Anno 2017, almost all marketing gurus are shouting the same message: “Data should be one of your main drivers when determining your (future) business direction”. Unfortunately, it often just remains a good-looking quote and nothing more. Our Yearly Marketing Survey shows that currently only 25% of Belgian marketers personalise their marketing approach and a mere 11% have begun exploring the world of predictive analytics. For those who are currently looking for successful ways and tools to implement data in their strategy, the Netflix story may be a great source of inspiration.
Netflix, initially founded in 1998 as an online DVD rental company, scored its first win by offering customers the most efficient business model in the market for watching DVDs. Very quickly, however, by 2002, the company’s business model became obsolete. DVDs were no longer the main source for movie and TV show entertainment. To continue profiling them as the number one supplier in the long term, new strategies needed to be put in place...
This is where Netflix succeeded in dazzling their competition and consumers for the second time, by tapping into the unique world of “data-driven strategies”. Mainly thanks to this unique strategy, Netflix 2.0 - with the online streaming business as we know it nowadays - has been growing continuously for the past 15 years, going from fewer than 1 million subscribers in 2002, to nearly 100 million monthly subscribers in 2017.
So, which smart data practices did Netflix introduce?
The Netflix Prize: engaging and listening to consumers
One of Netflix’s most valued assets is its recommendation system. With ‘The Netflix Prize’, a datamining and machine-learning competition, the goal was to increase the accuracy of its movie-rating predictions by 10%, because accurate recommendations are one of their business drivers. Over 40,000 teams from all over the world entered the competition. The contest was a huge win for Netflix for several reasons: algorithmic innovation, brand awareness, the attraction of high-potential data engineers and the enhancement of the user experience due to the personalisation of its movie recommendations.
Learning? Take the time to sit down regularly with your clients and employees. You’ll be amazed by what information and knowledge you will gain for improving some of your key business drivers.
Netflix’s endless desire for more data
Consumer ratings are Netflix’s main sources of consumer data and personalisation opportunities. For many years, consumers rated movies and TV shows on a scale of 1 - 5 stars. However, the stars were a ‘misunderstood hero’. It was generally believed that when a movie had 4 out of 5 stars, this was the average score for all users. This was not the case. It was a personalised rating. You might give 1 star for Iron Fist, whereas your Super Hero-minded nephew would give it 5 stars.
Netflix came up with a “simplified” model. To rate a TV show, you only need to give a thumbs-up or a thumbs-down and they also provide each movie or TV show with a ‘...% match’. So you instantly know if it’s in line with your preferences. With this new system, Netflix claimed to have twice as many consumer ratings compared with the old rating system. A win-win situation: Netflix gathers more data, the consumer gets an even more personalised Netflix front screen, with better recommendations.
Learning? Always challenge the status quo. Some processes might give the impression that everything is working just fine (such as the 5-star ratings), but by digging deeper, your company can gain so many more insights and data by changing the current way of working.
Taking risks – the $100,000,000 ‘gamble’
The most characterising example of the brilliant Netflix data strategy is the political drama series ‘House Of Cards’, which was created based on millions of analyses. House of Cards tells the story of Frank Underwood (actor: Kevin Spacey), a ruthless, dissatisfied politician who tries to climb as high as possible on the political career ladder.
$100,000,000. One Hundred Million Dollars. This astonishing amount was the cost of producing the first 2 seasons of House of Cards. With each season containing 13 episodes, the average price tag of one episode was $3,850,000. Combined with the fact that this was one of their first experiments with ‘Netflix Original’ content, it is safe to say Netflix made quite a risky investment, right? Actually, they did not. Rather, it was a well-calculated decision, based on the data of millions of users.
Their user analysis showed 3 remarkable outcomes:
- Movies of David Fincher – the director of ‘The Social Network’ – were watched from the beginning to end.
- The British mini-series ‘House of Cards’ – created in 1990 – received good reviews from its viewers.
- People who liked the British House of Cards watched movies with Kevin Spacey and / or David Fincher.
Netflix saw and took the opportunity that this analysis showed, but only gave the ‘final go’ when both David Fincher and Kevin Spacey decided to participate. They ‘knew’ it would become a success, because the show covered the most important needs and wishes of their target audience.
They even went one step further to reach and enlarge their target group, by creating different trailers. Fans of Kevin Spacey saw trailers highlighting his presence, whereas female users watching shows with strong female characters were shown trailers with Claire Underwood (the wife of Frank Underwood). The ‘critical’ movie-watchers were shown trailers where they could clearly see David Fincher’s touch and style.
It is claimed that, worldwide, 3 million non-users became subscribers in 2013 due to House of Cards. This increase in consumers repaid Netflix’s investment in a couple of months.
Learning? Taking calculated risked based on a data-driven strategy, combined with always putting the needs of the customer first, can really pay off.
As the story above already proved, Netflix can offer you some great examples on how your entire business can take advantage of a data-driven strategy. However, just like any other company, Netflix needs to make sure that they continuously evaluate and improve on their previous decisions. Recently, Disney decided to launch its own online streaming service. With the removal of Disney from the Netflix platform, a huge potential to gather interesting data from families will be lost. More than ever, it will be crucial to implement the right algorithms and make accurate predictions…