Get more out of your product recommendations — Part I: Set your aim

Avoid manual recommendations

Unless you have very few products and a static demand for those products, manual product recommendations are time consuming, static and often less relevant for the visitor. It is impossible to combine realtime store information with manual recommendations to deliver up-to-date relevant product recommendations. Displaying relevant alternatives for a product may be viable, but thats about it. It can’t be derived specifically for the individual visitor based on browsing history, prior purchases or items added to cart. And it quickly becomes unviable to manage a large set of products in this way.

The use of rule based recommendations

Rule based recommendations are derived out of rules provided by you (or some default setting). Examples can be to show products in the same product category, best sellers, products on sale, etc. These can be beneficial in cases where there is no or very little information available on the visitor. For example on the front page when the visitor just arrived. The issue with these, apart from being somewhat static, is that they generally drive more of the same sales. E.g. Best seller recommendations tend to be successful for a while because they promote the most popular products on the site, but they rarely introduce a new purchase pattern on the site. Furthermore, these recommendations may not present the most relevant products for the individual buyer. Overtime they also tend to become much less relevant for returning customers, since it is likely they already seen them and neglected them, or bought them.

AI opens a world of opportunities

What if we were to introduce logic and reasoning into the selection of product recommendations to display in any given moment? We could aim to show products based on:

  • Purchase behaviour of similar visitor profiles
  • Products that are often bought together with the product in question
  • Products that have been added to cart, viewed, searched, etc.
  1. Guide the visitor to the product they are looking for
  2. Make the visitor aware of products that could be of interest to them



Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Alexander Ekdahl

Alexander Ekdahl

AI and E-commerce enthusiast. Founder of We aim to help the world make use of its data.