Recommendations initialization

The Recommendations initializer configures product recommendation features including personalized suggestions, frequently bought together, and related products. Use initialization to customize recommendation algorithms and data models.

Version: 1.1.1

Configuration options

The following table describes the configuration options available for the Recommendations initializer:

Parameter	Type	Req?	Description
langDefinitions	LangDefinitions	No	Language definitions for internationalization (i18n). Override dictionary keys for localization or branding.
models	Record<string, any>	No	Custom data models for type transformations. Extend or modify default models with custom fields and transformers.

Basic initialization

Initialize the drop-in with default settings:

import { initializers } from '@dropins/tools/initializer.js';
import { initialize } from '@dropins/storefront-recommendations';

await initializers.mountImmediately(initialize, {});

Language definitions

Override dictionary keys for localization or branding. The langDefinitions object maps locale keys to custom strings that override the drop-in’s default text.

import { initializers } from '@dropins/tools/initializer.js';
import { initialize } from '@dropins/storefront-recommendations';

const customStrings = {
  'AddToCart': 'Add to Bag',
  'Checkout': 'Complete Purchase',
  'Price': 'Cost',
};

const langDefinitions = {
  default: customStrings,
};

await initializers.mountImmediately(initialize, { langDefinitions });

Tip

For complete dictionary customization including all available keys and multi-language support, see the Recommendations Dictionary page.

Customizing data models

Extend or transform data models by providing custom transformer functions. Use the models option to add custom fields or modify existing data structures returned from the backend.

Available models

The following models can be customized through the models configuration option:

Model	Description
RecommendationUnitModel	Transforms recommendation data from Adobe Sensei including product suggestions, recommendation types, and tracking identifiers. Use this to add custom fields or modify recommendation display logic.

The following example shows how to customize the RecommendationUnitModel model for the Recommendations drop-in:

import { initializers } from '@dropins/tools/initializer.js';
import { initialize } from '@dropins/storefront-recommendations';

const models = {
  RecommendationUnitModel: {
    transformer: (data) => ({
      // Add custom fields from backend data
      customField: data?.custom_field,
      promotionBadge: data?.promotion?.label,
      // Transform existing fields
      displayPrice: data?.price?.value ? `$${data.price.value}` : 'N/A',
    }),
  },
};

await initializers.mountImmediately(initialize, { models });

Configuration types

The following TypeScript definitions show the structure of each configuration object:

langDefinitions

Maps locale identifiers to dictionaries of key-value pairs. The default locale is used as the fallback when no specific locale matches. Each dictionary key corresponds to a text string used in the drop-in UI.

langDefinitions?: {
  [locale: string]: {
    [key: string]: string;
  };
};

models

Maps model names to transformer functions. Each transformer receives data from GraphQL and returns a modified or extended version. Use the Model<T> type from @dropins/tools to create type-safe transformers.

models?: {
  [modelName: string]: Model<any>;
};

Model definitions

The following TypeScript definitions show the structure of each customizable model:

RecommendationUnitModel

export interface RecommendationUnitModel {
  displayOrder: number;
  pageType: PageType;
  title: string;
  items: Item[];
  totalProducts: number;
  typeId: string;
  unitId: string;
  unitName: string;
}