As Google set a new all-time high Monday afternoon, my thoughts turned to one of the few companies in my portfolio that has significantly outperfomed Google: Akamai (NASDAQ:AKAM), which will report earnings for the quarter on Thursday.
(Coincidentally, I have advocated before that Google (NASDAQ:GOOG) should buy Akamai. Instead they used part of their big pile of cash to pick up YouTube, which I think makes plenty of sense given the weakness of Google Video.)
I don't have any idea whether AKAM or will meet or beat its numbers, though it's been folly to bet against them in the recent past.
But I do know that the Street will be watching closely. Akamai is one of those extremely rapid growers that everyone seems simultaneously infatuated with and very nervous about. Back when I was downsizing my margin exposure in the light of usurious interest rates, I sold a small portion of my Akamai holdings . But regardless of how fast it has climbed, and the trough from which it ascended, I can't get over thinking that this company remains in the sweet spot as broadband internet interactivity and rich media continue to grow (and as performance since then has indicated, I shouldn't have sold any of my AKAM).
For those unfamiliar with the company, Akamai essentially acts as a traffic manager and content server for the Internet. They place servers around the world, within ISPs in many places, that serve up software, images, video and other big files (as well as plain old interactive websites) in a way that is faster and much more reliable than a single centralized server solution.
Tom Leighton, co-founder of Akamai, summarized the capacity of Akamai in their analyst meeting last month:
The Akamai system is the world's largest on-demand distributor computing platform. It delivers all form of web content and applications for over 2,000 customers and 20,000 domains. We deliver content and applications from over 20,000 servers in 2,800 locations in 660 cities, over 70 countries. The servers themselves are physically located in over 900 different networks.
Akamai essentially offers four things to its customers, whether they want to run a network for their affiliates and franchisees, sell direct to consumers, serve big media files, or sell software as either a download or a service: speed, reliability, analytics, and security.
Mike Afergan, the Chief Technology Officer of Akamai, explains the basic premise of the company:
Instead of sending content through the core of the internet, where capacity is low and cost is high, we send a theoretical minimal amount of content across that core, instead choosing to distribute our content from the edge of the network, where capacity is high and cost is low.
For the end-user, that means when your computer goes out to your ISP's servers to look for the Ford website (just an example -- I don't know if Ford is a client), the Akamai server that's colocated with your ISP already has most of the site loaded and can quickly get any personalized information you need through essentially a pre-screened channel to the main Ford.com servers, and it also already has the Flash or Ajax program loaded on that local server that will allow you to customize and see your dream car online, so that big chunk of data gets to your computer much faster. This ensures that any necessary downloads go through quickly, and prevents the long lag times that we all know cause (increasingly impatient) Internet users to quickly give up and move on.
There are a few concerns with Akamai, and they've come up again and again over the past year or so as the shares have hit nosebleed growth levels but still trade at discount PEG ratios. Many folks want an excuse to sell Akamai after this runup, and these are a few that I've heard:
1. Net Neutrality will kill Akamai by building an open market for bandwidth and allowing content providers to buy plenty of bandwidth from network owners to do away with the need for Akamai. That seems utterly foolish to me, as Akamai doesn't sell bandwidth or buy it as their core competency. The core service offered by Akamai is speed, but they provide speed by reducing physical travel of data and optimizing data transmission across distances, not by selling redundantly huge bandwidth capacity -- which is useless unless it goes straight from the provider to the consumer without hitting any bottlenecks.
In their analyst day last month, CEO Paul Sagan said of the content providers and the bandwidth/network owners, "neither of them is hostile to Akamai, because we are actually improving performance for both constituencies and doing it more efficiently for both."
So regardless of the fact that politics is likely to prohibit any signficant changes to the way the Internet currently operates, Akamai should have ready customers and partners regardless of any likely changes that the net neutrality debate may bring.
2. BitTorrent will eat Akamai's lunch. BitTorrent and the similar peer-to-peer content delivery and optimization networks are certainly important players in a segment of this market. They can indeed speed up delivery of some content, and they can distribute data across the web with some efficiency. But this is a little bit like saying that StarOffice will eat Microsoft Office's lunch. The big corporations and others whose businesses increasingly rely on fast, reliable content delivery are not going to be comfortable using the modern equivalent of Napster to serve up their files. BitTorrent is good at what it does, but in terms of security and reliability (and even usually pure speed, to the best of my knowledge), there's simply no contest. Like it or not, the corporate Internet isn't often going to rely on peer-to-peer systems or open source code if it's not as good as (or controllable as) Akamai's proprietary system and software.
A recent IDC study found that Akamai's service paid for itself with increased efficiency and utilization in only 1.8 months, and other performance metrics for e-commerce sites show the clear benefit to saving even a second or two during a customer's shopping episode. I don't think companies are that eager to avoid paying for measurable performance like that.
3. The web infrastructure is so much faster now that Akamai will soon be unnecessary. I think that is a trifle optimistic -- and it assumes that the capacity will grow evenly, and faster than the content and interactivity load that's being moved to the Internet. While the core of the Internet, which is the big, slower heart of it all that Akamai generally tries to avoid, is indeed likely to move faster as more and more capacity is added, I think content providers will continue to be the ones pushing the envelope. iTunes wasn't developed because the extra bandwidth was sitting out there on the Internet, waiting to be used -- it was developed because the consumer need was there and Steve Jobs saw an opportunity to sell a lot of hardware if he met that need. The same is true of the burgeoning video and software delivery services. The companies that find and develop markets are going to move faster than the capacity grows. No one waited for the development of the interstate highway system to build a faster car.
It may sound right to say, "If you build it, they will come" about excess bandwidth or delivery capacity. But in reality, they're coming and coming and coming and you're going to need to build fast to keep up. The idea that there's excess bandwidth just sitting around to be used whenever anyone needs it may have had some credence in 2000, when there weren't many big files moving online and the physical buildout was so aggressively fueled by the silicon boom, but in my opinion, that was an aberration. Now the companies that provide broadband-intensive services or downloads are profitable and growing, and consumers are clearly more than ready to buy movies and software online if it can be delivered quickly and easily enough.
4. There are so many competitors getting big venture funding now, what if one of them is the next Akamai and has a better algorithm? It's true that Akamai has a lot of competition in this space, from companies like Netli and Limelight and many others. However, while it's certainly worth watching this space to make sure that Akamai is continuing to innovate and hold their customers, I don't think it makes sense to assume that Akamai's position as the first mover and the largest, most capable provider is as easily assailable as some folks seem to think. Akamai already works with Microsoft, Yahoo, maybe Google (they're not allowed to say), iTunes, and many of the largest online retailers. They're making money for those companies, and my guess is that on the whole it'll be hard to dislodge those relationships.
5. Akamai's biggest customers will just build their own networks -- they're huge now, why not just do it themselves and save some money? There is some fear here -- it may be true that Microsoft or Yahoo or Apple could build similar server infrastructures around the world and optimize them in some way to compete with the level of service that Akamai provides. But why would they? These companies all have different core competencies, and in general I think companies are much more willing to pay for a service than to invest heavily in performing that service themselves. Indeed, it's outsourcing that remains the key trend of the day, not insourcing.
Akamai was closely involved in serving up the beta version of Microsoft's new operating system, Vista, earlier this year. You may remember that bloggers around the web were spreading rumours that someone at Microsoft feared the massive demand for Vista downloads would swamp the Internet and prevent people from watching the World Cup, among other things. It didn't happen, thanks in part to Akamai, and that kind of capability is not as easy to replicate as some appear to think, even if you have a lot of venture funding and even with the cheaper, faster servers that seem to roll out every day. Akamai's reputation, customer contracts and history of satisfying those customers, are perhaps as important as their incredibly huge network and their capabilities -- as long as companies continue to need the best, I see no reason why they'd try to build it themselves when they can hire Akamai and get a pretty nice return on that investment in a very brief period of time.
Akamai's history is probably working against them in the stock market -- I expect plenty of people are very afraid of (or angry at) Akamai after their share price reached $250 or so in the dot com boom then collapsed to under a dollar in what seemed like a heartbeat. But if you consider the lessons that climb and fall taught them, their relationships with the most important companies in software, retailing and web services, and the time they've had to build a technological and sales lead without appreciable competition in the marketplace, I'd be hard pressed to bet against AKAM even as I acknowledge the temptation to sell shares that have appreciated by more than 200%.
Akamai's goal is to have a billion dollars in revenue by 2010 (about 3X last year's level), and to grow at at least 30% a year. I see no reason to doubt them at this point, given the recent growth rate of 40-50%, or to believe that a forward PE of 40 or so is anything but a reasonable price to pay for a company with such excellent growth and with profit margins that, while already high, may even increase as they continue to scale up the network and client base.
AKAM 1-yr chart:
For full disclosure, I do own Akamai and Google shares.