GLG News by Dan Rayburn

Implications: In March, Highwinds announced it had raised $55 million from General Catalyst Partners and Alta Communications. Highwinds plans to use the capital to do additional build out of their network named "RollingThunder" which includes a CDN offering.

Analysis:

While most CDNs are all going after the same customers, Highwinds is taking a different approach by going after those who want to resell content delivery services to their own clients. To date, most CDNs either don't have a reseller program at all or their tool sets don't support resellers properly. Resellers need products that allow for sub accounts, reporting based on multiple directories, self provisioning and many times custom branding. To date, most CDNs don't deal with resellers well, don't have the proper customer support for resellers or don't have the tools to allow for a lot of the specifics resellers need.

For me, the real question is whether or not there are enough resellers out there to really scale a business past $30-$40 million a year in revenue. I know some are going to point to Akamai's 2007 revenue of $636.4 million and say that about 20% of it came from resellers and hence the market is there, but we don't know how much of that nearly 20% in reseller revenue came directly from Akamai's content delivery services.

With Panther Express and Velocix (CacheLogic) having announced their funding a few weeks ago and now Highwinds, that pretty much leaves only Pando Networks, BitGravity and Voxel.net as the remaining CDNs from the list of 30 who have not raised money in the past 12-18 months or are public companies.

Implications: Of all the calls I do with analysts, money managers and others tracking the content delivery market, rarely do those I speak with ask about video birates. For me, the growth of the content delivery market and the very success of the CDNs relies heavily on the trend we see developing for increased bitrates. Most think CDNs can only grow their business by signing new customers or growing their existing customer base. But increased video birates has a huge affect on any CDNs bottom line and 2008 is the year that the bitrates of old finally turn the corner.

Analysis:

Five years ago, the average broadband video stream was encoded at 300Kbps. Last year, 300Kbps was still the norm but we saw many moving to higher quality. Within the last six months, the average broadband video is now at least 500Kbps. We have all seen bigger window sizes, better frame rates and all of that comes from increasing the bitrate. In just the past few days or weeks alone, MLB.com, MTV.com, Yahoo!, YouTube, and Daily Motion amongst many others have all increased the quality of their video.

Gone are the 300Kbps streams and small window sizes. Many content owners are now doing 750Kbps streams and come this year they are going to be delivering many more bits than last year. And with most content owners using some form of a CDN to deliver their content, the CDNs are poised for some explosive traffic growth this year. The average content owner I speak to says they expect to grow traffic 2-4x times this year, without the increased bitrate. Factor in moving from 300Kbps to 750Kbps and content owners could push 4-8x more bits this year than last. Of the over 1,000 content owners who took our CDN survey, 68.9% of them said they would increase their bitrates this year above 300Kbps.

Some say HD is going to be the biggest factor for CDN growth but HD viewing adoption in large numbers is years off. HD will have some impact this year, but most content owners are not encoding content in HD and are focusing on a bitrate around 700Kbps. But even with HD having a small impact this year, it still adds to the growth that CDNs are going to see in the next few quarters. Content owners are putting up more content, in more platforms, in higher bitrates and much of that content is longer in length. This all amounts to a huge increase in the number of bits being delivered via the CDNs.

That being said, the real question is will the CDNs be able to turn the additional traffic into additional revenue or will more traffic simply mean they will have to lower their price? As any content owner should know by now, the more traffic you do over time, the lower price you pay per GB delivered. But for most content owners, it's going to take a few quarters before they see that growth. Lower pricing is not going to come overnight and for those who aren't doing huge numbers to being with, even doubling your traffic may result in just small savings. Taking a look at contracts and pricing I see in the market, a large content owner doing say 300TB a month is only going to get about a 15% reduction in price on average for doubling their traffic to 600TB a month. Commitment levels and contract lengths play a factor in that percentage, but doubling traffic does not mean pricing then gets cut in half. For really large customers who grow their traffic quickly, they could see their price per GB cut in half. But if they are doing 4-8 times more traffic, even with the 50% price cut, the CDNs are still making more money. For some of the largest customers the CDNs have the increased traffic and lower price point could cancel each other out over time, but it will take many quarters for that to happen.

One of the best ways we could see the growth the CDNs are seeing from customers is if we knew how many streams the CDNs delivered each quarter. A few years ago, many CDNs use to give out that data each quarter or at least a few times a year. Some, like iBEAM, used it as a marketing vehicle to promote how quickly they delivered their 1 billionth stream. While the number of streams a CDN delivers does not equal a profitable company, as iBEAM found out, it would at least give us a way to see what type of growth customers and CDNs are experiencing. Today, the number of streams a CDN delivers each quarter and the growth they are seeing are one of the many data points that CDNs are scared to release to the market. Scared that if their number is lower than a competitor a customer may go with someone who did more volume. Scared because they won't be able to use some of the marketing language they use today if their numbers don't match. Scared because most CDNs in general are too focused on what customers "think" they do, as opposed to what they really offer. In the CDN landscape, they all seem to be focused on "perception" rather than reality, which keeps a lot of data from being revealed.

But the problem with this stay quiet approach is that as an industry, we need the CDNs to step up with this kind of data and show the growth. CDNs don't need to list out customers by name, but there is nothing stopping them from saying we delivered X amount of total streams last quarter. We need data points in the market that we can all rally behind to show Wall Street and others the value CDNs provide today and the role they are going to play when this becomes a billion dollar market. We need to show advertisers how many streams content owners are growing by each quarter and we need data points to show everyone how important the CDN market is. CDNs, think of the bigger picture. You hold the keys to being able to show everyone what the real growth patterns are for online video.

Implications: About six months ago, I was really convinced that P2P might start to get some traction in the new year and that we would see some content owners commit to the technology. But aside from the Pando Networks and NBC deal, as we enter Q2 not much has transpired from the end of last year. P2P networks are still talking up a storm in every interview I read about how much cheaper their pricing is as compared to CDNs and leading with price as the major value proposition. Most P2P companies are still missing the point that cost is not the only thing customers care about and if it was, then P2P networks would have a ton of business by now, which they don't.

Analysis:

Some are going to say that the announcements we have seen between P2P providers and the likes of Comcast and Verizon are a big deal, but those announcements are more for press than anything else. They are not creating any revenue for the P2P providers in the market, customers are testing the waters with P2P but not committing and most P2P providers are still all using the same marketing message and not distinguishing themselves from one provider to another.

Last year, I thought the biggest push that P2P might get would be from the content delivery networks. As much as some P2P providers think they compete with them, the fact is that P2P providers can't survive on their own. Around Q3 of last year, many of the major CDNs were investigating the purchase of P2P networks, white labeling a P2P based service, or re-selling one of the many existing P2P solutions on the market. Since that time however, the major CDNs have changed their minds and have done almost nothing with P2P. Limelight, Level 3, and Akamai all have no real interest in delivering a P2P based offering to the market at this time. They don't need the product today, the market opportunity is not big enough and even for a company like Akamai who acquired a P2P based company, you wont see a P2P based product from them anytime soon. Internap announced a deal to integrate Pando Networks P2P solution into their CDN, which was the first non-hybrid CDN to do such a deal, but since that announcement five months ago we have not heard of any customer deployments.

From what I can tell in the market, P2P is not as big of a story as it was at the end of last year. The topic has cooled off a bit except when its being discussed as it pertains to carriers blocking or filtering of P2P based traffic on their networks. Aside from that, customers are not asking me about P2P and 55.2% of those we surveyed about their content delivery needs said they did not plan to even look at P2P as a delivery solution for 2008. I hate to see how hard it is for P2P vendors in the market as I believe that the technology really does provide value to certain customers and with certain kinds of content. But until the CDNs start to offer P2P as just another one of the many ways they can deliver content, I don't see P2P getting any real traction anytime soon.

Implications: Limelight put out earnings this week and once again, many analysts are unfairly comparing Limelight and Akamai numbers and data. Why is it that so many analysts and reporters are willing to make very specific statements about CDN providers, but do so using general terms? Sounds confusing just saying it but I'll prove my point in a second.

Analysis:

Yes, there is no question that Limelight did not show revenue growth from quarter to quarter and its yearly revenue guidance of ($136M) is a lot less than the pre IPO revenue guidance they gave ($179.2) for 2008. Investors want to see revenue growth quarter to quarter, but to me, who has no vested interest in either Limelight or Akamai's stock, the number of net new customers by Limelight each quarter continues to grow very well. Continued, long term, steady growth of customers is a good benchmark. Yes, being profitable is important, I get that, but this is not a dash to the finish line. This market is only just getting started and for Limelight, they are in this for the long term. And even without revenue growth quarter to quarter, the next closest competitor to Limelight in terms of CDN revenue for the U.S. has less than 25% of Limelight's revenue, so its not like anyone is bumping them from the number two spot.

I find it interesting that no one is saying Akamai only had 19 net new customer for the quarter and Limelight had 170? Why is no one comparing that and saying look how good Limelight is doing over Akamai? Because it is not a fair comparison. But that does not stop analysts and others from comparing Limelight's 15% decrease in the monthly average customer revenue, versus 20% increase by Akamai. That's suppose to be a fair comparison? What percentage of Akamai's revenue came from CDN services last quarter? How many new CDN customers did Akamai sign up? We don't know as Akamai won't break out those numbers. But did anyone stop to think that Akamai's CDN business could of been flat for the quarter as well but all of the other higher grossing products they have made up for it?

Limelight lives or dies by two things. It's CDN product, specifically for video delivery and the media and entertainment customers, which is its core vertical. Akamai offers multiple products not related to video and targets other verticals like enterprise and government. So when reporting numbers, we know exactly what product or service customers are buying. With Akamai, we don't.

Now some will say that Akamai stated on their call that a) the writers strike did not affect their business and b) they saw significant seasonal strength in media and entertainment. I agree with both of those statements, BUT, they are general statements. Did the writers strike affect Akamai. Absolutely. Doesn't Akamai have just as many if not more major broadcasters on their network than Limelight? While it did affect them, Akamai answered the question correctly when they said that it did not affect their business. Since Akamai is diversified in their product line, the writers strike had little impact on their revenue overall. But is that then fair to compare that to Limelight since they don't offer those other products? No.

And as for Akamai's statement that they, "saw significant seasonal strength in media and entertainment..." I don't doubt that. But for what product? Why does everyone assume that just because it is a media and entertainment customer that means they are doing video? Media and entertainment customers need other services Akamai offers as well. Why doesn't anyone question the general statements many companies put out and ask for the data behind it?

And before anyone writes in and says I am taking sides or am defending Limelight, I'm not. Yes, Limelight is a sponsor of the blog but Akamai has also nicely been a sponsor of the blog on multiple occasions. If the roles were reversed, I would still be making the point. It does not matter who the company is to me, it's the principle of how data is reported and conveyed to the industry and the market.

I probably saw over two dozen articles yesterday alone about Limelight's earnings. One said, "...it’s at least as likely that customers are pulling away from Limelight because of the costly patent lawsuits brought by Akamai and Level 3..." Really? That's a very specific statement, yet made in a general term with no details. Likely based on what? Customers you spoke to? Something Limelight said? Something you can point to? None of the above. If anything, Limelight's continued growth of net new customers each quarter says the opposite.

No, I don't own any shares of Akamai or Limelight. I have never bought, sold or traded their stock ever. And I know some investors are going to say that I am not a financial analyst so what do I know. That may be. But don't you think that is exactly what is needed in today's market? People who don't have any vested interest in public companies questioning what they say and asking for details on the products that make up the numbers? Anyone can look at a spreadsheet and P&L and all of that. But all of those numbers come from the products and every time I listen to a earnings call, nearly all of the questions are about ARPU and lots of financial data. That data comes from the products and services being sold. Why not ask more about them so you can see how the numbers evolved into what they are?

I'm sure some will say, why the hell do I care about this so much? Why am I always ranting about apples to apples comparison when I have no vested interest in the stock of any of these companies? The answer is simple. I want this industry to grow. I have been in this market for almost 15 years and I want to see it grow for another 15. In order to do this education, data, metrics and examples are the_best way that is going to happen. Maybe I am a fool for wanting companies and Wall Street to be more straight forward with info, ask the right questions and provide real data. I know when it comes down to it it's all about politics and the companies decide what data they put out in the market and what "perception" they want the industry to have.

My relative Sam Rayburn, former Speaker of the House said it best with the quote of "you can never remove politics from politics". That's what this industry feels like to me sometimes, politics.

Implications: Last week, JumpTV announced that it was looking to sell its content delivery network and would be "refining its strategic focus toward high-value sports and Hispanic broadcast content." This is a great example of where trying to own and operate your own network specifically for the delivery of video does not makes sense. In the release, JumpTV said "The content delivery network is currently a significant cost center for the company, and the Company believes its sale will enable it to lower its ongoing operating costs."

Analysis:

When asked by investors who their biggest competitors are in the market, some CDNs choose to say "companies who do it themselves". That may be the case when it comes to the static caching of images and HTML, but for video, nearly no content owner builds out their own CDN. Yes, Google and MySpace deliver the majority of their video content themselves, and AOL does a lot, but how many companies are truly like those three? Certainly not JumpTV.

Delivering video to a mass scale, like JumpTV was doing for over 5,500 live events in the last quarter alone, takes a lot of money, a lot of effort and more resources than most realize. Yes, it is not rocket science anymore, but it is very capital and man-power intensive. For all the investors that think any company with some money can enter the market and easily give the top CDN players a run for their money, it's not that easy. And when any content delivery network says that "customers doing it themselves" is the bigger competition they face, ask them for what service they are talking about. It's not for video.

Think about some of the biggest users of video on the web today; MLB, NBA, CNN, MSNBC, FOX, ABC, CBS etc... none of them are building out their own CDNs for video delivery. The CDNs have no major threat from content owners building out their own distribution networks for video delivery.

Implications: The StreamingMedia.com survey on CDN pricing and trends has now been filled out almost 1,000 times. I'll be keeping it open for another week or so and then we'll start to compile all the data and give away the free iPhone. Taking a look at some of the preliminary data that has been collected reveals some great data points which I will be showcasing all this week on the blog.

Analysis:

When it comes to the length of CDN contracts, customer commit levels and dual-vendor strategy, there has a been a lot of speculation in the market on what the trends are. Here is the data we collected from nearly 1,000 respondents, 25%+ of which are doing over 25TB of delivery per month.

• 60.8% of respondents said they had a 12 month contract with their CDN. Only 12.9% had a two year deal.

• 73.5% said that when they renegotiate their contract, they don't plan to change the length. Only 12.4% said they planed to sign a longer term contract.

• 52.3% of respondents said their CDN contract had no revenue or bandwidth commitment at all. 25.3% have a monthly commit, 6.7% a quarterly commit and 11.4% have a yearly commit.

• 63.3% of respondents said they used only one CDN vendor for the delivery of video and static content. 26.% use two CDNs and 10.3% use three or more.

You can download the pie chart data for the questions here:(Download CDN-Contract-Data.jpg)

Implications: As I'm sure you've read by now, Yahoo! announced earlier in the week it had acquired Maven Networks for approximately $160 million. While I see some of the synergy of the deal, I think Yahoo! paid too much. Based on the price tag, Yahoo! paid about 11x the sales revenue that Maven had in 2007. It's a good deal for Maven shareholders, but for Yahoo!, that's a high evaluation in my eyes in today's market.

Analysis:

Clearly, Yahoo! bought Maven for their technology platform and IP, but at some point, you have to also look at the buy vs. build numbers. Acquiring Maven gives Yahoo! a platform today, as opposed to them having to build one themselves, but at what cost? Considering the state of the rest of Yahoo! business, selling a software video platform is very different than the way Yahoo! has sold everything else for years. And not really knowing what Yahoo! strategy is as a whole moving forward casts doubt on what will truly become of a Maven/Yahoo! integration.

While it sounds like the product will still be branded under the Maven name, Yahoo! should re-brand this immediately and bring it under the Yahoo! brand. With the sate of flux that Yahoo! is in as a company, I think it needs to do everything it can to put forth one clear brand, strategy and core set of products. I think over time the Maven platform could be a good core product for Yahoo!, but only time will tell how successful the integration will be and whether or not Yahoo! sticks to the set of current products offerings they have in the market today. In my eyes, it's hard to for Yahoo! to say to the market that it is dedicated to any product platform, while at the same time laying off 1,000 employees. How much will Yahoo! truly support the Maven platform with additional dev work and new products features and functionality?

Implications: I've been taking a look at some of the preliminary data that we have been collecting from our CDN survey and we've already got some great data points. While I expect a few thousand respondents over the course of a month, after a few days we already have over five hundred customers who have filled it out.

Analysis:

As Akamai's Q4 earnings from last night showed, much of the "pricing war" talk in the industry by analysts is overblown. Yes, there is some competition, especially when it comes to the largest customers, but for the most part, the "price war" is not as bad as analysts make it out to be and this data point below seems to reinforce that idea.

One of the questions in the survey is: How would you assess the pricing of your current CDN contract(s) versus your last CDN contract?

62% of those who have been surveyed stated their pricing was flat year over year. That point by itself is not enough as you really need to know what kind of volume those who fill out the survey are doing, but we're collecting that data as well and nearly half of those surveyed are doing over 50TB of transfer a month, so it's good sized customers.

The point is that not every customer is getting rock-bottom pricing and not every customer knows what price they should be asking for. If they did, we'd be seeing lower prices in the market more than we are, but in many cases, customers who already think they are getting a good price aren't trying to get the CDNs to lower pricing even more, unless the customer is willing to commit to more traffic, more revenue, or more services.

Implications: In October of last year, Level 3 announced that it would offer their content delivery services for video for the same price that it was charging for transit. Many in the industry incorrectly saw this as Level 3 saying they were undercutting other CDNs in the market and entering the industry as the low-cost leader, which is not the case.

Analysis:

If I remember correctly, Level 3 has over 2,000 customers for its IP transit services. It's only natural for them to roll out a pricing plan that they can use to target current customers buying other products like transit. But instead of many seeing it for just that, too many in the industry took it as a sign that Level 3 started a "price war" amongst the CDNs. For all those who have written about Level 3's pricing and a price war, why is it that none of them that I have seen have ever mentioned any real numbers of what Level 3 charges? Where is the data behind this Level 3 "price war" that so many talk about? How come no one will give an example of what Level 3 is actually charging? And if you aren't a Level 3 customer for IP transit services, their new pricing model gives you no indication of what they charge for CDN services without transit. Too many people want to make this more complicated than it is and want to talk to the "perception" of a Level 3 price war instead of finding out what they are actually charging for content delivery.

I've seen a bunch of Level 3 quotes and RFP responses and Level 3 is not undercutting the other CDNs in the industry just to win business. They are not the most expensive in the industry, nor are they the cheapest. From what I have seen, on large volume deals (over 50TB), Level 3 is a few cents more per GB delivered on average. But of the deals I hear Level 3 winning, like the one they announced yesterday with the Democratic National Convention, Level 3 is winning business where it involves more than just content delivery. In many cases it involves back-haul, transit, VYVX services, co-location and other products. Level 3's approach to the market is that they will gladly take on the customers who need high-volume delivery and commoditized CDN but they are really targeting the more complex business that involves more than just the CDN product. It's the same marketing angle and approach to the market that Internap is using and Level 3 and Internap are more similar in their services offered than Level 3 is to Limelight or Akamai.

I expect we will see more announcements about customer wins by Level 3 like the one from yesterday that involve multiple products and services outside of CDN. Is that to say that companies like Level 3 and Internap are eating into the market share of Limelight, Akamai and others? In most cases no, as they are all going after a core group of different customers with different needs. There is no CDN on the market today that is a perfect fit for every customers needs. Yes, all the CDNs are competing on some of the very commoditized business, but most of them are now distinguishing themselves by going after different sized customers, in different verticals, with different solutions. They have all become a lot more focused in knowing who they should target and who they shouldn't. A few years back, CDNs tried to be everything to every customer. Today, the majority of them are very good as knowing what business not to go after.

Implications: About two weeks ago, Akamai invited me to spend the day with them in Cambridge to allow me to get an better insight into their business and give me the opportunity to ask management and others more about the market, pricing, application delivery and a host of various other topics. They nicely gave me access to over a dozen individuals including senior management in sales, marketing, engineering, product development, investor relations and their CTO.

Analysis:

While some of what we discussed was off the record, there is a lot that I can talk about relating to some new product functionality, sales and marketing topics and a lot about the market dynamics, trends and my take on what the opportunity is for Akamai in the market. With so much to cover in detail, I won't be able to cover it all in one post. I will give a run down of the highlights here and then over the next two weeks I will follow up with specific posts about their application delivery product, their StreamOS content management system, customer service, analytics, some details about the network and I will post their answers to some of the questions that readers sent in. Akamai is nicely working with me on the future posts to give me customer names, specific examples of how they are using the services and examples we can see in action. There will be more to come on that later this week.

After spending the day with Akamai, there were a couple of really key high-level feelings I walked away with. For starters, Akamai has a very_clear understanding of what customers want in the market across all of their product lines. When I was getting new product demos and talking about new functionality for their services, at no time was Akamai guessing on what they think customers wanted. It was very clear that they spend a lot of time talking with their customers, getting feedback, going through review sessions and finding out what they can do to help their customers grow. For every new product or feature they showed me, they also included examples of real customers using the service and talked about how it allowed the content owner to grow their business or gave them more control applying business rules around their content. Many companies I speak to in the CDN space talk about new products and services they are working on but many times I get the sense that they don't know exactly what their current customers future needs are. While Akamai is clearly working on what customers need today, announcements they will make throughout the year will showcase that they are also deploying products and services customers will need two and three quarters from now. The biggest thing I saw from Akamai was that they have a very clear understanding of the market they are in.

Another key takeaway for me was that Akamai realizes that as the leader in the space, they need to do a better job of communication data to Wall Street and to investors and need to spend more time to educate customers. That being said, Akamai's is not saying that they are now going to provide every piece of data a financial analyst wants, but I gave them some examples of how they could put out more specific data to the market while at the same time protecting their customers, and they are open to the idea. They know as the leader in the space they need to step forward and do a better job of showing the entire industry how this market is growing.

They understand that analysts are saying that they want to buy more stock in the company but need more data to be able to do that and they said they would provide more breakout on their business at their next analyst day. While I don't deal with Akamai on a day to day basis as it pertains to their P&L, I do deal with many money managers daily and hear the kinds of questions they want answered. I think that over time you will see Akamai be more open to disclosing more data on their business.

Tied closely into the data discussions is the subject of Akamai's pricing and the pricing trends in general which we covered in detail. I gave Akamai my feedback on their pricing, what I am seeing in the market and how they compare to other vendors specifically for video delivery. This was an interesting discussion as I learned that in many cases, Akamai may not be the right fit for customers who want commoditized services. When I gave Akamai examples of deals I have seen in the market and showed them that they were two or three times higher than competitors, they were not surprised. They made it clear that if a customer does not care about being able to use multiple services, doe not care about good analytics (not just reporting), does not require a good SLA, not concerned with geographic reach etc... then that customer is probably not a good fit for Akamai and they should go somewhere else at a lower price.

If there was one thing I walked away from after this meeting it is that Akamai is laser focused on customers who have business problems, need to apply their own business rules around their content, need very detailed analytics, need geographic reach, need multiple services and are trying to solve the entire ecosystem problem from creation to distribution. While they will gladly take customers who want the simple task of pushing bits, that's not who they are going after. For the most part, that is why the whole "pricing war" people talked about last year was completely overrated. Yes, Akamai may have to compete harder on price for those specific deals where the customer just wants to just push a lot of bits. But for customers who are looking for much more than that and also need additional services like application delivery, pricing is not the problem.

While I have always covered content delivery for video very closely, application delivery is now something that I am going to also focus on. In my eyes the biggest market opportunity for Akamai going forward is their application delivery product. While it is a market that is just starting out, I think that in two or three quarters it will become a very big opportunity and I think it is the most under-estimated product in Akamai's portfolio. I'll have more on that later in the week along with details on Akamai's StreamOS product, the content management system that was acquired in the Nine System acquisition. From the product I saw in action, which Akamai has been working on for the past year since the acquisition, and will have announcements about shortly, I have not seen another CDN that has a solution that can do even 25% of what StreamOS does. And as more content owners start to become more sophisticated in regards to the kind of transcoding, tracking, analyzing and monetizing they are going to need, Stream OS will quickly become a revenue center for Akamai.

Overall, I didn't see or hear anything from Akamai that gives me any indication that they are currently losing market share. Yes, for video delivery they have some competition in the market, but most of those competitors are still very small and don't have a wide product portfolio. And when it comes to doing more than just pushing video bits, Akamai clearly has a handle on the market today and more importantly the solutions that will be needed by these customers a few quarters from now.

Implications: I've blogged before about how most CDNs don't give out data points in the market or to customers for them to try and fairly compare one CDN to another. It's impossible to measure performance, capacity and many other aspects of a content delivery service offering from one provider to another. I got an e-mail this morning from a content owner who summed up what I hear all the time from customers:

Analysis:

"How do you think we should proceed looking for a new CDN and how do we get any data that will help us determine who we choose? All the CDN sales reps are blaming the other one in terms of network size, service quality, etc. and we can’t see the difference between many CDNs. Do you have research in terms of how much bandwidth, capacity, scalability, performance, etc?"

I know many of the CDNs hate these questions and say it is not fair for them to have to answer them since their capacity constantly changes and they are all making upgrades to their network footprint all the time. But why can't CDNs give some sort of metrics on capacity per format in each region of the world? Look how many CDNs say they are global when they truly aren't. Many CDNs that are in the Adobe partner program and listed on the Adobe site as supporting live Flash video don't support live. And how is it that every CDN, large and small, global and regional, has gotten an A+ ranking from Keynote on their network quality?

I am not the only one who notices this stuff. If any content delivery network thinks I am pointing out things customers don't already know, I'm not. I hear these kinds of points from customers all the time. And with more CDNs in the industry and more competition it is harder than ever for content owners to try and figure out who really does what in the space and what any one company's limitations are.

While I would propose some suggestions to the CDNs of the type of data they should talk about to fix this, it's pointless as it would then clearly show the differences between the networks. And too many of the CDNs want to be compared to all the CDNs as being on the same level playing field when in fact, they know that many times they aren't. I don't think there is anything wrong with not being as big as someone else or having as much capacity as someone else. Why would you want to tell a customer you are global, sign them up, and then have them find out the hard way that you really aren't? Why start off the relationship on the wrong foot? This happens much_too_often with customers.

CDNs should highlight what they do well, what their core strength is and what type of customers they should and more importantly should not be going after. If MLB.com were to put out an RFP tomorrow for all of their video delivery needs, I bet nearly every CDN would want to bid on that business, even though those of us in the industry know there are realistically probably only 2-3 CDNs today that could handle that level of traffic all at once.

The CDN facet of this industry has to get smarter and providers have to start evolving much faster in terms of the message they are delivering to customers. The service is already starting to be seen by many as a commodity. Now is the time to make it clear to the market and content owners what your real strength is. Looking at almost all of the CDNs websites, it is nearly impossible to find out what verticals they specialize in, what format(s) they support, what regions of the world they have delivery in (network maps don't count), what type of reporting they have, (put up a demo account on your home page), what your message is to the market, (speed, reliability and global reach don't cut it anymore), who your customers are, (case studies please) what your products are in detail (where are your product sheets?!) and for those that say they help content owners "monetize" their content, show examples of exactly what that means.

A CDN that delivers bits and who sells content delivery only with no tools or applications, that delivery service is not a monetization service. Pushing bits is not enabling content owners to monetize their content. Giving them tools or providing services to do targeted delivery, advertising, DRM and very granular reporting - those are monetization services. Simply shipping bits is not.

Too many of the CDNs are so focused on only using networking language right now that they are not delivering any real marketing message to customers. Go to the websites of the CDNs and look at the message on their home page. The majority of them, but not all, are not delivering any clear concise message with any real identity.

In my eyes, the content delivery market is going into the next big stage as we will see more growth in the next 2-3 years than we have seen in the past five years. Now is the time for CDN vendors to deliver a very clear and concise message to the industry and to customers of exactly what it is they offer and what the differences are between their company and others.

Implications: While many seem to think I am crazy for thinking anyone would acquire Limelight, it would make sense for a company like Level 3. Yes, we all know Level 3 has had some problems with the integration of all the acquisitions it has made as of late. But putting that one hurdle aside, there are many reasons why this would make sense in particular for Level 3.

Analysis:

For starters, everyone seems to think that anyone acquiring Limelight would continue to operate their network. But for someone like Level 3 they don't need the Limelight network in operation. They need their sales reps, their customers, their revenue and their hardware. Transition as many customers as possible over to the Level 3 network and shut down the Limelight network. When Akamai acquired Speedera and Nine Systems they didn't keep those networks functioning. They took the customers and terminated the networks. Same thing happened when Internap bought VitalStream.

I don't think anyone would argue that you could do all of that overnight. It does take time and requires a great deal of work, but it's not difficult considering the product Limelight is selling to customers is very straight forward without a lot of customization. And any company that acquires Limelight would probably lose 20-25% of the customer base anyway so you'd be talking about having to migrate roughly 750 customers. That's not rocket science.

Would a company that buys Limelight have to pay some sort of royalty to Akamai while they transition the customers over and shut down the Limelight network? Maybe. But they might also use the appeal process to do all of that by the time the appeal goes to court and then show that they have terminated the product that was in question.  If that were to happen, I would expect Akamai would then file suit against Level 3 for the same 703 patent, which in my eyes they have not done to date as Level 3 has not been a serious threat to Akamai yet.

Some also say that Level 3 could not do this as they are not a real player in the CDN space and don't have the network to transition the customers to. That's incorrect. They have more customers for CDN than most realize, are continuing to add capacity each quarter and will become the number three CDN this year based on CDN revenue in the U.S. They are very quickly becoming a real option in the space and with the integration of the Vyvx products and the applications they acquired when they bought Servecast, they are laying the ground work for a true eco-system offering of more than just shipping bits.

Most would say that by Level 3 buying Limelight it would make them a target for a suit by Akamai. But they are missing the bigger picture. Level 3 is already lining itself us for a patent suit by Akamai. Based on the broad interpretation of the 703 patent, every CDN is already in violation. So why hasn't Akamai gone after Level 3 or any of the others? Simple. Even for Akamai a lawsuit is a lot of work and costs money. They are not going to go after any CDN until the CDN is a real threat to them in the market and doing enough revenue to make it worth their time. Cable and Wireless was around for years before Akamai went after them. Speedera was too. It wasn't until Speedera was getting traction in the market and revenue before the suit was filed.

And look at Limelight. Limelight was founded in 2001, yet Akamai didn't file the suit until five years later, when they were doing some real revenue and had become a real competitor to Akamai. Level 3 acquiring Limelight does not make them more of a target as they are already in the cross hairs and Level 3 knows it.

When Level 3 bought the SAVVIS/Cable & Wireless CDN assets they were buying a large patent portfolio to go along with the 800+ other patents in their portfolio. Clearly Level 3 knew what they were buying, knew the outcome of the Cable & Wireless and Akamai lawsuit and they would have spent a lot of time examining what their legal exposure may be with the patents before the acquisition. Based on Level 3 buying going through with buying the assets, they clearly feel they are prepared to defend whatever comes their way.

Is it an easy deal for Level 3? No. But it's not a crazy one and with the right pieces in place, Level 3 becomes the number two CDN overnight in terms of revenue and customers. And for all the people who still want to say how successful Akamai was in the Cable & Wireless suit, remember that Akamai sued Digital Island who was then bought by Cable & Wireless even though Digital Island was being sued. And we don't know how "successful" Akamai was in that suit as Cable & Wireless went bankrupt in the U.S. before any of the rulings were appealed. The one time cash payment by Cable & Wireless to Akamai was made in the final days of them closing down operations in the U.S. and was a small enough amount that Akamai didn't even need to mention it in any of their filings. (At least not that I could find)

And even with the ruling two years after the suit started, C&W said, "The injunction is a legal technicality about a legacy part of the CDN that was abandoned some time ago". So the idea that someone like Level 3 could take what they need and shut down the Limelight network is completely possible.

The biggest hurdle I see to this is the debt that Level 3 has and the problems they have had with all the integrations in the past. Those could potentially be deal breakers that keep this from happening. But if all the right pieces fall into place, Level 3 could make out nicely acquiring Limelight and propelling itself to the number two spot in the market.

Implications: The StreamingMedia.com survey on CDN pricing is nearing its end and to date we've had 1,041 respondents. We're starting to compile all the raw data for the final report and will be giving away the iPhone this week.

Analysis:

When it comes to the factors that are most important to customers when selecting a CDN vendor, price and customer service came in number one and two. We asked respondents to rank multiple factors on a scale of 1-5 with 5 being most important.

• Price: was given a number five by 55%
• Customer Service: was given a number five by 52.7%
• Geographic Reach of Network: was given a number five by 44.7%
• Flexibility of Contract Terms: was given a number five by 33.4%
• Number of Formats Supported: was given a number five by 28.7%
• Technology and Product Road Map: was given a number five by 28.1%
• Value Added Services: was given a number five by 20.2%

While CDN customers are not buying on price alone, it comes as no surprise that pricing and customer service, followed closely by geographic reach of the network are the most important factors overall.

Exactly 50% of the respondents came from the media, entertainment and broadcast verticals. Enterprise and education made up another 30% and pharma and government made up the remaining 20%.

Implications: I talk to a lot of content owners large and small and no matter what subject we are discussing, I always ask them for their predictions on their traffic growth for 2008. While many of them share their numbers with me but don't allow me to make them public, the vast majority of content owners are telling me that they expect to deliver 2-4x more bits this year when compared with 2007. For most of them, that does not take into account the fact that this year, many content owners are bumping up their encoding bitrates from 300Kbps to 500-750Kbps. Taking that into consideration, with increased bitrates they could see their traffic growing from 4-8x over last year.

Analysis:

In most cases, nearly 100% of this content is being delivered via content delivery networks, so it's the CDNs who really make out from this growth. But the real question is while the number of bits goes up, does the price per GB delivered come down to the point of where the volume really does not make up for the reduced price? From what I can tell, even if a content owner increases their traffic by 4x over last year, the typical price break they are going to receive is somewhere around 35%. On a 200TB a month commit, a customer who is paying say $0.20 per GB delivered now, is going to be paying on average around $0.13 per GB delivered for 1000TB a month. Now that's not what ever content owner is paying, some pay more, some less, but if we use that as an average, CDNs should still see higher overall revenue based on increased volume.

Now for larger deals, that may not be the case. For a content owner who does 500TB a month today and grows to 2000TB a month this year, I expect they will see close to a 50% reduction in price based on the additional volume. So for some of the really large customers, increased bits delivered may not equate to more revenue for CDNs. For larger customers it may cancel it out. Bottom line, content owners are expecting to deliver a lot more bits this year and don't see the recession cutting back on their growth or spending.

Implications: Insight Research is forcasting that streaming content will generate almost $70 billion in the U.S. by 2013. I don't know how they come up with that number as I have not seen the full report, but $70 billion? They say the revenue prediction comes from audio and video files transmitted over the Internet, via an IPTV network or to mobile phones. They say that advertising revenue will fuel this growth and that "Questions surrounding consumers’ willingness to pay for content have been dispelled by the popularity of satellite radio and iTunes." I would disagree. Customers are willing to pay for music via iTunes, but so far, not videos on a mass-market scale. Over time, yes, more video specific content via iTunes will be purchased but you have to  back up the $70 billion number with more than just iTunes as an example. And what does satellite radio have to do with streaming?

Analysis: They also say that if pre-stream costs drop faster than expected, or IPTV or 3G takes off faster than expected "it could blow the doors off of our forecasts, propelling this industry into explosive growth." I am all up for reports that show growth and make predication based on accurate data, but $70 billion is just so far away from reality. If someone has a copy of the full report, I'd love to see how the $70 billion number is calculated.

Implications: This past Monday, Oprah webcast her second class online and peaked at 200,000 simultaneous viewers. While that was down from the 500,000 simultaneous viewers she had for her first class, it's to be expected and is not any indication of failure.

Analysis:

Having done a lot of series of webcasts myself over the years, the first webcast typically gets more traffic, more promotion and has more excitement around it since it is the first one. And with eight more classes to go, Oprah still has a lot more viewers she will be picking up along the way.

One of the interesting metrics of the webcast is that the average user stayed on for almost the full 90 minutes. Most webcasts don't have an average viewing time that long and you see a lot of people dropping in and out of the webcast along the way. Considering the nature of the content Oprah is talking about, I think you have to be pretty into it to begin with, so it didn't surprise me that the average viewing time was so long.

Some in the industry have criticized Oprah for not making it a pay-per-view event or for doing more advertising during the webcast. They are almost seeing it as being a failure since Oprah didn't monetize it the way they think she should have and some even went on to say how she didn't leverage the traffic properly. I think they are all missing the point.

For starters, Oprah did have three sponsors for the event; Chevy, Post-It and Skype. I saw commercials run just before the webcast started for Chevy and Post-It and those ads are still running where you to go to watch the archives. But the biggest point I think people are missing is that Oprah does not need to make it a pay-per-view event. Not every piece of video on the web needs to be monetized or charged for. As an industry, we need to stop being in the mindset that if we don't somehow charge for every piece of content or show enough ads with the content that it is a failure. Many companies use webcasting and on-demand video all the time as marketing and promotional platforms and are quite happy with the results.

These webcasts give Oprah a way to reach an audience outside of the one hour a day she is on TV. In many cases, she is reaching a different audience in parts of the world who can't get her show on TV. It enables Oprah to further expand her reach, get more awareness and increase her brand. She's taken her content and made it available on demand and portable via iTunes. She is using many different platforms to reach the widest possible audience on many devices. She's doing what any smart content owner would do who already has a huge audience they can tap into and easily grow.

I expect once all of her webcasts are over you are going to see her talk about just how well this platform worked for her, how many total views she got from the content and she will make a commitment to do more events on a regular basis. I think Oprah will very quickly become one of the biggest advocates of webcasting on the Internet and will become a regular webcaster.

Implications: This morning, Google indexed a filing from December 17th from Level 3 Communications who filed a patent infringement suit against Limelight Networks (Case number 2:2007cv00589) in the District Court of Virgina. I have not been able to get a copy of the filing as of yet but I was able to speak to Level 3 about the filing this morning.

Analysis: While Level 3 is not giving out many details of the suit, as is common in these cases, they were able to confirm that the patent(s) in question were acquired by Level 3 from SAVVIS as part of the sale of SAVVIS's CDN assets. Level 3 would not say how many patents are in question or give out patent numbers at this time, but they did say that the infringement relates specifically to content distribution. They would not say what kind of content distribution, video, static, applications etc... but were willing to say that the patent(s) in question have never been previously asserted in any patent infringement case.

Level 3 has over 850 worldwide patents and patent applications in their portfolio and did say that when they decided to get into the CDN business that they viewed having patents pertaining to the product as being a key factor that would help determine their long term success. At this time, Level 3 would not comment on any other networks who might be infringing on their patents or speculate on additional infringement cases.

I've put in a request to Limelight Networks and will update the post if I hear from them this weekend. I also expect to have a copy of the filing shortly and am looking to see what records exist that talk about the SAVVIS patents, which really comes from the Cable and Wireless patents that SAVVIS acquired.

Implications: Hot on the heels of EdgeCast announcing it had raised $6 million dollars, content delivery provider CDNetworks announced late yesterday that it had raised $96.5 million in private placement. For CDNetworks, this funding now gives them the ability to make a serious push in the U.S. market for content delivery of video and static content. The funding announcement also coincides with the launch of a new company website.

Analysis: To date, CDNetworks primary business has solely been in Asia but over the past few months they have been out in the U.S. market selling services and ramping up their sales team. From the pricing I have seen in RFPs, they are not the low cost leader in the space and they are not, I repeat, not undercutting Limelight and others on price just to win market share. I think all CDN providers understand by now that they can't give this stuff away just to win market share and expect to stay in business. The pricing "war" that so many analysts are "speculating" on is completely overblown in the market. Just because a new provider enters the market does not mean they are selling on price.

Implications: As we near the end of 2007, I've seen a few reports estimating the CDN space to have been anywhere between $800 million and $1.4 billion for the year. The reports I have seen do not break out what specific "CDN" products they are talking to and how those numbers were calculated. So with that in mind, here is how I sized the market for 2007 and the data I used to come up with the number. This does not include revenue numbers for P2P only based vendors.

Analysis:

The number I am talking to is for the outsourced delivery of video in the U.S. market and is specific to video delivery, be it streaming, progressive download, live or on-demand. I also looked at all outsourced CDN services and not companies who sold products or services for internal delivery, such as Cisco's CDN solution.

While not all of the companies listed below provide public data on their revenue, many of them tell me off the record what they are billing or I have other data to know their revenue. In these cases, I have grouped some of those companies together below so as not to expose data they have given me privately. In no particular order:

- Internap (run rate of about $24 million for 2007, nearly all of which comes from the U.S.)

- Limelight Networks (estimated to do about $105 million for 2007 and I estimate about $95+ million of that to be from the U.S.)

- Akamai (it is estimated that about $400-$450 million of their approxiametly $625 million $900 million in revenue comes from their CDN offering. What percentage of that $400-$450 million is specific to video and comes from the U.S. market is up for debate, but I estimate it to be about $300 million.) Note: The $900 million number I originally quoted is their 2008 revenue guidance, not 2007.

-  Level 3 (didn't do much in the way of video delivery in 2007 since their streaming product only recently launched, however their CDN product for video downloads has been around for about half the year. Estimated 2007 revenue about $2 million.)

- VeriSign (why I don't have exact numbers for VeriSign, taking the European business out of the picture, I estimate the CDN revenue in the U.S to be about $8 million for the year, the majority of which was P2P based from the Kontiki product.)

- Mirror Image, CacheLogic, Panther Express, CacheFly and Advection.NET (combined, they will do about $20 million in video delivery for 2007, U.S. based.)

- EdgeCast, CDNetworks and BitGravity (combined, I estimate these companies did about $5 million for the year as would be expected since they all just recently launched their services in the later half of the year.)

- PEER 1, NaviSite and Ignite Technologies (combined, I estimate they did about $8 million in 2007 for video delivery services in the U.S.)

- Regional service providers. While not typically not classified as CDNs as they tend to go after small and medium sized business, they still provide outsourced video delivery services and tend to focus in the U.S. market in particular. (all of these companies combined did under $20 million in 2007.)

So based on that data, the market for outsourced video delivery services in the U.S. for 2007 comes in at $482 million. Factor in an error margin of $20-$25 million and the market for video delivery services in the U.S. for 2007 was between $450-$500 million. Very different than the $800 million and $1.4 billion number that is being reported.

The real question is what it will grow to in 2008? Based on what I am seeing in contract terms, increased volume of bits, higher bitrates, etc... I expect to see the U.S. video delivery market grow to about $800 million for 2008. If you factor in the revenue for P2P delivery networks in the new year, that number could go up another $50 million.

I know some may disagree with my market size numbers and that is fine, but if you can't provide the data to back up the number you published in a report, it's hard for anyone to take that number seriously. Anyone who wants to quote any of my numbers above in a report, press release, on their website or any other format is welcome to do so as long as they attribute it to me by name.

Implications: This morning, I had about half a dozen e-mails from analysts asking about AT&T and their CDN business. Yesterday, at AT&T's analyst day they announced that they will grow their streaming and caching services by 6x and said they were going to increase their CDN capabilities. Because of this, an analyst at Cowen and Co. downgraded Akamai on concerns of greater competition from AT&T.

Analysis:

Now I am the first one to say I am not a financial analyst, I don't own shares in any company and I have no vested interest at all whether Akamai or any other stock goes up or down. And while I don't pour over the numbers like financial analysts do, I listen and hear what is really taking place in the market from customers. Numbers and spreadsheets only tell you so much.

For the past 4+ years AT&T claims to have been in the CDN market for the delivery of static content and video. Yet in that time, I have never spoken to or heard of a single customer using AT&T for any CDN services pertaining to video. I have never seen AT&T even bid on any RFP, have not seen any customers listed on their website, have not found a single customer case study, have not seen AT&T put out any data on their CDN business and have not seen anyone from AT&T speak at any conference or event about their CDN business. If AT&T is in the CDN business today, and I don't mean via the way of customers delivering content via AT&T's co-location or IP services, where is the business?

Could AT&T be in the CDN business? Yes. But are they today? No. I find it amazing that analysts are going to think AT&T is competition to Akamai or anyone else when they don't have a real offering today, and even by the analyst's own omission, AT&T would not be a competitor until late 2008. You're downgrading a stock based on what a company "may" do a year from now? Am I the only one who does not see the logic in this? It took Level 3 acquiring the CDN assets of SAVVIS and 12 months of build out just to get a basic offering out the door. If AT&T does not acquire anyone, how long would it take them to  be at even 30% of the capacity Akamai is at today? They can't do that in a year. And what about streaming support? Whatever limited CDN service AT&T has today, it does not support delivery via streaming, live or on-demand, has no support for Flash, no content management system, no video reporting etc.... all things Akamai and others have today.

In the report the analyst wrote, "AT&T’s extensive network reach could shrink the average distance between a CDN node and a customer to as little as 100 miles versus Akamai's 25O-plus miles." Ok that may be, but what does that mean for Akamai? How does that affect them? That statement alone does not say how that is suppose to impact Akamai's business. Plus, AT&T is not going to place servers for CDN services at every location in their network, just like Akamai doesn't, so it's not valid to look at AT&T's entire network and say they can leverage that for one specific service offering like CDN.

If AT&T were to go out and acquire someone like Limelight Networks, which they should do if they are serious about being in the space, then they would have a real shot at the getting into the CDN business and providing real competition to the market in the next 12 months. But if they don't acquire any company or assets, they will have little to no offering when compared to Akamai or others 12 months from now.

Also, does anyone remember back around 2000 when Quest, MCI, AT&T, Sprint and others all had CDN offerings and divisions? They lasted about 12-18 months in the market before they all decided to no longer be in the CDN business. Yes, they had a lot of factors going against them in those years, especially being in the market in the wrong time, but how many people "assumed" they would make it just because they are big named networks?

In my opinion, many analysts are too quick to listen to what vendors tell them without doing enough research to really know what is taking place in the market. Speak to customers. Look at RFPs. Evaluate pricing trends. Know what products companies actually offer. Company product to product, not company to company. Anyone can say they are going to be a competitor in any market, but they don't get any creditability in my eyes just because they are a big company.