As consumers we are “drowning” with choices. This is especially true when it comes to online shopping. Take for example Amazon, if you were to look up any item on Amazon you are likely to get back hundreds if not thousands of choices. A quick visit to your local grocery store to buy a bottle of olive oil could leave you dizzy with the choices you see. The digital media market is not any better, there are over 30 Million songs on Google Play, and more than 10 thousand titles of movies and shows on Netflix. In short, we are really lucky to have so many choices but it comes at a cost, selecting the right item you are interested in. The figure below shows some stats on the plethora of products/choices some retailers have.
As consumers we are “drowning” with choices. This is especially true when it comes to online shopping.
Believe it or not, this is more of a problem to the retailers than it is for us the consumers. Retailers are interested in showing you things that you are likely to buy. To achieve this retailers have to comb through massive amounts of data to show you what is only relevant to what is it that you are looking for. If you are on Amazon shopping for men’s clothing they better not show you clothing for infants or night gowns. If you are into listening to classic music then Pandora better add classic albums to your stream. The more relevant the stuff you see from a retailer the more likely you are to buy, and to keep you engaged longer with their site. The figure below shows Amazon’s home page for one of Analytica’s Data Scientists. As you can see, Amazon is recommending Data Science related books to him on the top, and since he had shows some interest in sports clothing they are showing him Under Armor shirts. The stuff in the middle about the kitchen was recommended because his wife had used his laptop to view some kitchen items!
But how do they do it, how can Amazon tell what items I am likely to buy, or how does Netflix know which movies to recommend to me that are actually relevant to my taste. Enter the Recommendation Engine Algorithms!
Recommendation Engines and Decision-Aiding Algorithms
Recommendation engines are a type of Machine Learning algorithms that basically predict how a user will like, or not, an item. The item could be a song, a movie, a clothing item, a trip destination, an investment portfolio, a car, or pretty much anything else. These algorithms put a quantitative value on the likelihood of the item being liked. A retailer will then only show items with the highest likelihood of being liked by the given customer.
But you will then ask, how do they do that? Well, the math is pretty complicated, see figures below for some simple examples 😉 
But if you are asking on how they really do it then here it is. You will be surprised how much data about you is being collected. You might be surprised to know that there are companies whose sole purpose is to collect data about consumers and sell this data to retailers. Data about you is constantly being collected. If you use Facebook, or any other social media platform, you are being data mined, if you use your credit card to buy stuff, this data is collected and you are being data mined, if you use the cute key chain tag retailers give you, you are being data mined, if you google something, you are being data mined. In short, if you interact with something that is connected to the internet, you are being data mined. Some retailers collect this data and use to build a user profile about you the consumer. Things you like, things you don’t like, your demographics, age, etc. your political affiliations and so forth. They also have lots of data about the items they are selling of course. The Recommendation Engine algorithms they employ will find patters in this massive amounts of data to make the recommendation for you. These algorithms learn about your past behavior/purchases/likes etc. They find users who are similar to you to see the things they liked, they find similar items in their databases. The algorithms will then make the decision on which items to recommend for the user.
You can learn more about Recommendation Engines by reading our presentation “Recommendation Engines, the Rise of Decision Aiding Algorithms“
If you would like to learn more about our work, or if you would like to get our free demo please don’t hesitate to contact us.
[contact-form][contact-field label=”Name” type=”name” required=”true” /][contact-field label=”Email” type=”email” required=”true” /][contact-field label=”Website” type=”url” /][contact-field label=”Message” type=”textarea” /][/contact-form]