Recommendation systems(RS) have become ubiquitous in the past few years. All 10 of the most visited websites on the internet use recommendation systems of some form on their landing page. Given this prevalence, what are the most common types of use cases?
At their core Recommendation systems(RS) are matching algorithms that pick items from a large set and recommend them to an item from a second set. For instance, recommend the most appropriate videos among a massive set of videos to a user from the user’s set. Depending on the type of items in the first set, there are 4 main types of RS:
Content
Content is anything that has audio, video, text, or any combination of the three. Additionally, content could be short-form or long-form.
For instance, Tiktok or IG Reels recommend Short form videos with audio and/or text, Youtube recommends long-form videos and Twitter advises short-form text.
Currently, no platform recommends audio or long-form text. Building this kind of platform is not hard so it is most likely the case that there is no market for this.
People
People-based recommendation systems recommend people to other people. All the dating apps fall under this category. People are recommended based on their previous user interaction history and user preferences. Another platform that recommends people is Linkedin. It has a “people that you may know” and “people similar to” feature which recommends accounts.
Additionally, Linkedin recommends jobs to users. It picks the best jobs that it thinks the user will apply to. However, there is no platform built yet which recommends users to Jobs i.e. a platform used by recruiting teams to pick candidates based on their preferences. This could be an opportunity as the market for this is huge. Every company needs to hire people and it would be great if the recruiters could provide a few examples of what they are looking for and the RS takes care of the rest!
Products
E-commerce platforms like Amazon, and virtual stores like the apple store or the play store use recommendation systems to recommend products to users.
Ads
In contrast to using recommendation systems to market products in an in-house marketplace like Amazon, most companies market their products on a third-party platform like Facebook or Google as they can leverage the data they collected on users to show better ads. These companies, as a result, build RS that recommend ads to their users.