Yale SOM Case 07-014 Netflix

yale case 07-014 february 20, 2007 Netflix Will Its Operations Model Be a Short Subject or a Long-Run Feature?

Jean W. Rosenthal1

Reed Hastings often told the story of his inspiration for Netflix: a $40 late fee from Blockbuster. He said, “It was all my fault. I didn’t want to tell my wife about it. And I said to myself, ‘I’m going to compromise the integrity of my marriage over a late fee?’”2 Still chagrinned over the late fee, Hastings, a dot-com multimillionaire, formed Netflix, a company that would rent DVDs through the mail for a monthly subscription price, with no postage charges or late fees. Hastings’s model for Netflix seemed simple enough. Netflix subscribers would create a wish list of DVDs on the company’s website, and Netflix would send a new title from the list when the previous rental was returned.

Behind the simple model however, Netflix’s success had been built on attending to every detail of its operations and adapting to the company’s various constituencies. For subscribers, Netflix designed a recommendations engine that customers liked and that allowed Netflix to shift subscriber interest from new releases. By attending to United States Postal Service (USPS) processes, Netflix had located its 41 warehouses, created processing procedures, and even designed its envelope in such a way as to minimize both operating costs and turnaround times. By working with the film studios, Netflix had reached agreements through which it reduced its risk in holding large numbers of DVDs from new releases.

The attention to detail paid off. Nine years after its April 1998 launch in the San Francisco Bay Area, Netflix generated net income of $49 million on revenues of $996.7 million. The firm boasted 6.3 million subscribers and carried an inventory of 70,000 titles on 42 million discs. Netflix’s website, in 2006, was rated the best website for retail satisfaction for the third year in a row.3 (See Exhibit 1 for Netflix financial data and stock prices.)

In spite of the company’s operational success, Netflix faced two big challenges in 2007. First, in 2006 Blockbuster had made a major move into online rental. In Blockbuster’s new service, subscribers could bring mailers directly to a Blockbuster store and immediately rent a DVD, getting the instant gratification denied to Netflix subscribers. By January 2007, Blockbuster had grown its online business to two million customers.

Second, a number of firms were beginning to offer video on demand (VoD). Netflix announced its own internet service in January 2007. The service complemented the existing subscriber service, generating no new fees. Netflix had budgeted $40 million to develop the system, but some analysts questioned whether that was sufficient to cover server data centers and licensing fees. Others argued that VoD would kill off the DVD rental business in general and that, for all its operational savvy, Netflix’s time had passed.

Challenges to Operations: Netflix in January 2007

Netflix offered subscribers a choice of monthly subscription plans, with the most popular priced at $17.99/month for three movies out (at home or in transit) at one time, allowing as many movies as a subscriber wanted over the month. Netflix charged no late fees. (Indeed, the company’s business model depended on people holding onto DVDs.) Late fees were a strong initial point of differentiation from Blockbuster, whose in-store late fees (“extended viewing fees” in Blockbuster parlance) had been a primary source of Blockbuster customer complaints — and revenues — until Blockbuster began its new online service.

To be able to make money on its customer service, Netflix had optimized their processes. Reed Hastings and his cofounders had come from dot-com companies and applied their knowledge of software and the internet to create Netflix’s operational strategy. They tailored their business to make the most of their relationships: adapting to studios as content suppliers to build inventory, adapting to customer interests to shift demand, and adapting to USPS procedures for quick and efficient delivery. But with challenges from Blockbuster and VoD to its core business, Netflix had to worry whether the procedures were sufficient, or whether Netflix would have to find further economies in order to survive.

70 Thousand Titles, 42 Million Discs The Netflix inventory did not include adult films or video games, but it had almost everything else. As Reed Hastings said in a December 2006 interview, “If it’s ever been put on DVD, chances are it’s in the warehouse.”4 (See Exhibit 2 for selected Netflix operational statistics.)

When Netflix began in 1998, the average DVD player cost nearly $600, and DVD players were in fewer than 7 percent of homes with televisions, while videocassette recorders (VCRs) were in 86 percent of homes.5 In spite of the much larger market at the time for VHS videocassette tapes, Netflix rented only DVDs, because of the disc’s small size, inherent strength, and much lower mailing cost than a videocassette. (A DVD in a mailing envelope required only one first-class stamp.) As Hastings noted in an interview, “DVDs are very compact, so you don't have to have a big warehouse. You know, you can fit about 10,000 DVDs in the back of an SUV.”6 In its early years, Netflix did not buy advertising, but instead included free rental coupons in new DVD packaging. Hastings’ focus on DVDs foreshadowed market development. By 2006, more than 81 percent of homes had DVD players, while VCRs had fallen to 77 percent.7

As its list of available titles and its customer base grew, Netflix required more discs in inventory. By December 2006, Netflix offered more than 70,000 titles and had an inventory of 42 million discs. The discs were located in multiple centers, designed to provide most customers with a new title within one or two days of a return. Faster turnaround also made it possible for Netflix to control its inventory while satisfying its growing customer base.

Incorporating the 100,000 new purchases that arrived every day into the Netflix’s inventory created its own operational challenges. In a 2005 interview, Tom Dillon, the company’s Chief Operating Officer and Chief Information Officer until his retirement in January 2006, described how Netflix had reduced labor costs related to receiving by 50 percent — and he was aiming for 75 percent. “When the company first started up, the receiving clerk would pick up a box from the pallet, find the paperwork, and match the box with the purchase order. Now the receiver just slaps a label and bar code on the box, and the system automatically matches the code with the purchase order.”8

2 NETFLIX

Netflix distributed its inventory across 41 mailing centers, although a large portion, 26 million discs out of the 42 million total, were housed in its original center near San Jose, California.9

A 2002 article stated that Netflix’s satellite centers deliberately maintained “zero inventory.” It described the process in regional centers in 2002 as follows:

Discs not mailed out (about 300 DVDs per day) are temporarily stored on a wire shelving unit. All the unshipped DVDs fit inside one cardboard container. At the end of every week, any unused inventory is shipped back to the main distribution center in San Jose for storage.”10

Netflix introduced software in January 2003 that dramatically reduced shelf time. As discs were returned by customers, many were matched to a customer request list and shipped back out the same day. Discs that were not matched for customer send-out that day were not sorted and shelved. Instead, they were tossed into bins to be rescanned the following day. If a disc remained in a regional center for more than a few days, it was returned to the California warehouse. With this software, Netflix reduced labor costs by about 15 percent, and the vast majority of discs never touched the shelf before they went out again. 11

Although Netflix had to grow its DVD library to deliver titles to subscribers with minimum delays, outside estimations of Netflix’s year-to-year inventory value were problematic, since Netflix changed its accounting methodology for valuing its DVD library several times. In its financial statements, Netflix showed its DVD library as a productive, non-current asset. In valuing inventory, Netflix considered the useful life (including utilization and loss and damage) of new release DVDs as one year, and back-catalog DVDs as three years.12 For DVDs that it had purchased directly and expected to sell, Netflix estimated a salvage value of $3.00. For DVDs obtained through revenue sharing agreements, after a fixed period, usually a year, the disc was either returned to the studio, destroyed, or purchased by Netflix.13

41 Distribution Centers, 1.6 Million DVDs a Day – One Disc at a Time From its inception, Netflix designed its service around post office procedures. By late 2006, Netflix was shipping more than 1.6 million DVDs a day, making it one of the post office’s top ten customers for first-class mail. The company made fast turnaround part of its brand identity, and saw delivery under three days as essential to retaining customers. It had located its mail centers so that, by 2006, close to 90 percent of the U.S. population — and 90 percent of Netflix’s customers — was within one-day delivery of its distribution centers, via first class mail. In January 2007, Netflix announced that it planned to reach 95 percent of customers with one-day service by the end of 2007, by adding nine additional mail centers and expanding its fleet of trucks to link those mail centers to an additional 50 regional post offices.14

Netflix had initially focused on the San Francisco Bay Area, and still enjoyed its greatest market penetration there, with 15 percent of households subscribing. During the company’s first three years, it had mailed all DVDs from its warehouse in the Silicon Valley near San Jose, and the Silicon Valley facility remained the largest, at 115,000 square feet.15 The facility also housed the Netflix receiving and storage center, its customer service center, and the processing and shipping center for the San Francisco Bay Area.16 As it had expanded nationwide, Netflix opened five to seven new centers each year. These satellite fulfillment centers were near post office distribution facilities in major metropolitan areas. They were typically around 8,000 square feet and employed eight to fifteen workers.17 According to one source in 2003, the typical cost for setting up a center was $60,000.18 All Netflix facilities were in rental spaces in lightly utilized industrial areas, so Netflix could adjust the size of centers to demand.19

3 netflix

Netflix trucks picked up returned envelopes from regional postal centers early each weekday morning and delivered presorted envelopes in the evening, in time for next-day delivery. As employees in each fulfillment center opened returning envelopes, they scanned the bar code on each disc. Software then retrieved the name and address of the next person on the waiting list for that DVD and printed a label for mailing out that afternoon. The software also generated emails to customers as discs were scanned in and again as they were mailed out, and controlled the movement of discs as they were sorted, shipped to the next customer, or restocked into storage.20 A 2003 PC Magazine article gave this description of the process:

An order placed by a customer in Manhattan will be assigned to the distribution center in nearby Flushing, New York. If the DVD is not available, the system will poll the next- closest distribution center, in Stamford, Connecticut. If that center doesn't have it, the system contacts the next closest, and so on until the DVD is located (even if that means sending it from Netflix’s main library in San Jose, California). If the disc is not found, the system will look for the customer's second choice back in Flushing. No matter where the disc is sent from, the system knows to print a return label to the Flushing facility to minimize return-mail times.21

An August 2006 New Yorker article described a similar process at the Netflix mail center in Maryland, which provided new titles by Wednesday to customers who had mailed movies on Monday. The center received and sent out 126,000 discs on Tuesday, the busiest day of the week for returns. The 40 employees started at 6:30 A.M., each opening between 450 and 1,150 returned envelopes an hour. By 11:00 A.M. they had completed the cycle of opening each envelope, recycling the empty envelopes, and checking the titles, discs, and sleeves. After an hour for lunch, employees begin stuffing: inserting an outbound disc into an electronically addressed envelope, flipping the envelope closed, peeling a sticker strip, and pressing the envelope together. The fastest workers could stuff 1,000 mailers an hour. All workers joined in hourly ergonomic exercises to prevent repetitive motion injuries.

The envelopes then entered a machine that placed tabs closing a third side, at 6,000 an hour, and then placed the envelopes into one of four zip code sorters, which ran until 7:30 p.m. to complete the presort (presorted mail qualified for discounted postage fees). As the warehouse closed, a truck drove one mile to deliver the envelopes to the post office, which accepted presorted mail until 8:30 p.m. for next day delivery.22 A December 2006 segment of 60 Minutes described the same procedures in the Silicon Valley main warehouse.23 (A link to the segment is on line at http://www.cbsnews.com /stories/2006/12/01/60minutes/main2222059.shtml.)

With 1.575 million envelopes a day by late 2006, Netflix spent nearly $400 million on postage, even with discounts for presorted sendout. In January 2006, Netflix hired Bill Henderson as Chief Operations Officer. He had been the 71st Postmaster General of the United States and had served as the USPS Chief Operating Officer from 1994 to 1998. Reed Hastings described Henderson as “the only guy on the planet who looks at our volume of mail and thinks of it as quite small…. It's a trickle of mail to him, where to anyone else it's a torrent.”24

May We Have the Envelope, Please? Netflix operations were adapted to post office sorting machinery as well as to post office procedures and schedules. In interviews, Reed Hastings’s version of Netflix’s creation story included a recounting of his purchasing several DVDs and mailing them to himself to see if they would survive.25 Jim Cook, who built and managed financial and operational functions from the company’s founding in 1997 until 1999, wrote about the effort that went into understanding

4 netflix

post office operations. He noted that he had spent hundreds of hours in the largest regional postal centers. He realized that the high-spinning circular drums used to process standard-sized letters would destroy DVDs. However, he noticed a separate conveyer belt system that sorted larger pieces of “flat mail” that would work well for mailing DVDs. Cook wrote,

I found out that if an envelope had certain dimensions and other characteristics, it would be sorted by this alternate system instead of the large, crushing metal drums. Better yet, this flat mail-sorting machine would read a bar coded delivery address and could automatically sort the item into “carrier walking route” sequence. Now the wheels were really turning. The fact that we could provide the right-size packaging, bar code, and other characteristics would make it possible for extremely fast processing of a mail piece with absolutely no human intervention or other physical touch.26

With this knowledge, Netflix went on to design its patented mailer. Cook noted that “Above all else, this envelope was our ‘product.’ It was the only thing that our customers would touch and see. Therefore, it had to have all the key features of a great marketing piece.” Cook described a number of additional criteria for the envelope:

 It had to effectively hold and protect the DVD.  It had to meet stringent post office criteria so we could mail with the equivalent cost of a first-class stamp.  It had to transform into the return envelope, so that DVDs would find their way back to Netflix quickly and in good condition.  It had to be “operational” — easy to insert and remove the discs and something that could be pre-printed in mass quantities. 27 Years of experimentation went into creating the perfect DVD envelope for the post office and the customer — lightweight, recyclable, sturdy enough for the round trip, and efficiently loaded and sealed at a mail center and opened on its return. Cook described the evolution of envelope design:

Since 1998, there have been over 150 versions of this little red package. It wasn't always red, but we determined that red was the easiest color to see in the post office. It was not always paper thin — our first package was much thicker, and we shipped three DVDs in the same envelope. By testing, learning, and improving, we did such non-intuitive things as print the inside return address upside down to make processing more efficient. We had to build a mini “pocket” inside the package to ensure that the Post Office stamp-canceling machine wouldn't break the DVD inside. A very recent change added a little cut-out on the outside of the package, enabling one to check the disc in by “seeing” the inside disc barcode without having to open the package. The amount of time and motion this one little step saves is enormous, given today's processing volumes of over 20 million DVD returns every month.28

(A link to a gallery illustrating the evolution of the Netflix envelope is available online at http://money.cnn.com/2006/04/20/ technology/business2_netflixgallery/index.htm.)29

Although Cook described the current envelope as “perfect,” Netflix’s operational model depended on its ability to continue to conform to post-office operations and DVD technology. For example, new HD-DVD or Blu-ray discs might turn out to be too fragile for the current envelope, requiring more padding, which would increase postage costs dramatically. Other changes in post office equipment or requirements would require Netflix to adapt its procedures and design.

5 netflix

Adapting to the Studios Netflix success also depended on its ability to interact with the producers of the content contained on DVDs: the movie and television studios. As a new medium, DVDs had offered Netflix a window of opportunity. Prior to the introduction of DVDs, movie studios had negotiated exclusivity contracts with premium cable channels to gain financing for new films. The contracts delayed the general availability of movies through subscription services, but the contracts did not cover DVDs, so Netflix was able to rent out new titles before they were available on premium cable channels.

Netflix negotiated revenue-sharing agreements with the studios, with studios providing Netflix with DVDs for new titles at cost, in exchange for a share of subscription revenues.30 This approach reduced inventory costs appreciably. (Blockbuster had pioneered revenue sharing for VHS, but had not made a similar arrangement for DVDs.31)

The Netflix annual report Form 10(k) described the revenue sharing as follows:

“These revenue sharing agreements generally have terms of up to five years. Revenue share agreements typically enable us to increase our copy depth of DVDs on an economical basis because of the low initial payment. Additional payments are made only if our subscribers rent the DVD. Under a purchase arrangement, we must pay the full wholesale price, regardless of whether the DVD is rented. In addition, revenue sharing agreements generally provide for studio promotional support of the associated DVD and our service as well as permit us to own the DVD following expiration of the revenue sharing period, typically no more than 12 months following street date.”32

To have most titles immediately available to subscribers, Netflix purchased multiple copies of titles, based on a title’s expected popularity, as predicted by comparing the new title to existing titles in its customer rating software. In a 2003 interview, Ted Sarandos, Chief Content Officer at Netflix, described the range of holdings. In that year, Netflix had bought 60,000 copies of My Big Fat Greek Wedding, as well as a few hundred copies of an independent title from a director just out of film school.33

Sarandos looked beyond the big studios, to film festivals and boutique films, for new offerings. The company also resolved the number of copies to purchase as a software challenge. It recreated algorithms that looked at box-office results and rentals for similar movies, as well specific indicators such as the number of subscribers listing a movie on rental queues before the title was available on DVD. Other movies, such as Hotel Rwanda, with limited initial box-office appeal, had become top rentals on the Netflix system, primarily through the recommendation system.34 Documentaries, which traditionally had limited box office appeal, found new life on Netflix. Netflix also began Red Envelope, an independent movie distribution and production company supporting small films.

Netflix’s sharing contracts and other arrangements with studios depended on its negotiating success and the studios’ willingness to allow companies like Netflix to have the opportunity for rental and sales. The studios’ control of timing and pricing made Netflix dependent on its ability to adapt. Netflix’s own entry into film development exposed it to a much riskier world, with a more volatile income stream.

Demand Creation: The Netflix User Experience Customers made their movie selections on Netflix’s website. According to Jim Cook, Netflix initially had leaned heavily on Amazon’s web design:

6 netflix

Why reinvent the wheel if someone has already come up with an easy-to-use, useful, and elegant solution? When designing the Netflix website, we turned to the best: Amazon. Some of the ideas that we adapted: product and button placements; overall color schemes; size of DVD images for fast page loading; customer reviews and movie reviewer articles; easy-to-use search with categorized searching by movie genre; overall website navigation.”35

Netflix added to this mix its movie queue and its recommendation system. In May 2003, Netflix filed a patent application for its “approach to renting items for customers.”36 In April 2006, the same day the patent was issued, Netflix sued Blockbuster’s new online rental program for infringement of its patents, particularly for the method of prioritizing and ordering from the queue. Blockbuster announced it would contest the suit.37

Netflix used its online recommendation system to help customers build their queues and to stimulate customer demand for titles outside the most recent releases. Customers received individualized recommendations for titles, based on their ratings of movies they had already seen. To the extent that Netflix could move customers out of recent releases and into titles that were requested less frequently, it could maximize the use of its inventory and reduce the number of copies it needed of new releases.

Netflix’s recommendation engine, CineMatch, was built on an Oracle database, and included over a billion ratings, with two million additional ratings added every day. The proprietary software organized the Netflix library into clusters of similar movies, and then analyzed how customers rated them. Those who had given similar ratings to the same movies in a cluster were then matched as like-minded viewers. Then CineMatch looked in real time at an individual subscriber's ratings and recommended films that had been highly rated by matched viewers.38

In 2006, Netflix launched a highly publicized contest with a million dollar prize for new software that could improve the accuracy of CineMatch’s customer recommendations by 10 percent. It posted interim results on its website; by January 2007, top teams of software developers world- wide had entered, and the top entries had found ways to improve matches by over 7 percent.

Netflix continued to develop additional customer experiences for its website to widen the list of movies that subscribers might consider adding to their online queues. For example, through a “Friends” page, subscribers shared their lists and recommendations with others. Subscribers were also offered lists of the top 100 rentals, and the top rentals in a subscriber’s local zip code.

By mid-2006, Netflix subscribers were renting 35,000 to 40,000 different titles every day, and two-thirds of its rentals were from the company’s backlist, rather than recently released films.39 Increasing backlist rentals increased customer satisfaction with the rental experience and allowed Netflix to reduce its inventory of new releases and thus avoid the revenue-sharing payments that recent releases required.40 As Sarandos noted in a 2006 interview,

If you’re in the hits fulfillment business, the top 10 movies of the year, like Blockbuster and most people are, the margin pressure is enormous…. We are completely upside- down from that, so 70 percent of what we ship is deep catalog, and 30 percent new releases…. There’s a lot better margin in demand creation than in demand fulfillment.41

While Netflix’s web design enchanted customers, the company’s advantage was not necessarily sustainable. Amazon, which was also highly rated by consumers, had a service in the United Kingdom, DVD “Rental by Post,” which was similar to Netflix. Amazon had never brought that service to the United States, but that strategy could change.

7 netflix

Throttling Customer Demand – Too Much of a Good Thing? Fast turnaround was both a key — and a challenge — to Netflix’s success. Customer satisfaction depended on Netflix’s ability to send out movies within one or two days of receiving a return. However, Netflix found that some customers were very heavy users of the system, turning over their rentals almost right away. The most common plan was three movies at a time for $17/month, with free shipping; the average subscriber rented two to 11 DVDs per month, With round-trip costs approaching $0.85 to $1.00 per rental and round-trip postage alone roughly $0.68, a subscriber under that plan renting more than 17 or 20 movies a month would cost Netflix money.

Some customers who were heavy users of the system found their accounts “throttled,” or in Netflix terms “smoothed,” by receiving DVDs from centers that were two shipping days away instead of one, and other unexplained delays.42 One enterprising subscriber tracked his accounts and published the results in April 2003. (See Exhibit 3 for a summary of Michael Muegel’s results.) He compared the accounts of heavy users to light and new users and found that heavy- user accounts received return DVDs much more slowly and were less likely to receive titles highest in their queues.43 These and similar complaints led to a consumer class-action lawsuit, Frank Chavez v. Netflix, Inc, in September 2004. The suit alleged false advertising, in relation to claims of “unlimited rentals” with “one-day delivery.” (See Chavez v. Netflix Inc., No. CGC-04- 434884; settlement approved at Cal. Super. Ct., San Francisco County Oct. 27, 2005) Class Action Litigation Reporter, Volume 12, Issue 10, November 16, 2005.)

Netflix settled the suit in October 2005, without admitting wrongdoing or apparently changing its service. The settlement gave all subscribers a one-month free upgrade to the next highest level of service, paid Chavez $2,000, and covered legal fees and costs of around $2.5 million.44 In January 2006, the company did change its “terms of service” agreement with customers to acknowledge that those renting fewer titles get faster service and titles higher in their queue. (See Exhibit 4 for the resulting terms of service from Netflix’s website.)

While Netflix retained the right to provide slower delivery to more frequent users, the policy seemed to create a penalty for those who most enjoyed the service that the company offered. Some analysts and a number of bloggers argued that this policy could create friction with loyal customers. On the other hand, alternative solutions, such as setting a maximum number of rentals per month, might interfere with marketing efforts to the majority of potential subscribers, who were not likely to have such heavy usage patterns.

In-Store AND Online Rental: The Blockbuster Challenge

Netflix had overcome several challenges to its leadership in DVD rentals. At first, the company had had to overcome consumer reluctance to wait a few days to get a movie rather than grab one immediately from a local store. As Netflix grew its online rental business, competitors entered the market. In 2002, Wal-Mart began an online rental service, but by 2005, Netflix had driven it out of the DVD rental business.

But by January 2007, Netflix faced a revamped threat from Blockbuster, which had developed a new rental service that combined online rental with in-store renewal. In 2005, Blockbuster, the largest brick-and-mortar video rental company, had eliminated late fees in most of its stores, and at the end of 2006 it created “Total Access Service,” combining online and in-store rentals, and began a major push to ramp up its online customer base.

Blockbuster itself was in trouble. With the market shift to DVDs, the availability of DVDs through the mail, and the low price of DVD purchases, Blockbuster’s stores had become a

8 netflix

liability. At the end of 2003, it had 5,670 stores (4,579 owned by Blockbuster, and 1,091 under franchise agreements). By 2006 it had reduced the number of U.S. stores to 4,000 and had announced a plan to continue to reduce stores by around 800 a year as leases expired.45 Netflix rentals had cut into Blockbuster’s in-store rental volume, and eliminating late fees had cut into Blockbuster revenues.46

In retrospect, Blockbuster had made a number of strategic mistakes. In 1998 it had refused to accept Warner Bros.’s offer to entrench a DVD rental window under a shared-revenue program (the same program that Netflix later negotiated). Then Blockbuster turned down an opportunity to purchase Netflix for $50 million. By 2006, the company Sumner Redstone of Viacom had bought in 1994 for $8.4 billion had a market value of under $700 million.47 (Exhibit 5 shows Blockbuster financials and stock price.)

By the end of 2006, Blockbuster owned a library of 65,000 titles for its online service and added 30 regional mail centers in addition to its central warehouse in Texas. It initially created a completely separate inventory system for its online service, locating discs in mail centers rather than having the central warehouse provide discs for stores.48

After starting its new offering with one free monthly in-store rental, Blockbuster added several twists: It allowed customers to return a mailed title to a local store, giving customers both faster mail-out of the next title from their queue as well as an immediate, free store replacement (under rental-store rules). “Frequently asked questions” on Blockbuster’s website suggested that it was having trouble with DVD breakage during mailing. In addition, the company’s website was not as easy to navigate nor as informative as Netflix’s. However, the company’s ability to provide DVDs instantly to subscribers willing to visit a store did offer an advantage over Netflix.

Blockbuster had announced that as part of its new service it expected to reduce the number of stores and reduce the number of titles carried in each of the remaining stores in order to emphasize “hot new titles and high-value promotional movies” for in-store rentals. It also planned to use the 1,000 largest stores as mail centers for the online rental service, to provide faster turnaround times.49 With a greatly increased advertising budget, Blockbuster promoted its combined service by promising instant gratification (you would always have a movie to watch), and no throttling (“we will never slow you down”). By the end of 2006, it had grown its online customer base to 2.3 million subscribers, 2 million of them paying. Although some analysts were still predicting Blockbuster’s demise, others saw it surviving to rival, if not kill, Netflix’s service.50

Blockbuster’s own online service put additional pressure on profitability of its U.S. stores, and it was not clear whether its own brick-and-mortar stores could withstand the challenge of online service from both Netflix and Blockbuster itself. As for the impact on Netflix, although Reed Hastings did not describe the Blockbuster challenge as a serious threat, he acknowledged that the new service creating a “headwind” for Netflix.51 Netflix’s customer growth was slower in 2006 than it had been in prior years, and the company had increased its spending on advertising.

Delivering Movies, not DVDs: the VoD Challenge

Although Netflix used online ordering and communication, its rental service was, in the end, the round-trip delivery of a physical object through the U.S. Postal Service. For years, technologists had been predicting that video-on-demand services would be available in homes “just a few years down the road.” At some point that forecast might finally come true for more than just a few, specialized services. Many believed VoD would cause the market for DVD rentals and purchases to collapse. VoD would end the reliance on moving a physical object, would be available immediately, and would avoid Netflix’s major expenses of postage and handling.

9 netflix

However, true VoD over the internet faced a number of unresolved problems. As of 2007, no technology offered an easy interface between computers and televisions. Downloads were time- consuming, particularly as high-definition products with large files came on the market. The February 2007 Consumer Reports noted that a typical five-gigabyte, high-definition movie would take 9 hours with low-speed DSL, 2.8 hours with cable, and 17 minutes with a fiber connection.52 In addition, studios were reluctant to make digital versions of movies readily available. They were fearful of pirating and the loss of copy protection. Yet another limitation to the growth of VoD had been the studio contracts with premium television channels, which provided the channels with exclusive digital rights for defined periods.53

By 2006, on-demand video encompassed a wide range of quality and availability. True VoD was available in some hotels and via limited online offerings. Cable service offered “near” on-demand service — titles were broadcast every 20 minutes, so customers could watch at any time, but titles were limited. Comcast, for example, offered only 800 movies. Video, although at a lower quality, was available through iTunes, telephones and YouTube.

Netflix launched View Now in mid-January 2007 for the first wave of subscribers, and it expected to make the service available to all subscribers by the end of June. About 1,000 films and televisions shows were available for watching at the launch of the service, and the library was expected to reach 5,000 films by the end of 2007. To use the new feature, subscribers would be required to accept a one-time download of a browser applet, which would enable a movie to play on their computers within 10 to 15 seconds. Subscribers with a 3Mbps broadband connection would have DVD-quality viewing on their computer, and those with slower broadband speeds could still use the service.54 The movies streamed to computers with Windows software, and were not available on Apple computers or other operating systems. In addition, the movie had to be watched on the computer, with no easy access to home television sets.

Netflix was using its own servers, locating them across the country. It streamed movies to the subscriber’s computer, beginning immediately after the subscriber clicked “play.” The number of hours available corresponded to a subscriber’s monthly subscription plan. Subscribers at the $5.99 per month level would be able to see six hours of video or TV shows monthly, and those paying $17.99 would receive 18 hours each month. The amount delivered was measured by time, not the total length of any title, so that subscribers could use their time to browse through many titles, watching only a few minutes of each, or only particular scenes, or only the beginning, or the end.

Netflix announced that it was spending $40 million in 2007 for the development of online video, and in January 2007 increased the estimate to “more than” $40 million, as some analysts argued that $40 million was an unrealistically low forecast of first-year costs. Although Netflix declined to provide a breakdown of costs, observers suggested that securing rights was the largest portion of the spending, since hardware prices were low.55 Netflix had negotiated arrangements with NBC Universal, Sony Pictures, Metro-Goldwyn-Mayer Inc, 20th Century Fox, Paramount Pictures, Warner Bros., Lionsgate, New Line Cinema, and a number of television studios, but a few studios such as Disney had not yet agreed to offer its movies.56

As he announced the new service, Reed Hastings made it clear that he did not expect the streaming service to be more than a “microscopic” portion of Netflix’s movie delivery for the immediate future. He believed that the website customer experience would keep Netflix in the market even as VoD grew, but, he stated, “DVD is going to be a very big market for a very long time.”57

Hastings predicted that online video would develop over time into three separate markets: first, advertising-supported video, with sellers like YouTube and Yahoo; second, download for

10 netflix

purchase, dominated by companies like Apple, Amazon, and Wal-Mart; and third, video rental, which he saw as Netflix’s expertise. (The Wal-Mart announcement on February 6, 2007, of a test version of its new download-to-sale service of new titles at retail rates, offering movies as well as television shows, seemed to confirm Hastings’ assessment.)

Hastings stated that streaming service was a first step in this evolving market area. With no way to predict consumer demand for the product, this first step would provide Netflix with more data to evaluate which of its subscribers were interested in this service and how powerful a draw it would prove to be. If the service turned out to be popular, it could replace rentals through the mail, particularly for the latest titles, eliminating postage and handling costs. In addition, although the company did not say so explicitly, the service provided a key competitive advantage. It provided instant gratification — a movie now — without even a drive to the rental store. On the other hand, Netflix’s service was limited to streaming video, not a downloadable file that the customer could keep or play on another device. Whether customers would be satisfied with online rental at this level, rather than purchase, remained to be seen.

Netflix had succeeded by adapting to its suppliers, its delivery system, and its customers. It now had to look for ways to adapt to new competitors and technology. Would the operational excellence that had driven their early success allow it to meet these new challenges? Or, as so many companies before them, would they fade away as their niche disappeared?

This case was developed under the guidance of Arthur J. Swersey, Professor of Operations Research at the Yale School of Management.

This case has been developed from published sources and is intended for pedagogical purposes. The case is not intended to furnish primary data, serve as an endorsement of the organization in question, or illustrate either effective or ineffective management techniques or strategies. Copyright 2007 © Yale University. All rights reserved. Reprint only with permission of Yale University School of Management. To order copies of this material or to receive permission to reprint any or all of this document, please contact the Yale SOM Case Study Research Team, 135 Prospect Street, New Haven, CT 06520.

Endnotes

1 Case writer, Case Study Research Team, Yale School of Management. 2 Reed Hastings, “Out of Africa, Onto the Web,” as told to Amy Zipkin, New York Times, December 17, 2006. 3 ForeSee Result, December 2006 poll. 4 CBS News, “The Brain behind Netflix,” 60 Minutes, December 3, 2006. 5 Nielsen Media Research’s Home Technology Report, as reported in MediaWeek, December 19, 2006. 6 “Netflix Opens Tempe Distribution Center,” The Business Journal of Phoenix, February 26, 2003. 7 Nielsen Media Research’s Home Technology Report, as reported in MediaWeek, December 19, 2006. 8 Larry Stevens, “Growing Pains: Netflix IT Has to Deliver High Efficiency, Support Fast Growth,” CIO Insight, October 15, 2005, http://www.cioinsight.com/article2/0,1540,1875170,00.asp.

11 netflix

9 CBS News, “The Brains behind Netflix,” 60 Minutes, December 3, 2006. 10 Sarah Mason, “Movies on Demand: Efficiency Combines with New Technology in Netflix’s DVD Rental Service,” IIE Solutions, October 2002. 11 See Lucas Conley, “90,000 DVDs, No Shelves,” Fast Company, No. 74, September 2003, http://www.fastcompany.com/magazine/74/netflix.html; and also Christopher Null, “How Netflix Is Fixing Hollywood By Finding a Market for Niche Titles — and Keeping Discs in Constant Circulation — the Online DVD Rental Pioneer Is Shaking Up the Movie Biz,” Business 2.0 Magazine, July 1, 2003, http://money.cnn.com/magazines/business2/business2_archive/2003/07 /01/345263/index.htm. 12 Netflix, 2005 Annual Report, p. 24. 13 Netflix, 2005 Annual Report, p. 25. 14 Reed Hastings, Netflix analyst conference call, January 24, 2007, online at http://ir.netflix.com/eventdetail .cfm?EventID=33072.

15 Joan O’C. Hamilton, “Home Movies: Now that He's Put Hollywood in Your Mailbox, Reed Hastings Is Watching Out for the Future of Entertainment,” Stanford Magazine, January/February 2006 http://ed.stanford.edu/suse/news-bureau/displayRecord.php?tablename=susenews&id=139. 16 Netflix 2005 Annual Report, p. 20. 17 Aude Lagorce, “Netflix's Next Challenge,” Forbes, October 2, 2003, http://www.forbes.com /2003/10/02 /cx_al_1002nflx_print.html. 18 Kim S. Nash, “Netflix: Box-Office Star,” eweek, January 16, 2004, http://www.eweek.com /article2/0,1895,1456971,00.asp. 19 Reed Hastings, Netflix analyst conference call, January 24, 2007, online at http://ir.netflix.com /eventdetail.cfm?EventID=33072. 20 Joan O’C. Hamilton, “Home Movies: Now That He's Put Hollywood in Your Mailbox, Reed Hastings Is Watching Out for the Future of Entertainment,” Stanford Magazine, January/February 2006 http://ed.stanford.edu/suse/news-bureau/displayRecord.php?tablename=susenews&id=139. 21 Alan Cohen, “Netflix: DVDs at Your Door.” PC Magazine, February 19, 2003, http://www.pcmag.com /article2/0,4149,894278,00.asp. 22 Susan Sheehan, “Moving Pictures: Tear, Slap, Clack,” The New Yorker, “Talk of the Town,” August 28, 2006. 23 CBS News, “The Brain behind Netflix,” 60 Minutes, December 3, 2006. 24 Katie Hafner, “Postal Service Finds a Friend in the Internet,” New York Times, August 2, 2006, http://www.nytimes.com/2006/08/02/business/02postal.html?ex=1312171200&en=52b1770033b09 30f&ei=5088&partner=rssnyt&emc=rss. 25 CBS News, “The Brain behind Netflix,” 60 Minutes, December 3, 2006. 26 Jim Cook, with Suzanne Taylor, “Five Lessons from the Netflix Startup Story,” Marketingprofs.com Newsletter, April 11, 2006, http://www.marketingprofs.com/6/cooktaylor1.asp. 27 Ibid. 28 Ibid. 29 G. Pascal Zachary, “The Evolution of the Netflix Envelope: The Key to Netflix's Lean Operations Is Its Lightweight, Versatile Mailer, Business 2.0 Magazine, April 21, 2006, http://money.cnn.com /2006/04/20/technology/business2_netflixgallery/index.htm. 30 Netflix, 2005 Annual Report, Form10(k), p. 13.

12 netflix

31 Jeffrey M. O'Brien, “The Netflix Effect, Wired, December 12, 2002. 32 Netflix 2005 Annual Report, Form 10(k), p. 13. 33 Christopher Null, “How Netflix Is Fixing Hollywood by Finding a Market for Niche Titles — and Keeping Discs in Constant Circulation — the Online DVD Rental Pioneer Is Shaking Up the Movie Biz,” Business 2.0 Magazine, July 1, 2003, http://money.cnn.com/magazines/business2 /business2_archive /2003/07/01/345263/index.htm. 34 CBS News, “The Brain behind Netflix,” 60 Minutes, December 3, 2006. 35 Jim Cook, with Suzanne Taylor, “Five Lessons from the Netflix Startup Story,” Marketingprofs.com Newsletter, April 11, 2006, http://www.marketingprofs.com/6/cooktaylor1.asp. 36 United States Patent: 7024381. 37 Netflix Inc. v. Blockbuster Inc., No. C 06 2361 (N.D. Calif.), Blockbuster Inc. Form 10-Q, Filed May 9, 2006, for period ending March 31, 2006. 38 Alan Cohen, “Netflix: DVDs at Your Door,” PC Magazine, February 19, 2003, http://www.pcmag.com /article2/0,4149,894278,00.asp. 39 David Leonhardt, “What Netflix Could Teach Hollywood,” New York Times, June 7, 2006. 40 Jeffrey M. O'Brien, “The Netflix Effect,” Wired, December 12, 2002. 41 Ted Sarandos, Chief Content Officer at Netflix, speaking at Cowen and Company 2nd Annual Internet Conference, Thursday, Dec 7, 2006, online at http://www.corporate-ir.net/ireye/confLobby.zhtml?ticker =NFLX&item_id=1415440. 42 ABC News, February 11, 2006. 43 Michael S. Muegel, “An Analysis of Netflix's DVD Allocation System,” Originally published April 22, 2003. http://dvd-rent-test.dreamhost.com/. 44 Documents from the lawsuit are online at http://www.sftc.org/Scripts/Magic94/mgrqispi94 .dll?APPNAME=IJS&PRGNAME=ROA&ARGUMENTS=-ACGC04434884. 45 John Antioco, Blockbuster CEO and Chair, speaking at the 17th Annual Entertainment, Media & Telecommunications Conference, January 9, 2007, online at http://www.veracast.com/webcasts /citigroup/emt07/14305203.cfm. 46 Edward Jay Epstein, “Hollywood's New Zombie: The Last Days of Blockbuster,” Slate Magazine, Monday, January 9, 2006; and Edward Jay Epstein, “Hollywood's Biggest Blunders: Inexplicable Decisions of the New Moguls,” Slate Magazine, Monday, September 19, 2005. 47 Edward Jay Epstein, “Hollywood's New Zombie: The Last Days of Blockbuster,” Slate Magazine, Monday, January 9, 2006. 48 Janet Rae-Dupree, “Movie Moguls,” CIO Insight, August 5, 2005. 49 John Antioco, Blockbuster CEO and Chair, speaking at the 17th Annual Entertainment, Media & Telecommunications Conference, January 9, 2007, online at http://www.veracast.com/webcasts /citigroup/emt07/14305203.cfm. 50 Rick Aristotle Munarriz, “Blockbuster's Millions,” Motley Fool, January 4, 2007, online at http://www.fool.com/investing/general/2007/01/04/blockbusters-millions.aspx. 51 Reed Hastings, Netflix analyst conference call, January 25, 2007. 52 “Internet Service: Fiber Joins the Fray,” Consumer Reports, February 2007. 53 David Leonhardt, “What Netflix Could Teach Hollywood,” New York Times, June 7, 2006, http://www.nytimes.com/2006/06/07/technology/07leonhardt.html.

13 netflix

54 Steven Schwankert, “Netflix Launches Direct Download Service for PCs,” IDG News Service, 1/16/2007, http://www.itworld.com/Net/3589/070116netflix/. 55 Reed Hastings, Netflix analyst conference call, January 24, 2007, online at http://ir.netflix.com /eventdetail.cfm?EventID=33072. 56 Gina Keating, “Netflix Launches 1,000-Title Online Movie Feature,” eweek.com, January 16, 2007. 57 Miguel Helft, “Netflix to Deliver Movies to the PC,” New York Times, January 16, 2007.

14 netflix

Exhibit 1: Netflix Financial Data

Annual Income Statement (Standardized) (millions of dollars)

12 months ending 31-December 2006 2005 2004 2003 2002 ------Reclassified Normal 31-Dec-05------\ Total Revenue 996.7 682.2 500.6 270.4 152.8

Cost of Revenue, Total 626.1 464.6 330 179 97.5

Gross Profit 370.6 217.7 170.6 91.4 55.3

Selling/General/Administrative Expenses, Total 266.2 185.7 130.9 70.3 51.4 Other Operating Expenses, Total -4.8 -2 -2.6 -1.2 NA

Total Operating Expense 932.2 679.2 481.3 265.9 163.5

Operating Income 64.4 3 19.4 4.5 -10.7

Interest/Investment Income, Non- Operating 15.9 5.8 2.6 2.5 1.7 Interest Income (Expense), Net Non-Operating 15.9 5.3 2.4 2 -10.3 Income Before Tax 80.3 8.3 21.8 6.5 -20.9 Income Tax - Total 31.2 -33.7 0.2 0 0 Income After Tax 49.1 42 21.6 6.5 -20.9 Net Income Before Extra. Items 49.1 42 21.6 6.5 -20.9

Net Income 49.1 42 21.6 6.5 -20.9

Source: © 1983-2007 Reuters Research Inc. All Rights Reserved. Published by OneSource Information Services, Inc., February 2007 http://businessbrowser.onesource.com/web/Reports/cia.aspx?KeyID=L44098532&Process=CP

15 NETFLIX

EXHIBIT 1 (CONTINUED)

Netflix Annual Balance Sheet (Standardized) (millions of dollars)

December 31, 2006 2005 2004 2003 2002*

Cash, Short Term Investments 400.4 212.3 174.5 135.2 103.6

Prepaid Expenses 14.2 13.1 7.4 3.5 3.1 Other Current Assets 13.8 18.3 5.4 0.2 0.4 Total Current Assets 428.4 243.7 187.3 138.9 107.1

Property/Plant/Equip., Net 55.5 40.2 18.7 9.8 5.6

Intangibles, Net 105.9 57.5 43.1 25.2 16.1

Other Long Term Assets, Total 19 23.3 2.6 2.1 1.8

Total Assets 608.8 364.7 251.8 176 130.5

Accounts Payable 93.9 63.5 49.8 32.7 20.4 Accrued Expenses 29.9 25.6 13.1 11.6 9.1 Notes Payable/Short Term Debt NA 0 0 0 0 Current Port. LT Debt/Capital Leases NA 0 0.1 0.4 1.2 Other Current Liabilities 69.7 48.5 31.9 18.3 9.7

Total Current Liabilities 193.4 137.6 94.9 63 40.4

Total Long Term Debt NA 00 0 0.5

Total Debt NA 0 0.1 0.5 1.7

Other Liabilities, Total 1.1 0.8 0.6 0.2 0.3

Total Liabilities 194.6 138.4 95.5 63.3 41.2

Total Equity 414.2 226.3 156.3 112.7 89.4

Total Liabilities, Shareholders’ Equity 608.8 364.7 251.8 176 130.5

Total Common Shares Outstanding 68.6 54.8 52.7 50.8 44.9

* Reclassified Normal 31-Dec-03

16 NETFLIX

EXHIBIT 1 (CONTINUED)

Netflix Annual Cash Flow (Standardized) (millions of dollars)

12 months ending 31-December 2006 2005 2004 2003 2002*

Net Income/Starting Line 49.1 42 21.6 6.5 -20.9 Depreciation/Depletion 15.9 9.1 5.9 4.7 5.9 Amortization 141.2 97.9 82.3 46.3 20.6 Other Non-Cash Items 15.6 -20.6 16.8 10.8 20.3 Non-Cash Items 6.5 -24.2 14.3 9.2 18.6 Cash Taxes Paid NA -1 0 0 NA Cash Interest Paid NA 0.2 0.1 0.3 0.6 Changes in Working Capital 35.1 32.6 23.5 23.1 16

Cash from Operating Activities 247.9 157.5 147.6 89.8 40.1

Purchase of Fixed Assets -27.3 -27.7 -15 -8.9 -2.8 Purchase/Acquisition of Intangibles -170.1 -111.9 -103 -55.6 -24.1 Capital Expenditures -197.4 -139.6 -117.9 -64.5 -26.8 Sale of Fixed Assets 0 0 NA NA NA Sale/Maturity of Investment 12.9 5.8 50.6 1.8 2 Purchase of Investments NA 0 -0.6 -1.7 -43 Other Investing Cash Flow -1.3 0.6 -0.5 -0.3 0.6 Other Investing Cash Flow Items, Total 11.6 6.3 49.6 -0.2 -40.5

Cash from Investing Activities -185.9 -133.2 -68.4 -64.7 -67.3

Other Financing Cash Flow 13.2 0 NA NA 0 Financing Cash Flow Items 13.2 0 NA NA 0 Sale/Issuance of Common 113 13.4 6 6.3 88 Repurch./Retirement Common NA NA 0 0 0 Issuance (Retirement) of Stock, Net 113 13.4 6 6.3 88 Long Term Debt, Net 0 -0.1 -0.4 -1.3 -17.1

Cash from Financing Activities 126.2 13.3 5.6 5 70.9

Net Change in Cash 188.2 37.8 84.6 30.1 43.7 Net Cash - Beginning Balance 212.3 174.5 89.9 59.8 16.1 Net Cash - Ending Balance 400.4 212.3 174.5 89.9 59.8

* Reclassified Normal 31-Dec-03 © 1983-2007 Reuters Research Inc. All Rights Reserved. Published by OneSource Information Services, Inc., February 2007 http://businessbrowser.onesource.com/web/Reports/cia.aspx?KeyID=L44098532&Process=CP

17 NETFLIX

EXHIBIT 1 (CONTINUED)

Netflix Stock Price History

Source: Bigcharts.com. Downloaded February 4, 2007.

18 NETFLIX

Exhibit 2: Netflix Selected Statistics

General Statistics

END OF YEAR 2001 2002 2003 2004 2005 2006

Number of subscribers (thousands) 857 1,487 2,610 4,179 5,662

Price of most popular package $/mo 19.95 19.95 17.95 17.95 (3 at a time) 17.95

Number of Mail Centers 18 23 30 37 41

Titles in stock (thousands) 14.5 18 35 55 70

Discs in inventory (thousands) 5,100 12,000 24,000 25,000 42,000

Titles shipped 2,500 4,000 1,000 1,500 per week per week per day per day % of customers with 1-day shipping 50% 80% 85% 90% 90%

US TV households with 1999 DVD players (Adams 34% 50% 64% 75% 81.2% 7.6% Media) ` DVD library valuation total 35,039 58,795 198,216 304,490 NA (thousands) 110,360

31,406 17,417 80,346 96,883 NA annual amortization 43,125

8,851 24,070 102,971 113,950 NA Acquisitions 55,620 Proceeds from DVD sales -- 1,988 1,833 5,617 5,781 NA

3,633 9,972 42,185 57,053 NA DVD library, net 22,238 [changes in evaluation methods over period]

movie ratings in customer database (thousands) 2,000 300,000 525,000 2,500,000 NA

% inventory rented in last quarter 97% 99% 95%

EMPLOYEES full-time employees 567 940 985 1200 temporary employees 224 253 445 NA

Source: Netflix Annual Reports (10K), various years.

19 NETFLIX

EXHIBIT 2 (CONTINUED)

Fulfillment Costs (thousands of dollars)

END OF YEAR 2001 2002 2003 2004 2005 2006

Revenues (thousands) 35,894 74,255 150,818 270,410 500,611 682,213

Fulfillment costs 10,247 13,452 19,366 31,274 56,609 70,762

total cost of revenues 35,108 62,540 96,410 179,010 330,010 454,550

Gross profit 786 11,715 54,408 91,400 170,601 217,663

28.5% 18.1% 12.8% 11.6% 11.3% 10.4% Fulfillment as % of revenues

Net income (58,274) (39,182) (20,948) 6,512 21,595 42,027

Notes: values restated in 2005 to reclassify fulfillment expenses into cost of revenues Fulfillment expenses represent those expenses incurred in operating and staffing shipping and customer service centers, including costs attributable to receiving, inspecting and warehousing library. Fulfillment expenses also include credit card fees and other collection-related expenses.

Source: Netflix Form 10(K), various years (as filed)

Netflix Subscriber Churn *

*Churn is defined as cancellations in the quarter divided by the sum of beginning subscribers and gross subscriber additions, divided by three months.

Source: Netflix Corporate Fact Sheet, online at http://files.shareholder.com/downloads/NFLX/95060669x0x24776/1F1055CB-DCBB-475A-87CD-03F6091D4901/factsheet.pdf

20 NETFLIX

Exhibit 3

Description of results from Michael Muegel’s observations:

So what is affecting the availability score? I believe it is the number of movies you rented during your last billing cycle, or possibly last 2 or 3 billing cycle(s). Your billing cycle is shown on http://www.netflix.com/BillingHistory.

The following chart tracks availability scores during the entire test period for accounts "A" and "B". As I started testing I stopped returning movies from account "A" and started using account "B" exclusively. Thus over time A's number of rentals decreased and B's increased. The chart dramatically shows how on the first day of the new billing period for "A" (March 31st), the availability score dropped from an average of 42 for the previous billing period to an average of 25 for the new billing period. Likewise, when a new billing period kicked in for "B" (April 15th), it's availability score jumped from 13 to 40.

Source: Michael S. Muegel, “An Analysis of Netflix's DVD Allocation System,.”originally published April 22, 2003. Full study online at http://dvd-rent-test.dreamhost.com

21 NETFLIX

Exhibit 4: Netflix Terms of Use: Allocation, Delivery and Return of Rented DVDs

We reserve the right to process orders and otherwise allocate and ship DVDs among our subscribers in any manner that we, in our sole and absolute discretion, determine. In addition, we will, in our sole and absolute discretion, determine the quantity of DVDs we purchase for any particular title, their location within our distribution network and the level of staffing and number of shipments to be processed at each distribution center. As a result of the operational practices described in this section, we may not always send you the top choices from your queue, ship out your next DVD on the same day that we receive one from you, or process orders from your local distribution center.

At present, our goal is to ship you the DVDs listed highest in your queue. We currently try to ship you DVDs from the distribution center closest to you so that you get movies quickly. Generally, on the same day that we receive a DVD from you, we will ship the next available DVD from your queue. In certain instances, your next available DVD will not ship until at least one business day following our receipt of your returned movie. This can occur, for example, when your top choices are not available to you from your closest distribution center or the number of shipments to be processed by the distribution center on that day has been exceeded. When one or both of these conditions exist, your DVD will likely ship on the next business day and may come from an alternate distribution center.

In determining priority for shipping and inventory allocation, we may utilize many different factors, including without limitation, the number and type of DVDs you rent through our service, the subscription plan you select, as well as other uses of our service by you. For example, if all other factors are the same, we give priority to those members who receive the fewest DVDs through our service. The type, number, mix and weighting of the various factors impacting shipping and inventory allocation will change from time to time and will be made in our sole and absolute discretion. As a result of your viewing habits and our operational practices, the actual number of DVDs you rent in any month may vary, and you may experience differentiated service during the course of your membership. Also, such service may be different from the service we provide to other members on the same membership plan. The type of differentiated service you may experience includes, without limitation, (i) the shipment of your next available DVD occurring at least one business day following return of your previously viewed movie, (ii) delivery taking longer, as the shipments may not be sent from your local distribution center and (iii) the movies you receive coming from lower in your queue more often than our other subscribers. These effects will not occur unless we are faced with limited inventory at your local distribution center or when the number of shipments to be processed by that distribution center on that day is exceeded. In our unlimited plans, we do not establish a monthly limit on the number of DVDs you can rent.

Other factors that may affect delivery times, include, but are not limited to, (i) the distance between the distribution center from which your DVD was shipped and your delivery address, (ii) the timing of your placement or adjustment of movies in your queue, (iii) circumstances impacting delivery by the U.S. Postal Service, and (iv) any unplanned downtime of our computer systems.

We receive mail from the USPS and process returns Monday through Friday, except Holidays. Generally all of the returns are processed on the same day received. We will send an email to you soon after processing your return, and another email letting you know when we have shipped your next DVD, including the anticipated day of delivery. We make no guaranty as to the shipping and delivery of DVDs and may, in our sole and absolute discretion, change our business practice regarding allocation, delivery and shipping, without notice. We may from time to time revise these Terms of Use but we will not necessarily provide you notice of the revisions. It is up to you to review the Terms of Use frequently to determine if there have been changes. We believe that our operational practices, including how we prioritize shipping and inventory allocation among our subscribers, provide a great value to all our subscribers. http://www.netflix.com/TermsOfUse, accessed February 1, 2007.

22 NETFLIX

Exhibit 5: Blockbuster Financial Data Annual Income Statement (Standardized) (millions of dollars)

12 Months Ended December 31 2005 2004 2003 2002 2001 --Restated Normal 31-Dec-04-- Revenue 5,792 5,961 5,815 5,480 5,050 Other Revenue 73 92 97 86 107 Total Revenue 5,864 6,053 5,912 5,566 5,157

Cost of Revenue, Total 2,647 2,441 2,390 2,359 2,421 Gross Profit 3,145 3,520 3,425 3,121 2,629

Selling/General/Administrative Expenses, Total 3,056 3,111 2,785 2,619 2,532 Depreciation/Amortization 231 250 268 234 424 Impairment-Assets Held for Use: unusual expense (income) 357 1,504 1,305 0 NA

Total Operating Expense 6,291 7,306 6,748 5,211 5,376

Operating Income -427 -1,253 -837 355 -220

Interest,/investment, Other Non- Operating Income (Expense), net -93 -33 -27 -44 -61

Income Before Tax -524 -1,286 -867 312 -297 Income Tax - Total 65 -37 107 107 -56 Income After Tax -588 -1,249 -974 205 -241

Net Income Before Extra. Items -588 -1,249 -974 203 -240

Total Extraordinary Items (Accounting change) 0 0 -4 -1,817 NA

Net Income -588 -1,249 -979 -1,614 -240

23 NETFLIX

EXHIBIT 5 (CONTINUED)

Blockbuster Annual Balance Sheet (Standardized) (millions of dollars)

December 31 2005 2004 2003 2002 2001 Restated Normal 31-Dec-05 Cash, Short Term Investments 276 330 233 153 200 Trade Accounts Receivable, Net 128 178 184 185 150 Total Inventory 310 517 415 452 203 Prepaid Expenses 218 193 128 170 163

Other Current Assets, Total 491 482 NA NA NA Total Current Assets 1,424 1,699 960 959 716

Property/Plant/Equip., Net 724 854 787 875 909

Goodwill, Net 809 1,139 2,628 3,884 5,614 Intangibles, Net 27 35 389 451 353 Other Long Term Assets, Total 196 268 58 76 160

Total Assets 3,180 3,995 4,822 6,244 7,752

Accounts Payable 368 722 565 757 531 Accrued Expenses 765 697 610 583 550 Notes Payable/Short Term Debt 0000 0 Current Port. LT Debt/Capital Leases 36 26 145 133 181 Deferred Income Tax - Current 14813635 7 Total Current Liabilities 1,318 1,581 1,323 1,478 1,269

Total Long Term Debt 1,122 1,120 75 409 546

Total Debt 1,158 1,145 220 542 728

Deferred Income Tax - LT Liability NA NA 9 116 112 Other Long Term Liabilities 109 231 226 75 77 Total Liabilities 2,548 2,932 1,634 2,077 2,004

Convert. Pref. Stock - Non Rdmbl. 150 0 NA NA NA Common Stock 2 2 2 2 2 Additional Paid-In Capital 5,361 5,337 6,227 6,221 6,181 Retained Earnings (Accum. Deficit) -4,836 -4,248 -3,000 -1,955 -327 Other Equity, Total -45 -27 -41 -101 -107 Total Equity 632 1,063 3,188 4,167 5,749

Total Liabilities & Shareholders’ Equity 3,180 3,995 4,822 6,244 7,752

24 NETFLIX

EXHIBIT 5 (CONTINUED)

Blockbuster Annual Cash Flow (Standardized) (millions of dollars)

12 Months Ending December 2005 2004 2003 2002 2001 -Reclassified Normal 12-31-2005-

Net Income/Starting Line -588 -1,249 -979 -1,621 -240 Depreciation/Depletion 231 250 268 241 421 Amortization 357 1,504 1,305 0 NA Deferred Taxes 35 -94 -74 18 -108 Equity in Net Earnings/Loss NA NA NA NA -1 Other Non-Cash Items 55 -33 118 1,024 878 Non-Cash Items 55 -33 123 2,841 1,227

Cash Taxes Paid 11 44 98 13 9 Cash Interest Paid 90 23 34 46 79 Accounts Receivable 35 9 6 -28 35 Inventories 191 -84 55 -222 25 Prepaid Expenses -26 -98 -5 -39 -17 Other Assets 3 1 14 83 40 Accounts Payable -341 125 -222 188 -59 Accrued Expenses -22 91 102 1 72 Other Operating Cash Flow 0-500 NA Changes in Working Capital -161 39 -50 -17 95

Cash from Operating Activities -71 417 594 1,462 1,395

Cash from Investing Activities -114 -314 -188 -1,315 -945

Other Financing Cash Flow -8 -7 0 0 NA Total Cash Dividends Paid -8 -920 -14 -14 -14

Issuance (Retirement) of Stock, Net 145 3 18 39 29

Short Term Debt, Net 30 650 -310 -190 -422 Long Term Debt, Net -21 255 -29 -34 -34 Issuance (Retirement) of Debt, Net 9 905 -339 -224 -456

Cash from Financing Activities 138 -19 -336 -199 -441

Net Change in Cash -54 97 81 -48 6 Net Cash - Beginning Balance 330 233 153 200 194 Net Cash - Ending Balance 276 330 233 153 200

25 NETFLIX

EXHIBIT 5 (CONTINUED)

Blockbuster Stock Price History

Source: Bigcharts.com

26 NETFLIX