You can pass your product data to the Emarsys Marketing Platform in the Google Product Feed (GPF) format, which we automatically convert into our native CSV format. If you already have a GPF product file there is a good chance that you can use it as-is; just make sure that your GPF contains all the product data fields required by Emarsys.
Prerequisites
- A UTF-8 encoded GPF file.
If your GPF file is over 400 MB in size, please contact Emarsys Support before configuring the scheduled updates in the web UI.
We strongly recommend not using any whitespace characters (e,g. linebreak, space, tab) in the XML format GPF product file, unless they are explicitly needed. GPF files - when applied in XML format - work best without any additional whitespace characters. XML file specifications can be found here.
WEBP image format is not supported.
You must validate your GPF file in the Google Merchant Center before trying to upload it to Emarsys as we offer full product data validation for the Emarsys CSV format only.
Please pay particular attention to column names. Emarsys does not support multiple columns with the same name (e.g. additional_image_link) but Google will not show this as an error.
Google Product Feed fields in the Emarsys CSV format
The table below shows how we convert GPF fields into Emarsys product data fields:
GPF |
Emarsys CSV | Required in Emarsys | Comments |
|---|---|---|---|
id and item_group_id
|
item |
Yes | Emarsys discards product variants (i.e. similar products that differ by such attributes as size, color and material only) and creates a single parent-level item of each product group. We use the GPF field item_group_id to identify product groups. Learn more. |
| The parent-level product in the Emarsys CSV will inherit the attributes of the first item in the GPF product group. See the examples below. | |||
title |
title |
Yes | |
link |
link |
Yes | |
image_link |
zoom image and image
|
No | The original URL is pasted in the zoom_image field. |
| It is also used for the image field, but translates into a URL that creates the thumbnail image on the fly for email recommendations. Learn more. | |||
google_product_category or product_type
|
category |
No | To learn more about how we convert categories, click here. |
availability |
available |
No | GPF supports 4 values:in stock, out_of_stock, preorder, and backorder. In Emarsys it is either true or false.true: in stock, preorder, backorderfalse: out of stock, out_of_stock |
description |
description |
No | |
sale_price |
price |
No | If both GPF price fields are populated, this is how it converts. |
If only sale_price is populated, we use that value in both price and msrp in the Emarsys CSV. | |||
price |
msrp |
No | If both GPF price fields are populated, this is how it converts. |
If only price is populated, we use that value in both price and msrp in the Emarsys CSV. | |||
brand |
brand |
No | |
custom_label_[n] |
c_custom_[n] |
No | Read more about which GPF custom fields we support. |
Supported countries/regions and languages
We strive to offer support for all target countries/regions and languages supported by Google. If Google does support your country/region and language and you do not find them in the dropdown list when uploading your GPF in the Emarsys Marketing Platform web UI, please contact Emarsys Support.
If we do not support your country/region and language, you can always select United States as your default locale.
Using multiple feeds
We support multiple GPF feeds for clients who target multiple countries/regions or use multiple languages, but we do not support multiple feeds for the same country/region and language.
Multiple GPFs are merged into a single CSV product data file, based on the rules defined in our localization guidelines.
Only the products included in your default locale are added to the Emarsys CSV data file during the conversion. If you offer products in other locales that you do not sell in your default locale, make sure to include them in your default GPF file as well and set their availability to out of stock. Keep the same item_group_id and SKU across the default and locale product data files for the same product.
If you are not sure how to do this, please contact Emarsys Support.
item_group_id in local product data files is not fully supported at this point. Do not use item_group_id if you use local product data files too as some of the local products may be discarded from the final catalog (even if those are in stock in that local). Do not group your variants under the same item_group_id, but just keep them as separate id-s in order to reflect availability properly.
Product categories
GPF supports two attributes for specifying the category a specific product belongs in.
-
google_product_category - The product categories of your items according to the Google product taxonomy. This field has an important limitation: it can only include one category path for a product. It can be populated with either a numeric value (the standardized category ID) or the full path of the category (eg.
root > cat1 > subcat2). - product_type - The product categories you define for your products. In this field, you can include multiple category hierarchies. This field has an important limitation: Google does not support the use of commas ( , ) within the individual values submitted in this field as it is reserved for separating multiple values.
When the original GPF is processed and converted into our native CSV format, we will first look at the content of the product_type field. If it is populated, this is what we will use in the product data file. Only if product_type is not populated will we use the google_product_category field.
If you use numerical category IDs in the google_product_category field these will be converted to the corresponding human-readable category path. For example, 4745 will convert into Electronics > Computers > Tablet Computers in our native CSV format. When implementing our Web Extend data collection scripts, always pass the human-readable category path in the category command to report category page views on your website.
The Parent/Child catalog feature allows you to handle two-tier product structure and use that data for recommendations and personalization. For more information, see Parent/Child catalog feature.
Item groups
The Emarsys Marketing Platform currently works with parent-level products only and it does not use information about the SKU-level variants of the same product. But Google requires you to list product variants that only differ from one another by attributes like size, color, material, etc. as separate products in your GPF.
Make sure to use the item_group_id attribute to group the different variants of the same parent products in your GPF as Emarsys needs this field to identify product groups. When we convert your GPF into the Emarsys CSV format, we only keep product groups and discard individual variants.
This table shows you how we convert your GPF product groups:
| GPF | Emarsys CSV |
|---|---|
| id,title,item_group_id 100_40,Mens leather shoe size 40, 100 100_41,Mens leather shoe size 41, 100 100_42,Mens leather shoe size 42, 100 |
item, title 100, Mens leather shoe size 40 |
Here are the most common mistakes to avoid:
| GPF | Emarsys CSV | Comment |
|---|---|---|
| Id,title,item_group_id 100_42,Mens leather shoe size 42, 100_43,Mens leather shoe size 43, |
item, title 100_42,Mens leather shoe size 42 100_43,Mens leather shoe size 43 |
item_group_id is missing, all product variants will be listed as a separate item in the Emarsys product database. |
| Id,title,item_group_id ,Mens leather shoe size 42,100 |
[error] | Item is discarded because id is missing. |
| Id,title,item_group_id 100_42,Mens leather shoe size 42,100_42 |
[error] | Item is discarded because id and item_group_id are the same. |
If you use item_group_id in your GPF, you have to bear this in mind when implementing our Web Extend JavaScript API. In this case, make sure to pass the value from this field as the item attribute in the view, cart and purchase commands, otherwise Emarsys will not be able to identify these products.
Custom GPF fields
GPF supports two custom field types:
- custom attributes (feed rule)
- up to 5 custom labels, which are numbered 0-4 (e.g.
custom_label_1).
Currently, only custom labels are converted into the Emarsys CSV; custom attributes are ignored.