The story of Netflix, which disrupted the movie rental business and perhaps caused the bankruptcy of Blockbuster and other video rental chains, may be well known to many. The more important chapter of the Netflix story involves the experimentation underway right now
The DVD-by-mail business, which was the core of Netflix when it started, is a tiny part of the company today.
Through video streaming, the company has steadily transformed itself into a leading Internet television network that could replace the broadcast TV business model of the last century.
Internet TV is on demand (time-shifted, place shifted), personalized (uniquely based on individual viewing habits and predictions based on many data elements including Netflix’s links to Facebook), delivered from the cloud to any screen, (device shifted) anywhere (subject to legacy legal restrictions).
More than 75 million subscribers pay their fees in more than 190 countries, and each day, these subscribers enjoy 125 million plus hours of TV shows and movies.
Reed Hastings, the co-founder of Netflix, has been a consummate experimenter. Multiple trials over time have allowed him to steadily fine-tune his business model for the media web of the twenty-first century. All of his experiments have been at the edge of the television and entertainment industry, but the mainstream industry leaders have never taken them seriously, until perhaps now.
While most others in the movie rental business were focused on top movie hits based on general lists, Hastings focused on developing a “recommendation engine” software that would learn an individual’s film preferences and use that information to suggest other appropriate titles from its catalog.
At first, the engine relied on individual ratings and preferences from previously watched films to predict user ratings for other movies. Later, it pulled together data from similar profiles of viewing habits.
To further ensure the superiority of the recommendation engine, Hastings introduced the Netflix Prize, an open competition in 2006 for anyone to beat the company’s Cinematch filtering and recommendation algorithm. He wanted the best technical minds to challenge the algorithm so that it would continue to be at the leading position.
Then he experimented with adding unlimited streaming (albeit with a very limited content library during the early years) to his DVD subscriptions in 2007, which allowed him to understand early on how people were embracing online streaming, despite all the limitations of bandwidth and unreliable cellular networks.
Today, Netflix is the undisputed leader in video streaming, and personalization is a big reason for that success. Each subscriber’s homepage shows groups of videos arranged in rows with a title that conveys the meaningful connection between the videos in that group. Netflix uses its filtering and recommendation algorithm to select and personalize the rows, the items for each row, and the order of the videos in each row for every individual.
Netflix is a leader in personalization, which is done by machines and at scale and speed. Hastings’ experiment to work with Amazon for cloud-based streaming (instead of owning this digital functionality) was a well-reasoned analytical decision: there was no way Netflix could compete against Amazon’s cloud competence.
By systematically analyzing the pros and cons of creating Netflix hardware (like a Roku or Google Chrome or Apple TV) versus focusing on software, the company ultimately settled on Netflix software to make streaming work across multiple devices—television screens, set-top boxes, Netflix-ready television sets, personal computers, tablets, phones, and others.
Whereas other television networks compelled executive producer Kevin Spacey to prepare pilot shows to test audience reaction to his proposed show, House of Cards, Hastings and his team were able to assess its prospects by analyzing the treasure trove of data about subscriber viewing habits. Whereas networks traditionally select shows based on a pilot, followed by a few shows, then an entire season, Netflix, using its data to mitigate any risk, was able to commit to two seasons up front—which allowed the producers and directors to be more creative. Based on the success of this show, Netflix has gone on to create other original programs such as Orange Is the New Black, Bloodline, and Marco Polo.
So, at the edge of the media and entertainment industry, Netflix had been experimenting seriously and serially with different building blocks for Internet TV. Most traditional incumbents such as Time Warner, Disney (ABC), Fox, NBCUniversal, and CBS, woke up late to the power of recommendation software that drives the personalization that has become the hallmark of Netflix.
These companies were also late to recognize the power of data-driven insights that allowed Netflix to get into original programming, and most did not understand Netflix’s technological superiority to optimize streaming for different devices, and across different bandwidths, from distributed cloud architecture.
Such sequential experiments, refined over time, allowed Netflix to become a global powerhouse in Internet TV. Television incumbents turned up late with their own ways to compete against Netflix; their versions of Netflix-killers through apps and alternative alliances. As you make sense of Netflix, you realize that the experiments undertaken by Hastings and his team were not one-offs; they were sequenced to marry developments in technologies with plausible options for superior customer value and different business models that place digital characteristics such as algorithms and analytics, as well as streaming and cloud, as key anchors.
As you try to make sense of the different experiments in the media and entertainment space, you may be wondering ahead to Phase 3: will the digital future of media and entertainment be defined by tech entrepreneurs like Reed Hastings and digital giants such as Amazon (Prime Video), Alphabet (YouTube), and Apple (iTunes and Apple TV)? This question about how the foundations established in the experimentation phase could drive the reinvention phase is not restricted to Netflix. You will find that linkage to be an essential thread of the Digital Matrix.
This is an excerpt from Venkat Venkatraman’s The Digital Matrix. You can purchase the book here.