One feed. Every product checked.
Your Shopify catalog, ready for ChatGPT Ads — 15 of 24 items in our sample would be blocked by the schema today.
FeedMedic audits every product against OpenAI's product-feed schema, names the row and field that fail, and gives you the fix. Read-only, and it works alongside the native ChatGPT Ads app.
$19.99/mo at launch. No charge until you subscribe.
The native app gives you a number. Not a reason.
These are the failures the audit finds first, with the row and field attached.
Description empty on 15 items
description is blank on 15 of 24 sample items. The row fails validation before it can be considered.
Price sent as “free” instead of 24.00 USD
price is parsed with a currency code. “free” and “19.99” without a currency are invalid values.
Duplicate item_id on variant rows
item_id must be unique. Duplicates are disapproved or merged into one listing.
Description under 50 characters on 9 items
The row still uploads, but thin copy weakens matching. The audit lists it separately from blocking errors.
What a report looks like
Severity first, then the item, the field, the OpenAI diagnostic code, and the fix.
| Severity | Item | Field | OpenAI code | Fix |
|---|---|---|---|---|
| Blocking error | SKU-2 | price | invalid_value | Format as 24.00 USD |
| Blocking error | SKU-2 | availability | invalid_value | Use in_stock, out_of_stock, pre_order, or backorder |
| Blocking error | SKU-1 | item_id | invalid_value | Make every item_id unique and stable |
| Warning | SKU-1 | title | invalid_value | Remove “SALE” and “FREE SHIPPING” from the title |
| Warning | SKU-1 | description | invalid_value | Write at least 50 characters of factual product description |
Sample audit. Your report maps every issue to OpenAI's feed fields and upload diagnostic codes.
Pricing at launch
Start free. Paid plans raise the item limit; every plan includes the full diagnosis.
Free
Up to 250 items.
- Per-item schema blockers
- Prioritized fix queue
- CSV export of the mapped feed
Pro
Up to 5,000 items.
- Everything in Free
- Per-item schema blockers
- Prioritized fix queue
Scale
Up to 50,000 items.
- Everything in Pro
- 50,000-item catalogs
- Same full diagnosis
| Feature | Free | Pro | Scale |
|---|---|---|---|
| Items per audit | 250 | 5,000 | 50,000 |
| Per-item schema blockers | Included | Included | Included |
| “Why is my feed 0?” fix queue | Included | Included | Included |
| Prioritized fix queue | Included | Included | Included |
| CSV export of the mapped feed | Included | Included | Included |
| Read-only scopes | Included | Included | Included |
| Email support | Included | Included | Included |
Agency — multi-store workflows are planned after launch and are not available at launch. Join the Agency waitlist.
Launch plans: Free, Pro, and Scale. Agency multi-store is planned. Install from the Shopify App Store at launch.
Questions
Does this replace the native ChatGPT Ads app?
No. It diagnoses the catalog and the feed the native app creates. It tells you why items are missing or blocked and how to fix them.
Why is my feed at 0?
That is the first thing the audit answers: missing required fields, image URLs, price format, availability, variants, or eligibility flags. The report names the row, the field, and the OpenAI diagnostic code (missing_required_column, invalid_value).
How is feed health scored?
Feed health is the percentage of items with zero blocking errors. Only errors move the score; warnings do not lower it. An item with only warnings still counts toward the health score.
When does it launch?
Early access. Email us to request access; there is no charge until launch.
Which countries are supported?
The audit works on any Shopify store. ChatGPT Ads is US-only at the time of writing.
Is my catalog data safe?
The app uses the read-only read_products scope. It does not edit your catalog.
See the row, the field, and the fix before you resubmit.
Read-only. No catalog changes.