Recommender systems are a powerful tool that is being used in order to predict the preferences of a user in order to make appropriate suggestions. I’ve met many entrepreneurs that are not aware that recommender systems exist, and this always surprises me. Why? Because if you have a B2C model, then 9 times out of 10 you could reap huge benefits from using a recommender system.

Let’s see some examples where they are being used:

E-commerce

Recommender systems are used extensively in e-commerce to suggest new items. A very good example of this is Amazon‘s “customers who bought this…” feature:

amazon recommender system

Example of Amazon’s recommender system in action

Movies

Netflix’s Netflix Prize is probably the most famous example of a service that employs movie recommendations. Netflix plays a special role in the history of recommender systems, as the famous  helped the field get attention and grow significantly.

Social apps

Facebook employs a recommender system to suggest new friends. Tinder tries to predict whom you will find attractive. Twitter uses a recommender to suggest interesting people to follow.

There are, of course, many other cases where a recommender system is appropriate.

recommender system simpsons

It is important to plan for a recommender system from day 0, since the correct data has to be collected and in the right way. This is the job of the data science architect for which I have written in an older post. Doing that makes sure the recommender system is working at its optimal capacity. This leads to improved user experience and user retention, which in turn translates into increased traffic and sales.

There are different types of recommender systems, such as collaborative filteringcontent-based recommenders and hybrid systems. Which one you should use? This is where it gets complicated, since the answer depends on many factors, such as the type of your business and the kind of data you are dealing with.

If you have any questions about how this can apply to your startup or business, just drop me a line. I’ve designed  and implemented many recommender systems throughout my career and I’m always happy to help.

 

 

Facebooktwittergoogle_plusredditpinterestlinkedinmail