Recommended For You!

You might also like, Here’s what other people ordered, We think you’ll like this, and last but not least: “Recommended for you!”. They are all highly recognizable phrases to anyone who’s used any internet service or store ever. But how do these recommendations work, why do firms do it and could we go against what’s recommended to us?

How does it work?

In a 2013 interview, Netflix’s Carlos Gomez-Uribe, VP of product innovation and personalization algorithms and Xavier Amatriain, engineering director explained how they are controlling what you watch. They say that the metadata behind their content allows them to find similarities between movies/shows. Next, they track you, everything you do: what you played, what you’ve searched for, how you’re browsing through their services and even when you’re doing this and on what device. At this point they connect you with content that they either outright predict you will enjoy or content that similar users have already indicated they enjoyed.

Why do firms do it?

The answer to this question is fairly obvious and relates to a recent 2014 article by Huang, Zhong and Yao on targeted marketing. The better a firm can analyze and understand their customers and how their products are perceived, the better they can connect the right customer to the right product, thus creating additional value. The customer value resulting from Netflix’s recommendation practices is also well known, not having to spend the time to search for what you want to watch is perceived as pretty valuable by customers. A recent article on Forbes.com they concluded a price increase for Netflix would actually result in significantly higher profits.

Can we go against the recommendations?

Short answer: Yes, but doing so will take more and more effort as time progresses. As Netflix improves their processes, they will understand your preferences even faster while also making their suggestions more automatic, thus alleviating even more of the manual actions you would otherwise have to undertake. So in the extreme event that your personal preferences make a full 180 degree turn, you will have to put in more effort than ever before to change the input Netflix has aggregated on you. Which I assure you will take considerably more time than just changing the TV channel.

Sources:

Wired.com, Netflix Algorithm, as approached on October 5^th 2014, http://www.wired.com/2013/08/qq_netflix-algorithm/

Forbes.com, The impact of Netflix’s price rise, as approached on October 6th 2014,http://www.forbes.com/sites/greatspeculations/2014/05/15/the-impact-of-netflixs-price-rise/

Huang, J., Zhong, N., Yao, Y., 2014, A Unified Framework of Targeted Marketing using Customer Preferences, Computational Intelligence, Vol. 30-3 p. 451-472