Example Solution E-commerce Froomle can be used to personalize every aspect of your online interaction with your customers. Some examples are: Personalized e-mail campaigns A personalized homepage Personalized search results "You may also like" on product detail pages Froomle incorporates as much of your data as we possibly can. The following is an overview of the required steps to set up an integration tailored to your retail use case. There are three major steps: Metadata delivery Real-time tracking of browsing events Requesting recommendations Each will be discussed in detail on the page below. After reading this you should have a clear overview of the steps required to integrate Froomle and some inspiration for your first use cases. Metadata delivery In an e-commerce setting, Froomle needs at least your product catalog and up-to-date stock information to guarantee an optimal experience. Froomle provides an SFTP server to upload periodic data dumps. Reference information on the SFTP can be found here: Reference - Data Upload (SFTP). If the SFTP server cannot meet your needs, please contact the Froomle support team to discuss alternatives. Froomle refers to your products as items. For a complete overview of how Froomle deals with metadata, see: Concepts - Items. Product catalog Your product catalog is used to apply certain business rules, such as: "Never recommend products under €20". Additionally, it is also used to interpret results during analysis, so that your recommendations can be improved further. Typically a product catalog contains at least the following information on products: Product ID Category Tags Brand Price Product model The metadata you deliver to Froomle is by no means limited to the examples listed above. A product catalog is typically updated once a day. Stock information Stock information is essential to Froomle to decide what to recommend: Nobody likes being recommended a great product that turns out to be out of stock. Stock information is typically kept at the variant/size level, and not the product level, e.g. there are 10 sizes S and 15 sizes M of a certain t-shirt in stock. Even though Froomle identifies products at the product level, this is not an issue. It is however important to also supply a product identifier in a stock export, next to the variant identifier. A typical stock export contains: Product ID Variant ID Amount in stock online Real-time tracking of browsing events Froomle’s algorithms use your users' interactions with your website to make recommendations. We refer to your users' interactions as Events. These events are communicated to Froomle using the Events API. This is a simple REST API that can be integrated both front-end or back-end depending on your needs. Events for E-commerce Froomle supports a broad collection of event types to capture specific user behavior. To provide the best recommendations possible, it is recommended to record user events for all event types. The following table describes the priority of the different user event types. The highest priority of user events must be logged to obtain quality data models. Priority User Events Required for initial setup detail_pageview Important for model improvement item_interaction Important for model improvement (e-commerce) purchase add_to_cart remove_from_cart Required for analysis (KPIs) impression click_on_recommendation batch_open batch_click_on_recommendation batch_impression page_visit session_click_on_recommendation session_impression search_click_on_recommendation search_impression Example While we have listed all information required to integrate Froomle’s Events API, bringing everything together in an example use case might be more familiar. For such an example see: Integrating events on an e-commerce platform. Requesting recommendations Recommendations can be requested using the Froomle Recommendations API. For a full overview of the API specification see: Reference - Recommendations API.