Product Recommendations
A recommendation engine applies machine learning to visitor interaction profiles in order to surface the most relevant items (content or products) to each customer in their journey. Its primary aim is to increase cross-sells and up-sells in a retail environment, or engagement in the case of content, by helping customers find what they need or new items they would not have otherwise found themselves.
Using our visual Campaign Designer tool, you can easily do the following:
- insert a recommendations component (a widget with recommended items) on relevant page(s)
- configure the number of slots, information shown per item (name, price, etc.) and the widget's style
- choose the appropriate Machine Learning model and apply any suitable filters to the recommendations
- target audiences with the most suitable recommendations styling, positioning, and strategy.
- understand ROI and recommendations performance through our in-depth campaign reporting
A recommendations campaign is not always about products in a standard retail scenario. For this reason, we mostly use the term item recommendations. An item can be a product or a piece of content, such as an article or a blog post. An item can be characterized by the following elements:
- a unique Item ID that identifies it
- a name for the product or a title for the piece of content
- a URL to which a visitor can go in order to view details and/or purchase the item
- a category for the item (could also have subcategories)
- a thumbnail picture URL so that our platform can show the item's picture in the recommendation slot
- additional information such as price, stock level, brand, style (in retail), or any other item attributes may or may not be present.
If the above elements are to be used for filtering recommendations, then they should be contained in an Inventory File and imported into our platform. This step is a part of the one-time setup that enables you to use the solution.
There are two versions for the recommendations feature: Basic and Advanced. The main difference between them lies in the number of items that can be used, and the range of Machine Learning algorithms that are supported. For a detailed comparison, see the table below:
Basic | Advanced | |
---|---|---|
Inventory Size | Restricted - up to 2,000 items | Up to 200,000 items |
Algorithms |
|
|
Filters |
|
|
Note: To create a Recommendations campaign you need to complete the one-time setup and then use Campaign Designer to create, configure, and publish your campaign. See below for further details.
One-time setup tasks for Product Recommendations
Two essential tasks are needed to get this feature up and running. The first one lets our platform see all the items that are available for recommendations. The second tells us how to track the items (Item IDs) that your customers view, the ones they end up buying (in retail), or the ones they like, favorite, email, or share on social media. For more details on these tasks, see the following pages:
- One-time setup task: inventory file (importing, scheduling, and mapping)
- One-time setup task: visitor behavior tracking
Creating a recommendations campaign
After completing the setup tasks, create a new campaign. The variants of this campaign will act as recommendation widgets.
To configure your campaign:
- Navigate to the Content page and choose the page and position you want to use.
- Select a Div element you would like to use for your variant.
- From the Components panel on the right, insert a recommendation widget either above or below that element, or you may even replace the element.
The following video shows how to insert a recommendation widget below the hero banner on a landing page. For this example, we use the following demo site: https://sale.maxymised.com/
.
After having defined all the URLs in which you want the recommendations to display for the particular position, you can define the recommendations strategy and style your widget.
Note: When applying a strategy and configuring the layout, the order of what you do first is not crucial and can be changed around.
To set up the layout and style, you need to decide the following:
- the number of items you want to display (in the recommendations component, items appear as slots)
- the type of pagination if there are more items that can be seen on the page
- the information you want to include for each item (for instance, in a standard retail scenario, apart from the image and title of a product, you may want to show the price and style of the item)
The item characteristics should always be available and up to date in the inventory file. You can style the component using Campaign Designer's extensive customization tools. The following video shows you how to configure a layout:
A recommendation strategy is composed of the following:
- a suitable machine learning model
- any filters that you apply to the recommended items
Filters are not required by default and often are not needed, such as, when you want to simply display the most viewed items in a particular listings page. However, a model is always needed. There are a variety of models you can use. The following video shows you how to configure a strategy:
Before performing QA and publishing your campaign, you can choose who sees which variant in the targeting step. For instance, you may choose to qualify the audiences that enter the recommendations campaign in the first place. For more information on targeting, see the following pages: