Merchant Feed Doctor — Validate & Auto-Fix Shopping Feeds
Pricing
from $10.00 / 1,000 product checkeds
Merchant Feed Doctor — Validate & Auto-Fix Shopping Feeds
Validate and auto-fix Google Merchant Center product feeds from Shopify, WooCommerce or a native CSV/TSV. Finds price, GTIN, title, availability and identifier issues, applies only safe deterministic fixes, and returns a ready-to-upload TSV plus an HTML report. No Google API, no invented GTINs.
Pricing
from $10.00 / 1,000 product checkeds
Rating
0.0
(0)
Developer
Van Born Labs
Maintained by CommunityActor stats
0
Bookmarked
2
Total users
1
Monthly active users
3 hours ago
Last modified
Categories
Features
Share
Merchant Feed Doctor — Validate & Auto-Fix Google Shopping Feeds
Offline validator and auto-fixer for Google Merchant Center product feeds. Paste a Shopify export, a WooCommerce export, or a native Google CSV/TSV. Get a corrected feed, a scored report, and a fix log — no Google API, no invented data.
Stop losing impressions and getting disapproved products. Merchant Feed Doctor inspects Shopify, WooCommerce, or native Google Shopping feeds, reports every issue with a clear code and severity, and applies only safe, deterministic fixes — all without any network calls.
The problem
Google Merchant Center rejects or demotes products for dozens of silent reasons:
- Missing required attributes (id, title, description, link, image_link, price, availability, brand/identifier, condition)
- Titles that are too long, ALL CAPS, or stuffed with “free shipping”, “sale”, “!!!”
- Prices not in
19.99 USDformat, zero/negative prices, sale_price ≥ price - Invalid GTINs (wrong length or check digit)
- Non-https image links, spaces in URLs
- Clothing items missing color / size / gender / age_group or item_group_id (color or size alone does not make a product apparel)
- Duplicate IDs, ragged CSV rows, Windows-1252 encoding surprises
Fixing these by hand is slow and error-prone. This tool does it in seconds, offline, with a full audit trail.
What it checks
| Rule | Severity | Auto-fix |
|---|---|---|
| Required fields missing | error | — |
| Title > 150 chars, ALL CAPS, promotional text, heavy duplicates | error/warning | truncate at word boundary, sentence-case, strip promo phrases |
| Description > 5000 chars or contains HTML | error/warning | strip HTML |
| Price format, currency mismatch, zero/negative, sale_price ≥ price | error/warning | reformat to 12.50 USD (understands 1.299,99 €, 1,299.00, £12); drop a sale_price that is not lower |
| Availability not in {in_stock, out_of_stock, preorder, backorder} | error | normalize common variants (“in stock”, “yes”, “InStock”…) |
GTIN length 8/12/13/14 + GS1 check digit, spreadsheet damage (8.6E+11, '0123…) | error | strip apostrophe/spaces, restore leading zeros only if the check digit proves it; never invents a GTIN |
| shipping_weight without unit | warning | Shopify grams → 680 g, Woo kg → 0.2 kg |
| Ragged CSV rows | warning | — |
| link / image_link not absolute https or contains spaces | error/warning | http→https, URL-encode spaces |
| Duplicate ids | error | append -2, -3… |
| Variants missing item_group_id | warning | fill from Shopify Handle |
| Clothing/footwear missing color/size/gender/age_group | warning | — (not flagged on bedding, cosmetics, or other non-apparel just because they have Color/Size) |
| google_product_category not numeric ID or path | warning | — |
| condition values | warning | normalize |
Every fix is logged with before/after values.
Use cases
- Clean a Shopify or WooCommerce export before the first Merchant Center upload, or after a wave of disapprovals.
- Fix EU and US prices that spreadsheets have reformatted (
1.299,99 €,1,299.00,$19.9) into the19.99 USDshape Google requires. - Agency QA: run a client's feed, send them the HTML report and the fixed TSV, and charge per catalog instead of a monthly feed platform.
- Pipelines: hook the dataset,
fixed_feed.tsvandreport.jsoninto Make, Zapier, n8n or the Apify API.
Input
CSV or TSV. Auto-detects:
- Delimiter (comma, tab, semicolon)
- Encoding (UTF-8, UTF-8 BOM, Windows-1252)
- Source format (Shopify export, WooCommerce export, or native Google columns)
Real export layouts are handled:
- Shopify product CSV: variant rows inherit title/description/vendor/options from the first row of their Handle; image-only rows become
additional_image_link;Variant Compare At Priceis treated as the regular price andVariant Priceas the sale price; availability is derived from inventory qty/policy; draft/archived/unpublished products and gift cards are skipped (listed in the report);Google Shopping / *columns are mapped. - WooCommerce exporter:
variableparent rows are skipped and their variations inherit description/brand/images and getitem_group_id;Attribute Ncolumns become color/size; theImageslist is split;In stock?+Backorders allowed?map to availability;GTIN, UPC, EAN, or ISBNandBrandsare mapped. - Shopify and WooCommerce exports contain no product URLs: set
store_urland links are built as/products/<handle>(Shopify) or/?p=<id>(WooCommerce).
Mapped attributes: id, title, description, link, image_link, price, sale_price, availability, brand, gtin, mpn, condition, google_product_category, product_type, item_group_id, color, size, gender, age_group, shipping_weight.
Options
target_country(defaultUS)currency(defaultUSD)auto_fix(defaulttrue)title_max(default150)strict(defaultfalse)store_url(default empty) — shop base URL used to build product links for Shopify/WooCommerce exports
Output
fixed_feed.tsv— cleaned feed in Google Merchant formatreport.json— summary counts, issues found in the original feed,remaining_issuesleft after fixes, fixes list, skipped rows, health score before and after fixes (0–100)report.html— single-file dark report (charcoal background, white text, brand lime accent#B8FF2Cwith dark text on lime fills), sortable issues table, no external assets
Example report.json (excerpt)
{"summary": {"product_count": 62,"error_count": 211,"warning_count": 333,"fix_count": 148,"health_score": 50,"health_score_after_fixes": 90,"remaining_error_count": 0,"source": "shopify","encoding": "utf-8"},"issues": [{"code": "missing_link","severity": "error","message": "Required field 'link' is missing or empty","row": 2,"field": "link"}],"fixes": [{"row": 2,"field": "price","before": "29.99","after": "29.99 USD","reason": "reformat price to 'X.XX CURRENCY'"}]}
Usage
CLI (Python 3.11+)
python -m feed_doctor feed.csv --out out/python -m feed_doctor feed.tsv --out out/ --currency EUR --strictpython -m feed_doctor feed.csv --no-fix --json
As a library
from feed_doctor import process_feedfixed, report, html = process_feed(open("feed.csv", "rb").read())
Apify Actor
Input schema accepts feed_url or feed_text plus the options above.
The actor:
- Pushes one dataset item per product (
issues_found,remaining_issues,fixes,statusok/warning/error) plus a summary item - Saves
fixed_feed.tsv,report.html, andreport.jsonto the key-value store - Charges one
product-checkedevent per product item (Actor.push_data(..., charged_event_name=...)) and stops cleanly at the user's max-charge limit
FAQ
Does it call any external APIs?
No. Everything runs offline. GTIN check digits are computed locally. No Google API calls.
Will it invent GTINs or product data?
No. Invalid GTINs are flagged only. Fixes are limited to whitespace, HTML stripping, normalization, URL cleanup, title cleanup, price formatting, and ID deduplication.
What encodings are supported?
UTF-8, UTF-8 with BOM, and Windows-1252 (common in Shopify/Woo exports). Falls back to latin-1.
How is the health score calculated?
0–100 ≈ the share of products Merchant Center would accept. A product with any error counts 0, a product with only warnings counts 1 − 0.1 per warning (minimum 0.5), a clean product counts 1. The report shows the score of your original feed and of the fixed feed.
Can I use it on a private feed?
Yes. Paste the raw CSV/TSV into feed_text or run the CLI locally. No data leaves your machine.
Shopify / WooCommerce column mapping?
Yes. Common headers (Handle, Variant SKU, Variant Price, Body (HTML), Regular price, Stock status, etc.) are mapped automatically to Google attributes. Color/Size options become color/size.
Does it treat every product with a Color or Size option as apparel?
No. Gender and age_group warnings apply to clothing, footwear and wearable accessories (from the title, product type or Google category). Bedding, home goods and cosmetics are not flagged just because they have color or size variants, including makeup shade names like “Girl” or “Cardigan”.
Does a clean score mean Google will approve the feed?
No. This is a deterministic preflight of the file itself. Merchant Center also looks at your landing pages, account status, policies and additional attributes this tool cannot see.
Pricing
Pay-per-event pricing is live on the Apify Store:
- $0.25 actor start (
apify-actor-start) per run at the default 512 MB of memory. Apify bills one start event per GB of memory, minimum one, so a default run is a single $0.25 start. - $0.01 per
product-checked— $10 per 1,000 products checked
A 1,000-product catalog costs about $10.25 per run; 5,000 products cost $50.25. Skipped rows (Shopify drafts, gift cards, WooCommerce parent rows) are not charged. If a run hits the user's max-charge limit, the actor stops pushing products and says so in the summary.
The actor charges via Actor.push_data(..., charged_event_name="product-checked").
Disclaimer
Not affiliated with or endorsed by Google, Shopify, or WooCommerce. Google Merchant Center rules change; this tool checks the most common feed specification rules and does not guarantee approval.
License
MIT
Van Born Labs
Built for merchants and agencies who need reliable, auditable Google Shopping feeds.
More from Van Born Labs: https://arkadiuszvanborn.com/labs/merchant-feed-doctor