Download synthetic examples
These files use invented products, review IDs, and review text. They are provided to demonstrate the schema without publishing real reviewer data. The sample exports mirror schema 1.2; the analysis template is a separate working file for manual coding.
CSV and JSON structure
CSV is a flat table. Review fields and job-level metadata appear on every row so filtering a subset does not detach it from the export context. JSON keeps job metadata in meta, warnings in warnings, errors in errors, and review records in reviews.
schema_version, export_type,meta, warnings, errors, andreviews.Review record fields
| Field | Meaning |
|---|---|
| title | Review title when visible. |
| rating | Parsed star rating. |
| author_name | Public reviewer display name when visible. |
| date_and_location_text | Original date and location label from the page. |
| date | Normalized date when parsing succeeds. |
| body | Review text. |
| image_links | Array in JSON; pipe-separated URLs in CSV. |
| amazon_domain | Source Amazon domain. |
| asin | Product ASIN. |
| variation | Variation text when visible. |
| exported_at | Timestamp attached to the review record. |
| helpful_count | Parsed helpful-vote count when visible. |
| marketplace | Marketplace identifier selected by the extension. |
| review_id | Review identifier when present in the page. |
| review_url | Source review URL when available. |
| source_page_number | Observed collection page number. |
| country_or_location | Parsed location text when available. |
| verified_purchase | Boolean or unknown, depending on source visibility. |
A missing value means the page did not expose the field or the parser could not determine it. Downstream systems should preserve that uncertainty instead of silently converting it to a negative statement.
Job metadata and filter fields
| Field | Meaning |
|---|---|
| export_schema_version | Current schema version: 1.2. |
| export_job_id | Identifier for the user-initiated export job. |
| export_status | Final job status. |
| completion_reason | Why collection finished or stopped. |
| source_dataset_completeness | FULL, LIMITED_BY_REQUEST, LIMITED_BY_EXTENSION, PARTIAL, or UNKNOWN. |
| request_mode | Product page, reviews page, or manual ASIN workflow. |
| source_url | Starting URL when available. |
| requested_reviews_count | Count selected by the user. |
| available_reviews_count | Source count when it can be determined. |
| target_reviews_count | Effective collection target. |
| collected_reviews_count | Rows collected before export. |
| extension_review_limit | Current product limit: 100. |
| requested_pages | Page scope implied by the request. |
| target_pages | Effective page target. |
| processed_pages | Pages observed during collection. |
| sort_by | Selected sort option. |
| reviewer_type | Selected reviewer filter when applicable. |
| star_filter | Selected rating filter. |
| started_at / finished_at | Job timing. |
| dataset_exported_at | File creation timestamp. |
| warnings / errors | Arrays in JSON; flattened code and message columns in CSV. |
Completion and warning semantics
Completion metadata distinguishes a requested subset from an interrupted result. Common successful reasons include collecting all available source reviews, reaching the requested count, collecting the current page, or reaching the extension limit. A partial job may also report login, CAPTCHA, page, pagination, parser, tab, cancellation, or unknown errors.
Warning codes can report missing fields, removed duplicate reviews, early pagination end, skipped images, or suspected selector drift. Consumers should expose these conditions rather than dropping them during import.
CSV encoding, escaping, and spreadsheet safety
- The downloaded CSV uses UTF-8 with a byte-order mark for common spreadsheet compatibility.
- Quotes, commas, tabs, carriage returns, and newlines are escaped.
- Formula-sensitive text prefixes are neutralized before a spreadsheet opens the file.
- Multiple image links are joined with a pipe character in CSV and remain an array in JSON.
- The filename follows
amazon_reviews_marketplace_asin_timestampwith the selected extension.
Schema consumers should check export_schema_version orschema_version before relying on fixed columns. Future product versions may add fields while retaining documented versioning.