Netflix Case Study | Audio Transcript

Hello. This is Professor Michael Rappa from North Carolina State University in Raleigh, North Carolina, and I’m here today to speak about my course, Managing the Digital Enterprise.

If there is one digital enterprise I’ve made a point of watching in recent years, it’s Netflix, the online DVD movie subscription service, which provides us with a wonderful illustration of a very innovative and creative use of digital technology wrapped around a business model that has really made it a competitive force in an industry dominated by traditional players such as Blockbuster here in the U.S. The Netflix story is a wonderful entrepreneurial tale of an individual by the name of Reed Hastings who, perhaps like countless others, had grown extremely dissatisfied with the established predominant business model in movie rentals by the early to mid ‘90s, which was really the storefront neighborhood movie rental store that had become dominated by Blockbuster here in the U.S. as well as a few other competitors.

And the model there was really predicated on a kind of per video rental fee that was wrapped up with the need for late fees for the return of movies past a specified deadline, whether it was three days or five days. It’s important to recognize that the late fee was an important part of making that process work, keeping people on time in bringing back the videos, but it also represented a not insignificant portion of the overall revenue of that business model. In some ways you could argue that it was a business model predicated on human failings, our inability to keep to a schedule sometimes because life intervenes.

And, as the story goes, Reed Hastings, as he’s recounted in numerous interviews, found himself caught with a rather substantial late fee on a video, and having been a successful entrepreneur already came to the conclusion that there had to be a better way. And that was the genesis of the idea for Netflix. And I’ll refer you to some of the interviews that are readily available online, both through Digital Enterprise and other sites, so that you can learn more about just what provides that spur to an entrepreneur.

But by 1997 Reed Hastings had founded Netflix. And the basic notion was to create a club of movie viewers who would pay a kind of membership fee to be able to borrow a certain number of videos in a given time frame. And over the years it kind of evolved very quickly, especially as it kind of came to realize that there were many more people than he might have originally expected who were interested in exactly such a service. And over a period of a few years it evolved into a very clearly defined subscription service whereby the customer would be allowed to borrow from the collection a set number of DVDs. I think initially it started out as three, although the product mix is now varied. And the customer could hold on to those DVDs for as long as they wished, and when they returned one, they would receive in turn the next movie on a queue that they created over the Netflix web site.

And so the web and sort of the digital presence online became a very important, integral part of how Netflix reached its customers, enabled them to queue up movies that they wanted to see, and then as they returned movies, they could see the next one on their list, receive the next one on their list. And the subscription model basically set a given fee, monthly fee, and there was no need for any kind of late fees. The customer could hold on to the videos as long as they wanted. They could churn the videos as quickly as they might could. And whether they viewed three movies in a month or however many more they could fit into a kind of cycle of receiving and returning the videos, the fee was set.

Well, clearly Reed Hastings had caught onto something, and by 2002 he made the first public offering of shares in Netflix, raised almost $100 million. It was already nearing profitability by that time. They grew the collection of DVD titles into over 13,000 titles. Today it numbers in the tens of thousands. And by 2003, they had already reached one million subscribers, with a rapidly growing base. By 2005 it was already over three million subscribers.

So one important point about Netflix from the perspective of business models is that it took a well understood, common business model of subscription and took advantage of the technology around the Internet to take subscription into a realm where it really was not the prevalent model in the past, and has used it with great success. So it’s just one of those beautiful examples of someone taking a well-understood business model and leveraging technology to apply it in a new area where it hadn’t been used before. And of course the Internet is what greatly enabled this process. It’s not that it couldn’t have been done before the Internet, but that the Internet just made it that much more possible and increasing the likelihood of success.

Of course, the brilliance behind what Reed Hastings had done by recognizing that late fees were something that customers were generally annoyed by and didn’t appreciate, even though they might have been necessary to make that Blockbuster rental model work, that by focusing on the elimination of late fees, Netflix had a kind of ready-made lever on which to advertise the service which was going to resonate with the consumer base, especially a customer base which was becoming more and more web-enabled over that period of time.

So a lot of fortuitous things came together that made Netflix an increasingly attractive option to users. And I think one can have a very kind of long, fruitful discussion over what is it that customers want and what are the tradeoffs between, say, a Netflix approach of receiving DVDs through the mail versus the ability, if possible, to go down to a neighborhood rental store and pick out the movie that you want. And there are advantages and limitations on either side of that equation, but certainly when you look at Netflix you see some very appealing advantages, and perhaps most notably, going back to an article you would have read called “The Long Tail” by Anderson, the ability for an online business like Netflix to create a much larger catalog to select from than what’s possible in terms of what you can do in a neighborhood storefront in terms of how much inventory you can collect.

Whether it’s Netflix or iTunes or Amazon.com, the ability to create a very large catalog, with tens of thousands of products to select from, in this case movies, versus a storefront, which might only be able to collect together a couple thousand, and the tradeoffs involved with that, the experience of going to the local video store and not being able to find the movie that you wanted to view because it’s already been taken out, and a lot of the issues surrounding that, not having a very deep catalog to choose from overall.

Netflix really opened the door in a number of different ways. And there are so many facets of Netflix that one could highlight that are interesting as a digital enterprise. But I’d like to focus your attention foremost on the Netflix use of what are generally called recommender systems to help its customers select movies to view.

Now, before I get into this, let me just say that this is a very nice exercise that one can do on their own in terms of signing up for a trial subscription for Netflix and receiving movies and then having a chance to log in to the web site and see firsthand how the recommendation process works within Netflix. Recommender systems are a very important facet, I think, to how the whole Netflix model works, and so it’s worth some time exploring it and understanding its advantages in terms of helping customers choose movies.

Generally speaking, we all have an opinion about the movies that we watch, and so Netflix tries to capture these opinions through customer feedback. So each movie that you check out, or really any movie that you’ve seen, Netflix is constantly asking you to give that movie a rating. And it’s collecting those ratings for each of its customers into a database that uses some very powerful algorithms to cluster similar movies together and then analyze as how customers have rated them. And the purpose of this is so that customers who’ve given similar ratings to the same movies kind of cluster together and then are kind of matched with likeminded viewers.

So the basic idea is to use each customer’s opinions about the movies that they watched to understand what similarities they may have to other customers. And as that process moves forward, the technology looks at those clusters that you’ve rented from in the past and determines which titles you’ve yet to rent, and then recommends films that have been highly rated by other viewers matched with you that you have not yet seen. So that’s a very sort of simple description for what is otherwise a very kind of powerful algorithm-based approach to clustering people in ways that would hopefully yield some similarities in terms of what they’re likely to enjoy when it comes to watching a movie.

Of course, with millions of subscribers and tens of thousands of movies and literally hundreds of millions of movie reviews and evaluations, Netflix has to be able to crunch a lot of data and to do it very quickly in order to constantly feed its customers recommendations of movies that it might want to see. And so, although from a user point of view it all seems very simple and straightforward, I go to my Netflix account and each day I see different movies potentially that I’ve never seen before that there may be a good chance I might like to watch, that takes an awful lot of capability on Netflix’ part to make happen each day.

You can see this at work for yourself simply by signing up for Netflix, even if it’s just on a trial basis, and rating some movies. Usually a new subscriber is asked to rate a particular set of movies which probably provide a good sort of tool for differentiating a user or at least helping to get an initial clustering sort on where they might exist. I think it’s fair to say that some movie viewers probably cluster with other large groups, they have very common tastes in the kinds of movies they like to watch, but that the ability to predict for any given customer is kind of going to vary over a distribution. And there are probably a handful of customers out there in any pool of three million subscribers who are going to have extremely eclectic or different tastes in movies and so the algorithm may or may not work quite as well for them. But one can simply give this a try and see.

Along with its computationally based system for making recommendations, Netflix, of course, provides ample opportunity for its subscribers to provide feedback in terms of reviews of movies that they’ve seen and does all the kinds of leading edge things that we see on some other major web sites in terms of allowing those particularly energetic subscribers who love to provide feedback to web sites about products that they can really do that to their heart’s desire. And the goal here on Netflix’ part is to do everything it can to turn its customers into regular ongoing movie viewers. And so this is really the key point in terms of Netflix’ ability to succeed. It needs to do everything it can to cultivate a subscriber base that becomes regular movie viewers. That’s how they retain subscribers. The goal is to keep people excited and interested and wanting to view movies.

And so much of Netflix’ success, then, really hinges on this issue of how do we continue to feed our customers movies that they are going to enjoy and to keep them as interested subscribers into the future. Because subscribers are paying a set fee per month in advance, then this keeping them engaged in movie viewing is an important part of the equation that a customer is going to have in terms of whether they’re receiving value for that service. And it’s like any other subscription service. If you’re not reading – if you’re a newspaper subscriber and you’re not reading your newspaper every day, after a while you start going through that economic calculation and may decide to no longer subscribe. And so the same thing with Netflix. It has to cultivate a subscriber base of people who are active movie enthusiasts and viewers, people who want to see movies. And so much of, I think, that real competitive sustainable advantage for Netflix revolves around this recommender system and their ability to constantly help people select movies that they might want to see.

Now, of course, there are other means by which people choose movies, based on the recommendations of their family or friends, things that they’ve read about or seen in other ways, and so that whole issue doesn’t just revolve around Netflix itself. But I think Netflix can go a long way in terms of encouraging viewership of movies, especially by making available to its customers movies which they might not otherwise have come across, and yet, at least based on their algorithm, may be something that they might enjoy watching.

So as I’ve said, the quality and the ability of Netflix to make recommendations is going to depend a lot on the amount and quality of the data they collect about the user preferences on movies. I’m sure Netflix looks very closely at how successful they are in terms of recommendations they make and whether their users take up those recommendations and to what extent their users are happy with the movies that they watch flowing from that recommendation system.

If you think about it, today Netflix claims it has some 500 million movie reviews in its system by which it analyzes user preferences. This is really quite an amazing feat when one thinks about the longstanding dominant competitor in the movie rental business, Blockbuster, understood really far less about its customers and what their moviegoing habits were. I mean, sure, Blockbuster knew how many videos it was renting. It knew sort of what were the most popular movies, as would any movie theater would understand that in terms of the data collected. But Netflix is really capturing not just what people are watching but coming around and grabbing whether or not they enjoyed those movies, and really understanding something about moviegoers’ preferences.

And one can only sort of sit back and think about the incredible value that Netflix is creating by developing that kind of detailed empirical understanding of its customers’ preferences. I dare say that Netflix may understand more about moviegoers and their preferences today than even those folks who produce movies and distribute movies. They really have a very tightly coupled feedback loop on what people are watching and what in fact they’re enjoying or not enjoying. And so that makes the Netflix process a very interesting one to watch. And people have already observed how certain movies which didn’t get a broad release in movie theaters, were not pushed or advertised very much, have all of a sudden come to the surface and become movie hits on Netflix because they did find an audience, and that audience grew because of what Netflix was able to do in making recommendations.

Indeed, when you think about the countless tens of millions of dollars that the movie studios spend on advertising to generate an interest in a movie, Netflix has really sort of built that into the whole process within its subscriber base of being able to, in a sense, advertise movies in a very almost costless way. So it’s on this key front that Netflix has a really tremendous lead in both developing the technology to do this and also collecting the data that drives it. And when we look at Netflix, and we look at a lot of the discussion around how much larger competitors are going to come into this game and put this relatively small fledgling organization into a kind of difficult competitive straits, I think one has to sort of sit back and look at what Netflix has done and look at the tremendous investment that they’ve made and what they have which really no one else has, which is now this huge database of peoples’ movie preferences that it can continue to grow and leverage and cannot be replicated by some competitor in a short period of time, no matter how large they are.

With Netflix we see really a beautiful combination of taking a new business model into a product market segment and exploiting not just the technology of the Internet but digital technology in such a kind of powerful and creative way to build a sustainable competitive advantage that I think we’ve already seen competitors trying and finding it to be difficult being a follower in this business, no matter how much money you may have to spend. One can simply point to the example of Wal-Mart, which came into the online DVD rental business late but yet had enormous financial clout to explore this business, and after a relatively short period decided to throw in the towel and to stop its online DVD rentals and turned around and negotiated a partnership deal with Netflix to sell DVDs via the Netflix site.

So one has to sort of take a serious look indeed at Netflix as a case study and understand how in fact a sustainable competitive advantage can be built by tying together a business model and digital technology in a way that’s really hard to replicate. This is not to say that the competition is by any means over. The movie rental business is upwards of a $10 billion business in the United States. It’s still heavily dominated by the storefront rental market that Blockbuster has cornered, although is losing money or has been losing money on for a little while now. But we still have a long way to go with this. I just think that what Netflix has done is to kind of capture a sustainable early lead in the online segment of the rental business.

If anything, the increased competition will lead to benefits for consumers. We should see, I think, a lot more creativity coming out of companies such as Netflix as well as new companies to emerge. There has already been some movement by companies to pursue other kinds of business models in the online DVD world. One company by the name of Peerflix is a good example, and I’d suggest you take a look at that one if you have a chance, taking a different approach, although one which seems to have some serious promise to it. I think we’re going to see Blockbuster move more and more aggressively into online DVD rentals, and so they’re by no means out of this competitive game. And I think we’ll see sort of greater product variation. We’ll see, I think in the case of Netflix already, efforts to move toward that kind of community model of using now its large subscriber base in a way that gets them not just sharing their movie reviews with each other but becoming much more directly involved in making movie recommendations to each other, to networks of friends and so forth much more easily.

If you take a close look at Netflix, I think what you’ll see is efforts for them to help foster a customer’s network of friends for kind of sharing in a kind of more private way movie recommendations and suggestions. So I think now that just like we’ve seen with other cases, and maybe the most notable one being like an eBay, where when you have a large base of customers, then one has to kind of find fruitful ways to turn that customer base into a community of people who see themselves as kind of sharing views and opinions and becoming kind of loyal subscribers to a service because their network of social connections is kind of embedded in that process as well.

Lastly, I think as we go forward we’re going to see more sort of technological evolution in the world of video as homes get access to things like video on demand via their cable companies. We’re going to see this business have a kind of portfolio of different competitors taking different approaches to how to reach the household with entertainment. This is something, of course, Netflix itself understands and has already been sort of working with other potential partners like TiVo to find other ways of getting entertainment into the home and not just DVDs through the mail.

And so this kind of comes back to, again, what is the sustainable competitive advantage that Netflix may hold moving into the future. And, again, I think one has to point to that technology and the database that they have now about movie preferences that help their subscribers choose things. That’s at the crux of the matter, and I think it’s something that doesn’t go away no matter what technology, no matter what approach you use to reaching the home, ultimately people have to decide how to spend their time in an evening or on a weekend in terms of enjoying some entertainment. And so that issue is always there, and I think Netflix is trying its very level best to corner the market in terms of having the best technology to achieve that objective.

In conclusion, there are so many facets that we can explore in looking at Netflix. It really is a digital enterprise to watch with an admirable CEO in Reed Hastings, who is doing some of the more creative things in the digital world today. And I think whether it’s looking at the business model, looking at a tremendous approach toward technology, toward the distribution network, to fostering communities online, we could really learn a lot from what Netflix is doing.

This is Professor Michael Rappa wishing you all the best with your studies.

(Music)

[End of Audio]

Unedited transcript of audio podcast produced on October 26, 2005.

Audio source file: http://digitalenterprise.org/podcasts/netflix.mp3

Michael Rappa is the Alan T. Dickson Distinguished University Professor of Technology Management at North Carolina State University.

For more information, please visit: digitalenterprise.org