STILL WAITING FOR VIDEO-ON-DEMAND? FORGET FAT PIPES - WATCH YOUR MAILBOX. SPECIAL DVD BONUS: WATCH THE VIDEO RENTAL GAME GET SHAKEN TO ITS CORE!
On the first Friday of every month, the employees of Netflix, a Silicon Valley DVD rental company, head to a charming local movie theater for a noontime morale boost. Dressed dotcom-casual in shorts and sandals, they scarf pizza and mill around before the main attraction: a mild hazing for new recruits and an update on the company's growth figures, which so far have been impressive.
Carlos Serrao. Netflix CEO Reed Hastings: enemy number one of Blockbuster and Walmart.com.
Today, though, founder and CEO Reed Hastings has a different message to deliver. "A year and a half ago, we found out that Walmart.com doesn't want to work with us, but work on us," says Hastings, dwarfed by a huge PowerPoint projection. There have been rumors, but this is real. The world's largest corporation, with $218 billion in annual sales — more than 1,500 times the size of Netflix, in case you're curious — is gunning for an upstart that went public in May on negative earnings. Consumers love the Netflix rental model, which lets subscribers order DVDs online, receive them by mail, and keep them for as long as they want without late fees. Walmart.com likes it so much that it's launching a nearly identical service early next year. "They're printing packaging that is essentially identical to ours," Hastings informs the crowd.
Blockbuster is close behind. The nation's biggest video rental chain recently began testing an in-store subscription service that offers unlimited rentals for a fixed fee — and, Hastings says, has invested in Netflix rival DVDRentalCentral.com. "I encourage you to take them up on their offer for a free trial," he tells his staff with a smile, "and if you forget to return the discs, I understand."
Wall Street thinks it knows how this story ends — badly. Investors still remember Kozmo, Webvan, and a few pet store fiascos. Indeed, as Hastings speaks, 40 percent of all NFLX shares are being sold short. In the month that follows his presentation, Netflix stock drops from $14 to below $6. Still, Hastings is sure he's onto something. While other video-on-demand companies build businesses around broadband, Netflix is taking a half-step toward the digital future with mass-produced DVDs ($1 each) and an old-fashioned delivery mechanism: the US mail. The company has less than 2 percent of the $8 billion video rental market, but it doubled its customer base to 750,000 this year, and Hastings projects a million customers, and profitability, in 2003. Since subscribers pay up front, Netflix also has the steadiest revenue stream the industry has ever seen. "We're public now," Hastings reminds his troops, "with $90 million in the bank. For them to pull ahead will be almost unimaginable." He brings up the final PowerPoint slide: let the battle begin.
It already has. Hastings is competing not only with Walmart.com and Blockbuster but also with the pay-per-view and premium cable channels who know the same thing he does: The average video rental experience sucks. Consider Tom and Hilary, who decide to rent a movie on Friday evening. They drive to the video store looking for Kissing Jessica Stein only to find that all the copies are out. In fact, there are rarely enough copies of a modest hit film in its first few weeks of release because of a strategy known in the business as "managed dissatisfaction." Video stores don't stock large quantities of most new releases because demand is so fleeting that it's not cost-effective — they get stuck with extras. So Tom and Hilary settle for Me, Myself & Irene, and leave feeling disappointed. The movie is crap, they're too busy to return it, and then they get slapped with a $6 late fee. Not exactly a loyalty-building experience.
Netflix offers a low tech video-on-demand alternative. Jon and Alys subscribe to a $19.95 plan that allows them to rent as many movies as they want in a month, but no more than three at a time. After they watch a film and return it in a postage-paid envelope, the company mails them the next DVD from a list the couple maintain on their Netflix Web page, which offers recommendations based on their rental history. (Liked The Royal Tenenbaums? Try Ghost World.) If Netflix is out of Kissing Jessica Stein, Jon and Alys just receive the next movie on the list. On Friday, they have Donnie Darko, Piñero, and In the Bedroom on top of their DVD player. They love Donnie Darko, mail it back on Saturday, and by Tuesday, Y Tu Mamá También arrives. It's so convenient that the average Netflix customer watches five movies a month. Some subscribers rent twenty or more. (Which is a problem: Netflix loses money on postage for households that rent more than five a month.)
The convenience inspires a fervor among Netflix enthusiasts, and that's what caught the attention of Walmart.com and Blockbuster. But it's the recommendation engine that fascinates independent film producers and Hollywood studios, which see Netflix as a marketing vehicle for off-center movies that are difficult to promote through mass media. Rather than pushing the masses toward what's new — the way Hollywood does today — Netflix pushes subscribers toward titles they're likely to enjoy. Given a large enough customer base, the Netflix model could even change the way Hollywood develops movies.
Hastings sold his first company, Pure Atria Software, in 1997, for three-quarters of a billion dollars. After that, he spent a year as CEO of TechNet, the Silicon Valley lobbying organization; in 1999 he became president of the California Board of Education, where he has pushed to improve the state's notoriously underperforming public school system. That means monthly trips to Sacramento to oversee board meetings, craft motions, consider waiver requests, and review accountability standards. After one all-day assembly, he's unwinding over a latte at a Starbucks near the state capitol.
"What motivates you these days?" I ask. He lifts his mouth from his cup; a trace of foam clings to his upper lip. "The dream 20 years from now," he says, "is to have a global entertainment distribution company that provides a unique channel for film producers and studios." He nods toward the mermaid logo above my head. "Starbucks is a great example. Howard Schultz talks about building the brand one cup at a time. I'd love to be Howard Schultz. As Starbucks is for coffee, Netflix is for movies."
"But there's already a Starbucks for movies," I say. "It's called Blockbuster." He was waiting for that. "Blockbuster is the Dunkin' Donuts of movies. When Starbucks came out, Dunkin' Donuts said, 'Three fifty for a coffee just because it has steamed milk in it? That's a niche.'"
In newspaper interviews, Hastings likes to tell the story of how Netflix was born: He rented Apollo 13 from Blockbuster, forgot to return it, got socked with a $40 late fee, and realized there had to be a better way. It's a masterstroke of anecdotal marketing, in the tradition of the publicist who said eBay was built on a fondness for Pez dispensers. Ever aware of his audience, Hastings tells me that Netflix has its roots in a computer science course he took as a graduate student at Stanford. "The professor said, 'Never underestimate the bandwidth of a station wagon filled with backup tapes,'" Hastings remembers. Moving small bits of data at light speed can be far less efficient — and more expensive — than moving that information in bulk.
Netflix opened its San Jose distribution facility in 1998, when DVD players sold for $700. VHS tapes would have reached a wider audience, Hastings realized, but they were too heavy to mail cheaply. The small number of household DVD players held Netflix back initially, but not as much as its business model: $4 per rental plus $2 total postage. "We were a typical Internet company back then," Hastings says, "an ugly financial story, with not much hope of breaking even." In late 1999, the company abandoned pay-per-view for subscriptions. Especially in the Bay Area — where overnight delivery was possible from San Jose — consumers embraced the idea of paying one fee for all the movies they wanted.
Around the same time, Bob Pisano, a former MGM executive and current CEO of the Screen Actors Guild, joined the Netflix board and set up meetings for Hastings with studio moguls like Warren Lieberfarb, president of Warner Home Video. During the next year, most of the studios signed revenue-sharing agreements that called for Netflix to buy DVDs at cost and kick back a percentage of a customer's subscription fee for every movie rented. (Blockbuster pioneered revenue sharing in the VHS era but has no such arrangement with DVDs.)
In less caffeinated moments, Hastings acknowledges that Netflix could get crushed next year. Walmart.com rolled out a test of its new service in October and plans to go national in early 2003, says spokesperson Cynthia Lin. Like Netflix, it will offer 12,000 titles, delivery by mail, three movies at a time, and no late fees. The Walmart.com twist? As usual, it's undercutting the competition: $18.86 per month.
But the entrance of Walmart.com and even Blockbuster is just as likely to expand the DVD-by-mail market as squash Hastings. So far, the biggest struggle for Netflix has been educating consumers accustomed to in-store rentals and confused by the notion of a subscription. A Blockbuster-size marketing budget would help sell customers on the idea. In anticipation of such growth, Netflix opened 10 satellite distribution centers around the country this year. If the company can hold its own over the next 12 months, Hastings will expand into videogames.
In five years, Hastings sees Netflix as a $1 billion company with 5 million customers. Along the way, he may move beyond the US Postal Service — but not until broadband penetration increases and delivery costs come down. Today, it can cost more than $30 to send a DVD-quality film over the Internet. As that figure drops, Hastings will consider going digital. "In five to ten years, we'll have some downloadables as well as DVDs," he says. "By having both, we'll offer a full service."
Hastings might see his company as the next Starbucks, but Netflix owes much of its success thus far to Amazon. Netflix doesn't just deliver DVDs, it tells subscribers which ones they'll probably like, thanks to the Netflix recommendation engine — a combination of 29,000 unique lines of code and a database of 180 million film ratings known as CineMatch. It's based on the assumption that movie viewers all watch the same narrow range of big-budget films only because they don't know any better.
When it comes to mainstream movies, huge billboards and expensive TV spots are the most efficient way to drive the masses to theaters. With off-center movies, studios spend almost nothing on promotion; mass marketing is too pricey, so they just put the film out there and hope for the best. Indie films, meanwhile, often get lost in the noise. The rental market only reinforces this pattern: The hits get promoted while the smaller films are ignored. The average video store generates 80 percent of rental activity from 200 titles. It's called Blockbuster for a reason.
CineMatch doesn't focus on the mass market; it caters to the individual. Netflix encourages subscribers to rate the movies they've viewed, and CineMatch recommends titles similar to those well liked — regardless of a film's popularity at the box office. As a result, the average renter expands his or her tastes. Seventy percent of the movies Netflix customers rent are recommended to them on the site; 80 percent of rental activity comes from 2,000 titles. This decreases demand for popular new releases, which is good for Netflix, whose revenue-sharing agreements require larger payouts for such films.
Such customization is foreign to Blockbuster, which has a massive database of rental data and no idea which films customers enjoy. But the concept is hardly unique. The ability to build one-to-one relationships with customers has been growing for years. Television audiences have been fragmenting since the introduction of cable, and Internet retailers have pushed fractionalization further along. CineMatch may not be as ambitious or as well-executed as Amazon's engine, but it's more effective, because customers are more likely to accept the suggestions. Amazon's tool has the feel of a slick Circuit City salesman engaged in the hard sell. On Netflix, the recommendations are included in the cost of a subscription, making CineMatch seem more like the geeky clerk at a small video store. He sets aside titles he knows you'll like and tells you to return them whenever.
Clearly, Netflix users are not immune to Hollywood-style marketing. Ocean's Eleven, a box office smash, is the most popular rental of the year, even on Netflix. Yet many movies on the Netflix top 100 were box office disappointments. Take Monster's Ball. It grossed only $32 million in theaters, but found new life in the Netflix after-market, becoming the site's fourth-most rented film of 2002.
Can a recommendation engine help make hits? Look at Memento, which grossed a mere $25 million at the box office before becoming the seventh-most rented movie ever on Netflix. Lantana, an Australian indie, pulled in less than $1 million in the US — about 100,000 people. As of early October, an additional 40,000 viewers had seen it through Netflix alone.
This trend has little to do with wealth or good taste or any other characteristic typically ascribed to early adopters. As the service has grown, Netflix customers have started to look more Blockbusteresque: half male, half college graduates, 69 percent with full-time jobs, 62 percent with incomes below $75,000. The real difference between Netflix and Blockbuster is access to information about a variety of films.
Already, studios have begun to recognize what this means. "Netflix offers more personalized communication," says Pat Wyatt, president of Fox Home Video, whose film Office Space is enjoying a Netflix resurgence. "The more a customer shares, the better Netflix is able to customize," Wyatt explains. "For Fox, that's great." For Paramount, it's not so good. As a Blockbuster sister company, it has no revenue-sharing arrangement with Netflix. So while Netflix stocks Paramount movies, it doesn't promote them in any way. No matter how many Mel Gibson romantic comedies you rate highly, What Women Want will never be recommended to you. The Netflix effect? What Women Want was the second-most rented film of 2001 overall but didn't even make the top 100 on Netflix.
Later this year, Netflix will introduce an upgrade of CineMatch aimed at decreasing the frequency of off-base recommendations. (Like Muriel's Wedding? You'll love Lawrence of Arabia.) It will also begin to cache 500-Gbytes' worth of rating data, which should speed up a dreadfully slow site. As the rating system improves, and audience size increases, CineMatch could start having a real impact on Hollywood. Imagine studios, with a better sense of who likes what, embracing off-center movies — and concluding that big-budget films are just a big drag on the bottom line.
Take it a step further. It's easy to see CineMatch as a tool to help studios find audiences in the direct-to-video market. But Netflix could also identify those likely to enjoy first-run films — and then send targeted emails to drive theater traffic. It could bring about a "heightened incentive to create more diverse films — even for, perish the thought, adult audiences," says Peter Bart, editor in chief of Variety.
Today's Hollywood produces films for Middle America. If our tastes truly are so homogenous, then Walmart.com and Blockbuster, which have flourished by catering to the masses, should topple Netflix. But if Middle America has been a caricature — a market of convenience — all along, then CineMatch could be a real weapon. At first glance, it seems like a small thing. So did that $3 dollop of foam on your latte.
