Exporting to CSV
The CSV export is for when the PDF isn’t the right format — when you want to filter, pivot, sort, or paste into another tool. CSVs are clean, machine-readable, and one row per product.
In this guide:
- Trigger a CSV export
- What columns are in it
- Cost (same as PDF)
- File expiry (same as PDF)
Trigger a CSV export
From a search detail view in Product Research → History:
- Click the search to open the detail view.
- Click the CSV · 1 credit button (top right).
- The export starts generating automatically.
CSVs follow the same lifecycle as PDFs: 1 credit, 1-hour download window, automatic refund if export fails. → Export to PDF for the full details on export statuses and lifecycle.
What’s in the CSV
One row per product. Columns (in order):
| Column | Notes |
|---|---|
asin | Amazon Standard Identification Number. |
marketplace | US / UK / DE / etc. |
rank_in_shortlist | Where the product placed in the SellerScore-sorted list (1 = top). |
title | Product title. Quoted; commas escaped. |
brand | Product brand. |
image_url | Direct URL to the product image (Cloudinary-served). |
amazon_url | Direct URL to the Amazon listing. |
sell_price | Numeric. |
currency | ISO 4217 (USD, GBP, EUR, etc.). |
bsr | Best Sellers Rank at search time. |
bsr_category | Amazon’s rank category. |
review_count | Integer. |
avg_rating | Float, 1.0–5.0. |
seller_score | Composite, 0–100. |
score_demand | Sub-score, 0–100. |
score_margin | Sub-score, 0–100. |
score_competition | Sub-score, 0–100. |
score_trend | Sub-score, 0–100. |
score_seasonality | Sub-score, 0–100. |
score_cross_platform | Sub-score, 0–100. |
data_quality | 0–100. |
data_warnings | Pipe-separated list of warning codes. |
cogs_estimate | Numeric. |
cogs_source | aliexpress, benchmark, user_override, or default. |
fba_fee | Numeric. |
referral_fee | Numeric. |
net_profit_per_unit | Numeric. |
net_margin_pct | 0.0–1.0 (multiply by 100 for percent). |
score_reasoning | The Claude-generated paragraph. Quoted; newlines escaped. |
ai_brief_summary | A short version of the AI brief (the full brief is too long for CSV). |
Encoding
UTF-8, BOM-prefixed for Excel compatibility. Comma-separated. Quotes around any cell containing commas or newlines.
If your downstream tool prefers a different delimiter (TSV, semicolon-separated for some EU locales), import the CSV into a tool that lets you re-export. Native delimiter options are on the roadmap.
What’s NOT in the CSV
- Product images (just URLs).
- The AI brief in full (only a summary; the full brief is too long for a cell).
- Search metadata (the search keyword, filters, weights). The CSV is per-product, not per-search.
If you need a search-level summary, use the PDF — it has a cover page with that context.
Cost & expiry
Same as PDF: 1 credit, 1-hour download window. → Export to PDF for details.
Workflow patterns
Spreadsheet model
Open in Excel / Google Sheets → add columns for your COGS, freight estimates, expected ad spend → recompute net margin per your assumptions.
Supplier outreach
Filter by SellerScore > 70, sort by net profit per unit, copy ASINs into a supplier email. The CSV is the source-of-truth for the shortlist.
Cross-search comparison
Export multiple searches, concatenate in a spreadsheet, sort by SellerScore across all of them. Useful when comparing categories.
Pipeline integration
For developers, the CSV is the easiest path to ingest into your own tools (BI, CRM, project tracker). The schema is stable; if you script against it, columns won’t shift unexpectedly.