Google Merchant Feed Audit API

Turn messy product catalog rows into explainable Google Merchant feed audits - without inventing missing attributes.

FeedFix API accepts public product titles, descriptions, and existing attributes, then returns a stable JSON audit with taxonomy, normalized fields, missing data, violations, evidence, confidence, and a compliant AI-generated structured title. It is an enrichment API, not a product scraper and not a guarantee of Google Merchant approval.

✅ What you get / ❌ What this isn't

✅ One structured audit row per successfully processed product
✅ Evidence spans for extracted attributes
✅ Missing fields reported instead of hallucinated values
✅ Google Merchant structured_title source marker
✅ JSON, CSV, Excel, API, scheduling, webhooks, and Apify integrations

❌ No website scraping or image analysis
❌ No invented GTINs, brands, materials, or specifications
❌ No automatic upload to Merchant Center
❌ No guarantee that Google will approve a listing

🔎 Why use this product feed audit API

• Preflight a supplier feed before importing it into your store or ad workflow.
• Normalize inconsistent catalog attributes across multiple suppliers.
• Identify missing identifiers and product details at batch scale.
• Generate factual structured titles with the required AI source marker.
• Route low-confidence rows to human review using confidence and status.

📊 What data you get

Field	Type	Description
status	string	valid, needs_changes, or rejected
qualityScore	number	Feed quality score from 0 to 100
category	object	Best-supported Google taxonomy category
normalizedAttributes	object	Evidence-backed normalized values
missingAttributes	array	Important fields absent from the input
violations	array	Machine-readable issues and explanations
suggestedStructuredTitle	object	Suggested title plus the AI source marker
confidence	number	Overall confidence from 0 to 1
evidence	array	Input spans supporting extracted values

👥 Who it is for

• Shopify and WooCommerce agencies
• PPC and Google Shopping specialists
• Marketplace operators
• Catalog and PIM teams
• Feed management SaaS products
• Developers building product ingestion pipelines

⚙️ How to audit a Google Merchant product feed

1. Open the Actor input.
2. Paste one or more product objects into products.
3. Run the Actor.
4. Review the dataset and route low-confidence or rejected rows to a person.
5. Download JSON, CSV, Excel, or call the Actor through the Apify API.

📥 Input

{
    "products": [
        {
            "id": "sample-sku-42",
            "title": "Example Air Runner black men's shoes size 42",
            "description": "Lightweight black running shoes for men with a mesh upper.",
            "brand": "Example",
            "attributes": { "size": "42", "color": "black" }
        }
    ],
    "concurrency": 5
}

Each run accepts up to 1,000 products. A product must contain a title or description. Only send public, non-sensitive catalog data. Do not submit credentials, private customer information, confidential supplier terms, or proprietary documents.

📤 Output

{
    "inputId": "sample-sku-42",
    "status": "needs_changes",
    "qualityScore": 82,
    "category": { "id": "187", "name": "Apparel & Accessories > Shoes" },
    "normalizedAttributes": { "brand": "Example", "color": "Black", "size": "42" },
    "missingAttributes": ["gtin", "material"],
    "violations": [],
    "suggestedStructuredTitle": {
        "content": "Example Men's Black Running Shoes, Size 42",
        "digitalSourceType": "trained_algorithmic_media"
    },
    "confidence": 0.93,
    "evidence": [{ "field": "color", "value": "Black", "sourceText": "black running shoes" }]
}

You can download the dataset in JSON, HTML, CSV, or Excel format. Programmatic callers can use the API tab and dataset endpoint.

💵 How much does a product feed audit cost?

The planned Store pricing charges for each successfully audited product:

Event	Free	Bronze	Silver	Gold
Product audit, per 1,000 products	$5	$4	$3.50	$3

For example, a 100-product test costs $0.50 on the Free tier. Failed validation and internal API failures do not produce a paid dataset result. Final live pricing is always shown in the Actor's Pricing tab.

🔁 Run it on the Apify platform

Use Apify schedules for recurring catalog QA, webhooks for downstream imports, and integrations with Make, Zapier, Google Sheets, or your own application. Set a maximum run charge when calling a pay-per-event Actor to cap spend.

⚠️ Limits and data handling

• Model output can be wrong. Review material commercial or compliance decisions.
• Taxonomy IDs depend on the model's available knowledge and should be verified before bulk import.
• This Actor is designed for public, non-sensitive product content because its underlying API project may share inputs and outputs with OpenAI.
• Product text is treated as untrusted data and cannot override the audit instructions.
• Only successful product audit rows are written to the default dataset.

❓ FAQ

Is this a Google Merchant API?

It prepares and audits product data for Google Merchant workflows, but it does not call Merchant Center or upload feeds.

Does it scrape product pages?

No. You provide the product text and attributes. This keeps runs fast, predictable, and suitable for existing feed pipelines.

Will it invent missing attributes?

It is explicitly instructed to report missing values instead of guessing, and it supplies evidence for extracted fields. Model mistakes remain possible, so use confidence thresholds and human review.

🛠️ Support

For a problem, open the Actor's Issues tab and include the run URL, a sanitized example input, and the expected behavior. Never post API keys or confidential catalog data in an issue.