Field mapping is where you tell the app which piece of Shopify data feeds each Amazon attribute. It is opened from the Sync settings card on Product Sync, and it applies to your whole catalogue.
Most merchants never need to touch it. The defaults are sensible and cover a normal Shopify product. You come here when Amazon is asking for an attribute that Shopify does not hold in the obvious place, or when the default mapping produces content you are not happy with.
Out of the box, every push is built from your Shopify data like this.
| Amazon attribute | Fed by default from |
|---|---|
item_name (the listing title) |
The Shopify product title |
brand |
The Shopify vendor |
manufacturer |
The Shopify vendor |
product_description |
The Shopify description, with the HTML stripped out |
bullet_point |
The first 5 sentences of the Shopify description |
generic_keyword (backend search terms) |
The Shopify tags |
condition_type |
Fixed to new |
| The product identifier (UPC, EAN or GTIN) | The variant barcode. The type is inferred from how many digits it has |
| Images | The featured image plus the first 8 product images |
list_price and the offer price |
The variant price |
fulfillment_availability |
The variant inventory quantity |
Two of those deserve a closer look.
The description is sent to Amazon with its HTML removed, because Amazon does not accept your storefront markup in the description field. If your Shopify description leans heavily on tables, styled boxes or embedded media, what reaches Amazon is the plain text inside it, which may read poorly. That is worth checking on your first few products.
The bullet points are taken by slicing the first five sentences out of your description. This works reasonably well for descriptions that open with the key selling points and works badly for descriptions that open with a paragraph of brand storytelling. If your bullets look weak, either re-point bullet_point at something better, or fix them per market with the AI tools. See AI tools.
In the Field mapping modal, each Amazon attribute has a source selector. Change it and the new source is used on every push from then on.
The available sources are:
custom.my_field. This is the workhorse option. If you already keep structured data such as material, care instructions or a manufacturer part number in Shopify metafields, point the Amazon attribute straight at it.department of "unisex-adult" on a single-audience store.Tip
A Static value is the fastest way to clear an Amazon required attribute you cannot otherwise supply. It is not a cheat: Amazon simply needs an answer, and a correct constant is a correct answer. Do not use it to fake data that varies from product to product, because a wrong attribute value can get a listing suppressed.
The app exposes a further set of Amazon attributes that nothing feeds until you map them:
model_number, part_number, item_type_keyword, material, color, size, style, department, age_range_description.
You will typically meet these when the product type you chose declares one of them as required and the app reports "{n} required attribute(s) missing". Map the attribute to a metafield or a static value, save, and re-push. See Amazon product types and Required Shopify fields.
Two rules are worth committing to memory.
Mappings apply to every push. They are catalogue-wide settings, not per-product ones. Change a mapping and it takes effect on the next push of any product, including a scheduled resync. If you want a change to reach listings that are already live, run Sync all on the Product Sync screen.
A per-market AI override beats the field mapping. When you run one of the AI tools on a product and click Apply and save, the result is stored as an override for that specific product and marketplace. That override wins. Future pushes and resyncs will send the override, not the field-mapped value, until you use Clear override on the product. This is what makes the AI content stable: it does not get quietly overwritten the next time a resync runs.
Warning
Because mappings are catalogue-wide, a bad mapping is a catalogue-wide problem. After changing one, push a single product first and inspect the result on its detail page before you run Sync all across everything.