Product recommendations are a useful mechanism to boost your online sales by offering your other customers products in your catalogue that they might also like, or that are bought together.
Let’s take as an example, a store which is selling perfumes.
When a customer visits a page of a perfume, a good practice is to show your customer, similar perfumes, namely perfumes from the same brand or targeted for the same gender or with a related smell.
Another good type of recommendation is to suggest other products he could complete his order with. What this means is that, it is a good idea to recommend products that are frequently bought together as bundles in order to increase the volume of purchases.
At Jumpseller, you’re able to make use of these 2 types of recommendations.
1. Related products
We use Artificial Intelligence to analyse information about your product catalogue and detect which products are semantically more related.
We do this by processing the information that was filled up by you when creating or editing a product, namely the product names and descriptions.
We also take advantage of the other thousands of customers using Jumpseller, to analyze the sentences and words they used to describe their products in their stores in order to make the recommendation system more robust and accurate.
For instance, let’s take the initial example of a store which is selling perfumes.
By analyzing the product name and description we know that keywords such as “Boss”, “Hugo”, “Mujer”, “Perfume”, “rosa”, “frutal” and some others help characterize this perfume.
This way, other perfumes characterized by the same or similar words will be recommended when visiting this product.
If no good recommendations are found by this method, the default is to recommend products within the same category.
In this sense, you have some control over these types of recommendations, since the more accurate and descriptive you are when editing your product, the better the similar product recommendations should be.
The recommendations are usually shown below the product information at the bottom of the product page, but that behaviour can be changed by editing the theme.
2. Frequently Bought Together products
This type of recommendation is currently only available for enterprise customers.
We use an algorithm to process the recent order history from your store and check which products are bought together so that the next time someone is buying a certain product we can verify which products are more frequently bought with it.
We also take into consideration products that are not bought together at all, and for that reason are less similar in this recommendation case. Taking our previous example, look at the image below.
In this scenario, you have two perfumes being bought together in the same order. If this happens more frequently we can assume that the next time someone buys “PERFUME GOLDEN SECRET VARON EDT 200ML”, the “Perfume The One Hombre Edp 150ml” is a good recommendation to be suggested.
As you can see, the more orders you have in your store history, the more accurate this type of recommendation will be.