyword rules — or optional Gemini. 3. Keep ★ ≤ minStarsForTheme (default 3) for theme aggregation; flag spam/empty rows. 4. Write four outputs: By Review, Themes, Fix Queue (listings ranked by complaint volume), and Summary. ### Setup 1. Import this workflow and click Test workflow with the defaults (mock + rules). 2. Self-hosted: CSVs land under outputFolder (default /home/node/.n8n-files/review-radar/). On n8n Cloud, inspect the Rows: nodes. 3. Your data: set mode to "files" (CSV reviews*.csv) or "sheet" (paste your Sheet URL; tabs Reviews plus the four output tabs). Set output to "files" or "sheet". 4. Optional Gemini: create Header Auth named x-goog-api-key, select it on Gemini classify (optional), set themeEngine: "gemini". Free-tier rate limits apply; themes can be wrong — always read the weekly report yourself. On failure the flow falls back to rules. Etsy and n8n are trademarks of their respective owners. Stallkit Studio is not affiliated with them. Not legal advice; not a substitute for reading all reviews.

The steps

  1. Run
  2. Config
  3. Mode = mock?
  4. Mock reviews (fake data)
  5. Mode = files?
  6. Read reviews CSV
  7. Parse reviews CSV
  8. Normalize files reviews
  9. Read Reviews sheet
  10. Normalize sheet reviews
  11. Theme engine = gemini?
  12. Prep Gemini (optional)
  13. Gemini classify (optional)
  14. Merge Gemini / fallback rules

36 steps in total.

Similar automations