The Definitive Guide To Recommendation Engines
What are they? Why are they used? And why are they so popular & More!
INTRODUCTION
The good news: the internet is full of products and services that may be exactly what you are looking for.
The bad news: the internet is full of products and services that may be exactly what you are looking for.
How can you make the right choices without exhaustively searching through all options? This is where recommendation engines come into play.
Recommendation engines are data science tools that predict which products or services are best suited for each individual customer.
What are the important things to know before successfully adopting recommendation engines in your business?
We will see that over the next ten chapters of this guide.
By the end of it, you will be able to effectively communicate about why and how a recommendation engine should be added to your customer interface.
We will also see which companies and applications champion the recommendation engine scene.
Chapter 1: What is Product Recommendation?
A product recommendation is an item that you proactively present to your customer because you believe they are likely to buy.
Product recommendations do not need to always lead to a purchase. They are successful if they recommend interesting products.
In fact, having 100% successful recommendations is bad; it means that you are playing it too safe and are missing out on potential preferences of your customers.
Chapter 2: What Is A Recommendation Engine?
A recommendation engine is a software system that outputs product recommendations.
It does so by processing data generated by the customers, building a predictive model of their behaviour and outputting the items most likely to be desirable. Recommendations are not restricted to products. They can refer to services, content such as movies on Netflix and connections on social media.
One example of a recommendation engine is Amazon's "Customers Who Bought This Item Also Bought" feature. This feature shows customers items that other customers who bought the same item also bought.
Chapter 3: What Are The Types of Recommendation Engines?
There are three main types of recommendation engines: collaborative filtering, content-based filtering – and a hybrid of the two.
Collaborative Filtering
Collaborative filtering makes recommendations to a user based on their similarity to other users. Its main element is the preference matrix, where each user is a row and each product is a column:

If you are user E, then probably you don’t want a TV, because users B and C are the most similar to you and they also don’t want a TV.
An advantage of collaborative filtering is that it doesn’t need to analyse or understand the content (products, films, books). It simply picks items to recommend based on what they know about the user.

Content-based filtering Collaborative filtering makes recommendations to a user based on their past preferences. If you liked the Squid Game series, then you will probably also like the Parasite movie.
To make recommendations, algorithms use a profile of the customer’s preferences and a description of an item (genre, product type, colour, word length) to work out the similarity of items.
The downside of content-based filtering is that the system is limited to recommending products or content similar to what the person is already buying or using. It can’t go beyond this to recommend other types of products or content. For example, it couldn’t recommend products beyond homeware if the customer had only brought homeware.

Hybrid Model
A hybrid recommendation engine makes recommendations based on both the users’ similarity to others and their past preferences.
For example, one can generate a recommendation with both techniques and choose to show one of the two or both of them.
Netflix is the perfect example of a hybrid recommendation engine.
