Introduction to Recoms
Hello Retail Recommendations provides a flexible system for providing personalized product recommendations. The recommendations are often used on the various pages in a webshop. But can also be used in other places through the REST API. The recommendations can be configured to take the current context of the visitor into account when finding products. This could for instance, be the product or category the visitor is currently looking at or the current contents of the cart. Notice that the recommendation system is built to be able to load more than one product recommendation set at the same time. The system will ensure that products will not be included more than once in a set of product recommendations that are loaded at the same time.
Managed Product Recommendations
When using the Managed Product Recommendations the product recommendation instance must be configured in the Hello Retail backend. This configuration decides how the product recommendation will behave. Among other things you can choose the algorithm for selecting products and how many products to return.
The managed product recommendations can further more be used in two different ways. You can either make Hello Retail render the recommended products as HTML or just return the product data as JSON. The HTML will be rendered using a liquid template that can be configured in the Hello Retail backend.
In the backend you will also find settings that will make helloretail.js automatically inject the product recommendations into your site. This works by configuring a CSS selector in the backend. helloretail.js will look for the first element on your page that matches this CSS selector on every page load. If a matching element is found the product recommendations will automatically be rendered and injected into that element.
If you are not using the CSS selector to automatically inject the product managed recommendations you can do it your self by calling the
loadRecom SDK method from your website.
When using the managed product recommendations you need to identify the recom configuration from the backend. This is done using the product box key. This be found on on My Hello Retail.
General product recommendation box support articles, like algorithms & general setup, can be found here.
Unmanaged Product Recommendations
Unmanaged Product Recommendations can be used without configuring anything related to product recommendations in My Hello Retail. Instead you pass information about what algorithms to use to recommend products, and how many products to return to the API. This type of product recommendations can only return the JSON representation for the recommended products.