The parents among you know that having kids implies going back to the school calendar for a while. So here I am a 35 years old working mom on Easter holidays. I am trying to take it easy this time and not to make too many plans. With one exception. Like for my children, my personal archetype of freedom is staying up late watching TV. Right on the very first day, after I put the kids to bed, I make a bowl of popcorn and situate myself comfy in front of the display. You got it, time for netflixing!
15 minutes later I find myself still zapping through the impressive artwork of truly unimpressive movies, wondering how did Netflix come up with the idea that I would ever watch that.
If you are on school holidays too and struggle with all the free time you get, here's an idea to spend your time better. Go to Twitter and look for a combination of Netflix, recommendation, and algorithm.
You will find hilarious examples of recommendation failure.
According to the company, Netflix currently has around 100 million subscribers, i.e. different paying customers. Some of them have kids and spouses, too. That makes probably around 250 million different profiles. And the magic of Netflix is, so they claim, that all of them provide a personal experience. You can read about the numerous technical ways, and really interesting ideas on personalization technology on the Netflix Techblog on Medium.org. But then again, so many people feel like me, and ask:
I find it striking, because Netflix, like most of California rockstar companies, is famous to pursue a customer-oriented approach, and an UX driven product and innovation strategy, as a philosophy. According to Todd Yellin, Netflix's vice president of product innovation, it's all about the customer. You can read about his approach and opinion on A/B testing and how the Netflix recommendation algorithm works in an article by Libby Plummer "This is how Netflix's top-secret recommendation system works". But why does the customer then tweet about instead of watching stunning movies on Netflix?
It seems that customer-orientation has its limits.
Companies often start with the goal to make the world a better place, what I call a big heart mission. Very often the idea of a new venture gains traction around a strong vision that would change the lives of millions of people. In the case of Netflix, that would be the availability of a good movie selection at home.
At the beginning of a venture, a founders team is usually in love with their customers to be. They ought to be. In the case of Netflix, this resulted in a super simple interface and the idea of a recommendation algorithm.
With time, however, business is gaining priority and the romantic relationship with the customer fades away for profit. In an interview for Fortune back in 2007, the company's founder, Reed Hastings, admits that he finds it "heartbreaking", but it has to be.
Netflix has turned 20 years old last year and is way past its big heart mission. In simple words, Netflix has exchanged movie fans for the average TV consumer. As Todd Yellin, puts it in the interview the goal is the subscription, and it apparently depends on the recommendation algorithm.
What is the alternative?
Currently, Netflix focuses on the recommendation algorithm because 80% of the views come from the recommendations. No wonder, if the movie availability is strongly limited. Whenever I tried to search for a movie I could not find it. So at one point, I gave up. In my opinion, the importance of the recommendation algorithm is a choice and a self-fulfilling prophecy. But showing the same movies in the recommendation bar for months does not make me like them more, even if I watch them. 80% of the views coming from the recommendation might as well mean that the variety of available movies and the usability of the search function are below customers expectations. In a customer-focused company, it might be a reason to focus on that first.
Zach Schonfeld showcases the risks and downsides of the company's approach that focuses on the customer in terms of growth, instead of focusing on the customer in terms of quality.
Ths is important because it is a pattern. The world will not become a better place if all TV viewers simply switch to Netflix. This will just replace the one thing with the other and keep the problem.
#machinelearning #algorithms #datascience #netflix #personalization #recommendation #bigheartventures