Netflix IT Has to Deliver High Efficiency, Support Fast Growth
"Technology is the primary driving force" for just about everything Netflix does, says Tom Dillon, who serves as the company's COO as well as its CIO. "There's no daylight between technology and business in our company; they're basically one and the same. Whenever I look at operational challenges, I always consider how technology can meet them."
Dillon's dual role demonstrates the strength of the collaboration between IT and business at the online DVD rental company, based in Los Gatos, Calif. In fact, the connection goes further down the chain of command. The chief product officer, for instance, also has a major role in Web site development.
Dillon, who has experience in both operations and technology, started at the company as the COO and took on the CIO job when the person filling that slot left. "There are major benefits to having both roles, since we try to solve most operational challenges with technology," he says.
Given the company's rapid growth, and its strategy to ensure that it continues to grow, Netflix needs tight alignment between IT and corporate strategy. While it is still a small company, its growth since its launch in 1998 has been dramatic.
From 2002 to 2004, the company's subscriber base jumped from 850,000 to 2.6 million, and it's projected to reach 4 million by the end of this year. During the same period, its net income leaped from a $21 million loss on revenues of $153 million in 2002 to a $22 million gain last year on $506 million in revenue.
Further growth will depend on several factors: maintaining its informative and user-friendly Web site, building its already competition-beating DVD library, and maintaining quick DVD turnaround times. Meanwhile, fending off competitors like Wal-Mart Stores Inc. and Blockbuster Inc., whose August 2004 entrance into the market softened Netflix's bottom line last year, requires keeping prices and costs low.
Take the challenge of fast turnaroundshow long it takes from the time consumers return a disk to the time they receive the next one. This is an essential metric if Netflix is to be able to lure people out of their favorite bricks-and-mortar video rental store, where immediacy is the primary benefit.
So Netflix tries to send out the next video within a day of receiving the last one. Part of the solution is decidedly unaligned: The company simply bought up real estate and created 35 service centers nationwide. But the geographic strategy had to be aligned with IT-developed processes to speed the flow of DVDs as they enter and leave the centers.
Because the proprietary automation process is strategic to Netflix, it's a trade secret. But in general, when a disk comes in, it is first checked to make sure the right disk is in the right sleeve, and the serial number is scanned.
Then Netflix's proprietary software considers the total inventory of the title, the items on customers' wish lists of movies they want to see, and other factors. Then it either gets sent out again, placed in inventory or retired. Netflix says that it is able to check in a DVD and send out a new one within one day more than 90 percent of the time. And Dillon expects the systems to be able to scale up, although he admits that unexpected bottlenecks have plagued the company. "Bottlenecks are not something we can predict. We have to deal with them as they come up on a case-by-case basis," he says.
As to pricing and library size, such issues can be attacked with brute force, meeting the first challenge by cutting profits to the bone and the second by buying up DVDs willy-nilly. But Netflix augments those strategies with technological finesseallowing the company to withstand a price war with Blockbuster, which has much greater resources for such a battle, and forcing Wal-Mart to give up its online rental efforts and turn over its insignificant subscriber list to Netflix in May.
When Blockbuster entered the market last year, it aggressively priced its most popular subscription $7 below Netflix's, forcing Netflix to drop from $21.99 per month to $17.99. Blockbuster's size may allow it to withstand large losses for a time, but Netflix had to be more cautious. To sustain lower prices, Netflix would have to reduce costs. And that's where technology comes in. "Our theory is that by applying technology, everything can be made more efficient," says Dillon.
Netflix, for example, is flooded with 100,000 new disks a day. 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. Through such tactics, Dillon says he has been able to reduce labor costs related to receiving by 50 percentand he's aiming for 75 percent.
The price war was a factor in Netflix's smaller projected income this year of less than $11.9 million, but the company is doing better than Blockbuster, which showed a net loss of $57.2 million for the second quarter of 2005. In fact, Blockbuster recently hoisted the white flag, raising its subscription price in August to parity with Netflix.
While price is important, customers primarily come to Netflix to take advantage of its huge library. Netflix claims to have 50,000 titles: Blockbuster claims 40,000; neighborhood rental stores generally stock about 2,000. At first blush the battle of the libraries would seem to be one that's fought with cold cash: Whoever buys the most titles wins. But without IT-business alignment, Netflix's library size might give it bragging rights, while offering customers very little. "It doesn't matter very much to customers how many movies we have if most people only rent the top 100," says Neil Hunt, chief product officer at Netflix.
So the Web site is being continually tweaked to "generate traffic to more movies that otherwise may have languished in our library," says Hunt. Besides displaying Amazon-like recommendations when subscribers enter the site, Netflix recently introduced a wizard that lets users rate movies and then receive recommendations based on those ratings. Since its introduction, the average queue length for new customers has grown from 4 to 8 titles, Hunt says, while the number of new releases rented by members has dropped to around 40 percent.
So far, Netflix is the clear winner in a market of two. Like Amazon.com Inc., Netflix enjoys a strong online reputation and a very technology-friendly culture. As Wal-Mart and Blockbuster learned, name recognition offline doesn't necessarily translate into online success. And the technology needed to support stores is far different from that needed online. "Netflix is far more established in people's minds with this service, and it knows how to use technology to fend off much of what the big offline names are throwing at it," says Dennis McAlpine, a securities analyst with McAlpine & Associates.
Still, Netflix has a hard road to travel. McAlpine points out that Blockbuster has done a poor job of combining its on- and offline operations so far. But if it can figure out how to offer in-store and online rentals for the same pricea service which would probably require new technologyit could become a formidable competitor. An even more serious attack may come from Amazon, which is rumored to be considering a foray into this market. "Amazon has a lot of click-throughs. If they can divert some of them to an online DVD rental business, Netflix will have a fight on its hands," McAlpine says.
And no one knows what role new technologies, primarily downloadable movies, will play. Netflix is trying to stay ahead of the game. This year, it invested up to 2 percent of revenue developing a video-on-demand service that will offer movies downloaded onto a personal video recorder sold by TiVo Inc. and others. Netflix is expected to test the program with a limited number of titles and subscribers next year.
In the past year, Netflix was able to fight two retail dragonsWal-Mart and Blockbusterslaying one and wounding the other. But more dragons are in the offing. What will be required to maintain its impressive pace? More of the same, says Dillon. Current applications and processes should be able to handle the business at least until the company reaches 15 million to 20 million subscribers, although it will need some tinkering. "I think we're okay as long as we continue to make gradual improvements to our systems for the next five years. That's about as far in the distance as we can look right now."
The Role of Standards in Cloud Security
Security is often cited as a primary cause for concern...Watch Now
Ensuring Resources for Mission Critical Workloads
Application workloads can thrive in cloud environments,...Watch Now
Improving Security in the Public Cloud
One of the main concerns about moving data to a public...Watch Now