Indexing attributes

The cartridge can index various product, price, inventory and categories data from Salesforce B2C Commerce environments.

Default records

The cartridge can generate two types of records, depending on your record model preference:

Variation Product records: create one record per product variant (default).
Base Product records: create one record per base product, containing all the variants in a variants attribute.

For more information about which model best suits your needs, see Ecommerce records.

Variant-level default model

json

{
  "name": "Apple iPod Classic",
  "primary_category_id": "electronics-digital-media-players",
  "categories": [
    [
      {
        "id": "electronics-digital-media-players",
        "name": "iPod & MP3 Players"
      },
      {
        "id": "electronics",
        "name": "Electronics"
      }
    ]
  ],
  "in_stock": true,
  "price": { "USD": 219.99 },
  "image_groups": [
    {
      "_type": "image_group",
      "images": [
        {
          "_type": "image",
          "alt": "Apple iPod Classic, Silver, large",
          "dis_base_link": "/images/large/ipod-classic-silver.jpg",
          "title": "Apple iPod Classic, Silver"
        }
      ],
      "view_type": "large"
    }
  ],
  "url": "/electronics/ipod%20%26%20mp3%20players/apple-ipod-classic-silver-120gM.html"
  "__primary_category": {
    "0": "Electronics",
    "1": "Electronics > iPod & MP3 Players"
  },
  "objectID": "apple-ipod-classic-silver-120gM"
}

Product-level default model

json

{
  "name": "Apple iPod Classic",
  "primary_category_id": "electronics-digital-media-players",
  "categories": [
    [
      {
        "id": "electronics-digital-media-players",
        "name": "iPod & MP3 Players"
      },
      {
        "id": "electronics",
        "name": "Electronics"
      }
    ]
  ],
  "__primary_category": {
    "0": "Electronics",
    "1": "Electronics > iPod & MP3 Players"
  },
  "defaultVariantID": "apple-ipod-classic-silver-120gM",
  "variants": [
    {
      "variantID": "apple-ipod-classic-silver-120gM",
      "in_stock": true,
      "price": { "USD": 219.99 },
      "color": "Silver",
      "url": "/electronics/ipod%20%26%20mp3%20players/apple-ipod-classic-silver-120gM.html"
    },
    {
      "variantID": "apple-ipod-classic-black-120gM",
      "in_stock": true,
      "price": { "USD": 199 },
      "color": "Black",
      "url": "/electronics/ipod%20%26%20mp3%20players/apple-ipod-classic-black-120gM.html"
    }
  ],
  "colorVariations": [
    {
      "image_groups": [
        {
          "_type": "image_group",
          "images": [
            {
              "_type": "image",
              "alt": "Apple iPod Classic, Silver, large",
              "dis_base_link": "/images/large/ipod-classic-silver.jpg",
              "title": "Apple iPod Classic, Silver"
            }
          ],
          "view_type": "large"
        }
      ],
      "variationURL": "/electronics/ipod%20%26%20mp3%20players/apple-ipod-classicM.html?dwvar_apple-ipod-classicM_color=Silver",
      "color": "Silver"
    },
    {
      "image_groups": [
        {
          "_type": "image_group",
          "images": [
            {
              "_type": "image",
              "alt": "Apple iPod Classic, Black, large",
              "dis_base_link": "/images/large/ipod-classic-black.jpg",
              "title": "Apple iPod Classic, Black"
            }
          ],
          "view_type": "large"
        }
      ],
      "variationURL": "/electronics/ipod%20%26%20mp3%20players/apple-ipod-classicM.html?dwvar_apple-ipod-classicM_color=Black",
      "color": "Black"
    }
  ],
  "in_stock": true,
  "objectID": "apple-ipod-classicM"
}

To optimize record size, image_groups aren’t stored in each variant. For a given color, all size variations have the same image_groups. Instead of duplicating image_groups in all variants, only one is stored for each color: in the colorVariations attribute. The Algolia integration for Salesforce B2C Commerce uses a variant’s color information to determine which image to display.

Example

A site might have the following configurations:

hostname: example.com
site-id: clothing
Two locales: en_US, fr_FR
Two currencies: USD, EUR
Five additional product attributes: short_description, long_description, brand, color, size

This setup will generate the following indices in Algolia:

example-com__clothing__products__en_US
example-com__clothing__products__fr_FR
example-com__clothing__categories__en_US
example-com__clothing__categories__fr_FR

Algolia generates a separate product and category index for each locale to give you the most flexibility in terms of configuration. This setup lets you to have synonyms, rules, and other locale-specific settings on your index without interfering with the settings for other locales.

The prefix ${hostname}__${site_id} prevents accidental name collisions between your environments. It makes using a single Algolia application with multiple sandboxes or sites completely safe. You can always change the prefix with the Algolia_IndexPrefix custom site preference.

A typical variant-level record in these indices has the following structure:

jsonc

// index: example-com__clothing__products__en_US

// Base set (non-configurable attributes)
// theses fields are always added to your product record
{
  "objectID": "8987698M",
  "categories": [
    [
      { "id": "womens-tops-tshirts", "name": "T-Shirts" },
      { "id": "womens-tops", "name": "Tops" },
      { "id": "womens", "name": "Women" },
    ],
  ],
  "__primary_category": {
    "0": "Womens",
    "1": "Womens > Tops",
    "2": "Womens > Tops > T-shirts",
  },
  "in_stock": true,
  "price": {
    "USD": 12.99,
    "EUR": 11.99,
  },
  // configurable attributes
  "name": "Red T-shirt from PopularBrand",
  "short_description": "Short description of this red t-shirt",
  "long_description": "A much longer description of this red t-shirt",
  "brand": "PopularBrand",
  "color": "Red",
  "size": "L",
  "image_groups": [
    /* */
  ],
  "url": "/s/RefArch/womens/tops/tshirts/8987698M?lang=en_US",
}

Configure product records attributes

You can configure the list of attributes that are exported to your Algolia index. The list has a set of base attributes which you can extend to match your business needs. Configure the product attributes to send to Algolia by adding them to the Additional Product Attributes preference. You should do this before performing a complete (re)index of your data.

Go to Merchant Tools > Algolia > Algolia
Set up your Additional Product Attributes:

Base attributes (non-configurable)

The base set of attributes is always included in product indexing, except when using the attributeListOverride job step parameter, which overrides both the base attributes and any additional product attributes configured for that job.

Attribute	Description
`name`	Name of the product
`primary_category_id`	Primary category ID
`categories`	Array of categories assigned to the product, with all their parent categories
`url`	URL pointing to the product’s PDP (Product Detail Page).
`image_groups`	Array of product image URLs
`in_stock`	`true` if the number of products available is greater than the InStock Threshold you’ve set, `false` otherwise
`price`	Product price

Configurable attributes

You can choose to send additional Product record attributes such as the following to Algolia by defining them in Additional Product Attributes. You should start with the following attributes: short_description, long_description, then expand the list according to your use case (for example brand, color, size).

Attribute	Description
`EAN`	The European Article Number of the product.
`UPC`	The Universal Product Code of the product.
`brand`	The brand of the product.
`bundle`	Identifies if this product instance is a product bundle.
`bundled`	Identifies if this product instance is bundled within at least one product bundle.
`bundledProducts`	A collection containing all products that participate in the product bundle.
`color`	The product’s color.
`colorVariations`	An array containing an entry for each color variation. Each entry contains the `image_groups` of the variation, the `variationUrl`, and the variation `color`.
`long_description`	Full description of product.
`lsImage`	Expose the first product image at the root of the Algolia record. Used by the Looking Similar widget.
`manufacturerName`	The name of the product manufacturer.
`manufacturerSKU`	The value of the manufacturer’s stock keeping unit.
`master`	Identifies if this product instance is a product master.
`masterID`	ID of the master product.
`online`	The online status of the product. This is determined based on the product’s online status flag, and the onlineFrom and onlineTo dates.
`optionProduct`	Identifies if the product has options.
`pageDescription`	Returns product’s page description in the default locale.
`pageKeywords`	The product’s page keywords in the default locale.
`pageTitle`	The product’s page title in the default locale.
`productSetProduct`	`true` if this product is part of any product set, otherwise `false`.
`productSet`	Returns `true` if the instance represents a product set, otherwise `false`.
`promotionalPrice`	Product’s lowest promotional price, from the currently active promotions.
`refinementColor`	Search Refinement Bucket of the `color`.
`searchable`	Identifies if the product is searchable.
`size`	The product’s size.
`short_description`	Short description of product.
`unit`	The products sales unit.
`variant`	Identifies if this product instance belongs to a product master.
`newArrivalsCategory`	If this product is part of the new arrivals category.

To ensure the custom logic in the Algolia cartridge correctly displays the new arrivals facet:

Add the newArrivalsCategory attribute to the Additional Product Attributes list to send it to Algolia.
Declare newArrivalsCategory as an attribute for faceting in the Algolia dashboard.

Flexible attribute indexing based on default behavior

From version 24.5.0, the Algolia cartridge supports indexing any product attribute, even if they aren’t explicitly listed as configurable attributes. This feature simplifies the process of indexing custom or specific attributes, allowing for more flexible and comprehensive product data in your Algolia index without requiring extensive configuration. For attributes that aren’t explicitly listed, the cartridge uses the following default indexing configuration:

{
    attribute: attributeName,
    localized: false,
    variantAttribute: true
}

This means that by default, all unlisted attributes:

Aren’t localized
Are indexed in the variants array (for the product-level record model)
Use the default Salesforce B2C Commerce attribute name (with dot notation)

To change this default behavior, override it with advanced attributes configuration.

Nested attributes

The cartridge supports indexing nested attributes up to two levels deep. Nested attributes are grouped under their parent object in the Algolia record. For example: If you want to index activeData.impressionsDay and activeData.revenueDay, they will be indexed in Algolia as:

json

{
  "activeData": {
    "impressionsDay": 5,
    "revenueDay": 141
  }
}

If you prefer indexing these attributes at the root level of your Algolia records, use the method described in the advanced attributes configuration section.

The cartridge doesn’t support three or more levels of nested attributes or function calls in the attribute property.

Usage

To use this feature:

Add the corresponding Salesforce B2C attribute names (including nested attributes) to the Additional Product Attributes list in the Algolia Business Manager module Custom Preferences. Refer to custom attributes as custom.<attribute_name>.
Apply the Custom Preference changes and run a full re-index.

Check your Algolia index after the indexing job finishes. It should contain these new attributes in the product records now.

Advanced attributes configuration

Extend attributes declarations

The cartridge has an extension mechanism that lets you add new attributes, modify existing ones, or remove base attributes. Create a productAttributesConfig.js file in the int_algolia/cartridge/configuration/ directory. For an example, see this file in the cartridge code. This file must export a configuration object. The keys of this object are the name of the attributes in your Algolia records. Each key holds an object with the following properties:

Property Name	Type	Description
`attribute`	`string or function`	The attribute declaration. It can be: - The full path to the SFCC Product attribute name, separated by dots. - A function that takes the current product as parameter and returns data to be indexed in the Algolia record.
`localized`	`boolean`	(Optional, default: `false`) Should be true if the attribute is localized.
`variantAttribute`	`boolean`	(Optional, default: `false`) Only used for the product-level record model, to identify attributes indexed in the `variants` array.
`computedFromBaseProduct`	`boolean`	(Optional, default: `false`) Only used for the variant-level record model, to identify attributes computed once from the base product, and indexed into each variant.

The attribute property lets you declare attributes with their direct name (simple values like name, brand, etc.) or their “path”, for nested values with direct access. For example, you can access the revenueWeek attribute of a product’s activeData object by declaring activeData.revenueWeek. Other values are more complex and must be computed. For that, the attribute property accepts a function to define how to export such complex attributes.

/* productAttributesConfig.js */
module.exports = {
  /* ... */
  algoliaAttributeName: {
    /* Access directly a product's property (or nested property) */
    attribute: "activeData.revenueWeek",
    localized: false,
  },
  siblings: {
    /* Compute an array of variants IDs from the base product */
    attribute: function (product) {
      const variantIds = [];
      const variantsIt = product.variants.iterator();
      while (variantsIt.hasNext()) {
        var variant = variantsIt.next();
        if (variant.getAvailabilityModel().isInStock()) {
          variantIds.push(variant.ID);
        }
      }
      return variantIds;
    },
    computedFromBaseProduct: true,
  },
};

A more complete example is available in the cartridge code.

Any new attribute declared must be added to the Additional Product Attributes site preference in the Algolia BM module.You can also remove the base attributes by adding an empty declaration:

module.exports = {
    "in_stock": {},
}

The frontend uses the following attributes. Removing or modifying them can break the default UI:

id
primary_category_id
__primary_category
categories
variants

Post-processing customization

To customize your records just before sending them to Algolia, create a productRecordCustomizer.js file in the int_algolia/cartridge/configuration/ directory. This permits to do post-processing operations on the final records, after fetching all attributes. This file must export a function that takes as parameter the final Algolia record and does modifications on it:

/* productRecordCustomizer.js */
module.exports = function (productRecord) {
  if (productRecord.colorVariations) {
    productRecord["availableColors"] = productRecord.colorVariations.length;
  }
};

Configure extra attributes for `AlgoliaProductPriceIndex_v2` and `AlgoliaProductInventoryIndex_v2`

To extend the list of attributes sent by these jobs, select the job you want to configure in BM, go to the Jobs Steps tab, select the job step and edit the attributeListOverride parameter. Additional attributes can increase the job’s total runtime depending on the complexity of retrieving the attribute:

Nested attributes from B2C model classes take longer to retrieve than the product’s own properties
Calculated values take longer to retrieve than static values.

By default, these jobs send the following attributes:

AlgoliaProductPriceIndex_v2: price
AlgoliaProductInventoryIndex_v2: in_stock

If you extend the list of exported attributes, make sure to add the same attributes to the Additional Product Attributes site preference in the Algolia BM module as well so that the product delta export job also includes those attributes and updates them during its run.

Index active data such as revenue data

Indexing revenue metrics as an attribute in Algolia lets you use this data in your ranking formula. By default, the Algolia cartridge for Salesforce B2C Commerce supports many product attributes. Attributes like Active Revenue data, which aren’t supported, can be indexed in Algolia. Use the default behavior or specify a custom behavior in a configuration file.

Using the default method

You can refer to attributes in the Additional Product Attributes Custom Preference](/doc/integration/salesforce-commerce-cloud-b2c/getting-started/custom-preferences/) in the Algolia Business Manager module using the default behavior. For example, you can refer to activeData.revenueWeek to index revenue week data or availabilityModel.inStock to index the in-stock status of a product.

Declaring attributes in the configuration file

You can declare the desired behavior (for example naming, localization, nesting, or a transformation) using the extension mechanism.

In the extension configuration file (productAttributesConfig.js), declare each field with their path in the attribute property:

/* productAttributesConfig.js */
module.exports = {
  /* ... */
  revenueWeek: {
    attribute: "activeData.revenueWeek",
    localized: false,
  },
};

In your Business Manager, go to Merchant Tools > Site Preferences > Custom Preferences > Algolia. Find Additional Product Attributes and add all fields, such as revenueWeek.
Reindex your data by going in Administration > Operations > Jobs and running the AlgoliaProductIndex_v2 job.

Using a custom function

Using a custom function permits you to regroup the various fields under a common top-level attribute, such as revenueData.

In the extension configuration file (productAttributesConfig.js), add a revenueData declaration with the attribute property being a function that returns the data to index.

/* productAttributesConfig.js */
module.exports = {
  /* ... */
  revenueData: {
    attribute: function (product) {
      var activeData = product.getActiveData();
      return {
        quantityLast7Days: activeData.ordersWeek,
        revenueLast7Days: activeData.revenueWeek,
        quantityLast30Days: activeData.ordersMonth,
        revenueLast30Days: activeData.revenuesMonth,
      };
    },
    localized: false,
  },
};

In your Business Manager, go to Merchant Tools > Site Preferences > Custom Preferences > Algolia. Find Additional Product Attributes and add revenueData.
To reindex your data, go to Administration > Operations > Jobs and run the AlgoliaProductIndex_v2 job.

Integrations

Shopify

Salesforce B2C Commerce

Magento/Adobe Commerce

Commercetools

Bigcommerce

Default records

Variant-level default model

Product-level default model

Example

Configure product records attributes

Base attributes (non-configurable)

Configurable attributes

New arrivals facet

Flexible attribute indexing based on default behavior

Nested attributes

Usage

Advanced attributes configuration

Extend attributes declarations

Post-processing customization

Configure extra attributes for `AlgoliaProductPriceIndex_v2` and `AlgoliaProductInventoryIndex_v2`

Index active data such as revenue data

Using the default method

Declaring attributes in the configuration file

Using a custom function

Integrations

Shopify

Salesforce B2C Commerce

Magento/Adobe Commerce

Commercetools

Bigcommerce

​Default records

​Variant-level default model

​Product-level default model

​Example

​Configure product records attributes

​Base attributes (non-configurable)

​Configurable attributes

​New arrivals facet

​Flexible attribute indexing based on default behavior

​Nested attributes

​Usage

​Advanced attributes configuration

​Extend attributes declarations

​Post-processing customization

​Configure extra attributes for AlgoliaProductPriceIndex_v2 and AlgoliaProductInventoryIndex_v2

​Index active data such as revenue data

​Using the default method

​Declaring attributes in the configuration file

​Using a custom function

Default records

Variant-level default model

Product-level default model

Example

Configure product records attributes

Base attributes (non-configurable)

Configurable attributes

New arrivals facet

Flexible attribute indexing based on default behavior

Nested attributes

Usage

Advanced attributes configuration

Extend attributes declarations

Post-processing customization

Configure extra attributes for `AlgoliaProductPriceIndex_v2` and `AlgoliaProductInventoryIndex_v2`

Index active data such as revenue data

Using the default method

Declaring attributes in the configuration file

Using a custom function