You'll hear more about in the coming days, but Level 3 has now gone on record with me and other news outlets to say that they, "have secured a multi year CDN agreement with Netflix." They also go on to state that, "the capex mentioned on our earnings call associated with a large customer is for Netflix and the large customer is Netflix."
Sponsored by
Updated 6:42pm: "I can confirm that we have been awarded a deal with Netflix," a Level 3 spokeswoman told TheStreet.com.
As I am getting more calls and emails than I can possibly return about my earlier blog post about Akamai and Netflix, and because both companies are now willing to go on-record just a little, I've decided to give out more details on my post. Folks may still have questions after reading this post and I'm happy to answer them, if I can. All the emails and comments I am getting from crazy shareholders threatening me with bodily harm will be ignored.
Also, it should be noted that far too many websites that have picked up on my post have taken a lot of what I have said out of context or have implied things I never said. At no time did I say Netflix is, "no longer a customer of Akamai". Nor did I imply that, "Netflix won't have any relationship with Akamai in the New Year". I chose my words very carefully and third party sites need to use those exact words and not take what I said and imply something else from them.
Netflix is shifting their current video traffic from Akamai and over to Limelight and Level 3. Akamai has not denied it and neither has Netflix. Akamai did say for the record that, "it's no secret that Netflix has a multiple vendor strategy for its video service. Its for Netflix to decide how much traffic should be carried by each vendor based on their business needs. Our focus is always to work with our customers to ensure the best possible results for their business." There is also the strong and likely possibility that Akamai still gets some Netflix video delivery business for traffic not allocated to current contracts, new traffic from Netflix's expected International expansion and other video related offerings. Netlfix is constantly in contract discussions with many of their vendors throughout the year.
Akamai provides more than just streaming and they can lose a portion of Netflix's traffic, like video, while still retaining other portions of the business. The real debate seems to be why Netflix is shifting their video traffic to Limelight and Level 3 and whether or not poor performance is to blame. Normally I would not say how I got my info for my original post, but in this case I think it is important.
Technical folks inside Netflix have been telling me that in head-to-head tests between Akamai, Limelight and Level 3, Akamai has not always performed well. They have not always been the worst, but they have not always been the best. They have also told me that they have seen performance issues on Akamai's network, specific to video.
Some of the folks I spoke with at Netflix are also the same ones who told me earlier in the year that Netflix was moving their video traffic off Level 3 and over to Akamai. Two weeks after I wrote that post, talking about the change, Akamai announced that Netflix was a new customer. So my source for today's post inside Netflix has been accurate in the past. I don't see any reason why they would lie to me.
Interestingly enough, when I wrote back in March that Level 3 was losing all of Netflix's video traffic to Akamai, no one asked me for my source. No one even questioned the news. But this time, when the shift is from Akamai back to Level 3, everyone seems to want to know who the source is.
So lets get right down to it, has Netflix had performance issues with Akamai? Akamai says no and went on record to say, "as both companies have stated publicly, I’m happy to confirm that there have never been any performance issues re our support of Netflix’s business." A corporate communications person for Netflix went on to say that, "Akamai is a vendor and continues to be a vendor for Netflix. Reports of performance issues with Akamai are in accurate; we are and continue to be satisfied with their service. Netflix has a multi-CDN strategy."
While I appreciate the quotes from the companies, Netflix's is very generic. Netflix says they are satisfied with their "service" but don't say what service they are talking about. Akamai does more than just streaming for Netflix so Netflix can be happy with one service but not happy with another and their quote would still be accurate. Netflix says that Akamai is a vendor, but I never said they weren't. I didn't say Netflix was no longer a customer and my post was specific to Netflix's streaming only service.
Also, if Netflix had issues with any vendor, would they call them out in public? I doubt it. Netflix has had some major outages over the past year, not pertaining to CDN, and they won't give out any details on what the problem was, what caused it or what vendors had the problem. It is not in Netflix's best interest to speak badly about any one provider when they have a multi-vendor approach as they then lose leverage when it comes to negotiating.
But the real question to ask here is if Netflix is so happy with Akamai's service and hasn't had any performance issues, why are they shifting their video traffic to Limelight and Level 3? Why did Netflix not only sign a three-year contract with Limelight last month, but also allow Limelight to mention on their Q3 earnings call last week that Limelight would be, "expanding their role as one of Netflix's core delivery partners". Why is Level 3 spending $14M in CAPEX costs this quarter for one customer whom Level 3 didn't call out by name, but whom they said they signed a new contract with in October for business that will ramp up in the New Year? Neither Netflix nor Akamai will say why Netflix is shifting their video traffic from Akamai to others. If it's not performance and not price, why are they moving?
Is it a performance issue? I'll let you decide. I know what I was told from some tech folks at Netflix and I also know what Netflix and Akamai have now said on-record about the topic. We could cut through all this right away if Akamai and Netflix were to go on record with non-generic quotes, but neither company is willing to do that, which is pretty typical in these circumstances.
A few others things that I thing are important to discuss is the idea that some are telling me the news could not possibly be accurate as Akamai could never lose such a big customer. Yes, Akamai is the leading CDN in this space based on revenue, but many content owners jump from one CDN to another each year. We already know that Netflix was with Limelight, then Limelight and Level 3, then Limelight and Akamai, now Level 3, Limelight and Akamai, and soon Limelight and Level 3, and possibly Akamai again as well. And that's all within a span of about three years. For many content owners, that's simply the nature of the business.
I have been sent so many links in reference to my blog post today that I have not been able to read many of them yet. But I did see one that said Jefferies analyst Katherine Egbert wrote this morning that her checks find that, "Limelight likely has probably won back all of Netflix streaming, most likely due to performance issues, contrary to what Akamai is saying". Who her source is I don't know and I have not spoken to her, so you'd have to ask her directly. But the idea thay I wrote this blog post based on a hunch or assumption is not accurate. I was given info directly from the customer, who's info has been accurate in the past and public data put out by other vendors last week only helped confirmed a change.
What I make from all of this is that Netflix has decided to shift their current video traffic from Akamai to other providers for reasons that can be debated. As I said in my earlier post, I don't see this as a trend at Akamai, and unless I do, losing the video portion of traffic for one customer, to me, is not a big deal. It would be a much bigger deal and impact to Limelight if they were to lose their share of Netflix's business.
Also, I think it is important for me to let everyone know that Akamai was very responsive with regards to talking to me about this topic. We probably traded more than a dozen emails today about my post. Almost all of what we spoke of was off-the-record, but Akamai did a very good job of reaching out to me, being proactive and responding to my requests. I wish they could have said more on-the-record, but that's the nature of the business.
I also want to go on record once again for all of the folks who keep emailing me asking me if I made money from the news I put out today. For the record, I have never bought, sold or traded a single share of stock, in any public company – ever. And for those that have asked, yes, that implies to my wife as well. I don't trade stocks, never have and I don't write any post with the intention of thinking what it will or will not do to any company's stock price or value in the market.
If you have follow up questions, please put them in the comments section. If you want to disagree with me in any way, you're welcome to, but keep it professional. Any comments that are unprofessional will be removed.
My post is now up here: "Update On My Netflix and Akamai Story With Comments From Both Companies"
Sorry I can't return all the calls and emails I am getting today about by Netflix news, but in the next few hours, I will have an updated post on the subject. Both Akamai and Netflix are going on record with me with more details and I am working on the post now. The news about Netflix shiffing video traffic away from Akamai is accurate, but the real debate seems to be whether or not Netflix has experienced any performance issues with Akamai. I'll have more on all of this shortly and hope to have my post up by 3pm ET.
Updated 3:45pm: I have written a follow up to this post here: “Update On My Netflix and Akamai Story With Comments From Both Companies”
Updated 10:26am: I thought my post was pretty clear, but just in case there is any confustion, I am specifically talking about Netflix’s current streaming business, as it stands today. I am not talking about any potential business Akamai may have with Netflix outside of video streaming.
——————————-
At the beginning of this year, Akamai took the Netflix business away from Level 3 and become the primary CDN for Netflix with at least 51% of their video traffic. While it was a big win for Akamai at the time, it appears that it was short lived. Due to poor performance issues that Netflix has experienced with the Akamai network, Netflix now plans to move all of their “video” traffic off Akamai in the coming months and will go back to only using Level 3 and Limelight.
Last week, on Limelight’s Q3 earnings call, the company announced that Netflix had extended their contract with Limelight for another three years until the end of 2013. And while Level 3 did not mention any customer by name, they also said on their earnings call last week that part of their recent CAPEX spend was, “in support of a large content customer CDN contract we just won in October,” with Level 3 expecting to see revenue from that contract beginning to ramp in 2011. In addition, Level 3 also put out a release saying that they added 1.65Tbps of capacity to their CDN network in Q3 and added five new network locations, not surprisingly, two of which were in Canada. Clearly a move by Level 3 to offer more support for Netflix’s new streaming service in that region.
I sent Akamai an email to see if they wanted to comment on my post and they said, “It’s no secret that Netflix has a multiple vendor strategy. We continue to have an ongoing relationship with Netflix but we don’t comment on the specifics of customers, contracts or other providers.” So while Akamai is accurate and Netflix is still a customer of Akamai today, it won’t be for much longer.
While I am not going to go into Netflix’s pricing at this time, it is interesting to note that Netflix moved to Akamai for lower pricing, yet then didn’t get the performance they wanted. It’s another great example in the market that the CDN business is not just about the low cost leader, but rather good performance with reasonable pricing. Naturally, one of the things you have to wonder is why Netflix would have performance issues with Akamai, yet other video customers I speak to say they don’t have any. It’s hard to know exactly why that is, but part of it could simply be that Netflix has more video traffic going over CDNs than anyone else on the web, by far.
If there is one thing Netflix is good at, probably the best in the industry, it’s measuring the quality of their streaming. They constantly send out emails asking customers to rank the quality of the video they just watched and they have so much data on what works and what doesn’t. So when they choose one provider over another, they really have the data to back it up.
I know some are going to ask me what the value of Netflix’s business is to Akamai and while I don’t have exact numbers, it’s in the $10M-$15M range. (yearly) In addition to Level 3 picking up the Netflix business, Limelight also stands to get more of Netflix business in the New Year, as Limelight indicated on their earnings call, especially when it comes to Netflix expanding further out into the mobile space. And I heard at the Streaming Media West show last week that Netflix plans to release an Android app in the New Year, which would generate even more ridiculous growth for Netflix.
Another interesting point to note is that Limelight isn’t spending any additional CAPEX to support the increase in Netflix traffic, but Level 3 said that they estimate they will spend $14M in CAPEX in Q4 just to support “a large content customer” and already spent $10M in Q3 for additional CDN capacity. That’s $24M of CAPEX in two quarters for Level 3, just for their CDN business. That shows some of the differences in network capacity between Limelight and Level 3 and also gives you an idea at just how big of a customer Netflix really is.
While there is no question that HD quality video is growing online, the rate of growth is not as big as some want to suggest and not as dramatic as some vendors make it out to be. Delivering HD quality video online is not a technology problem nor an issue with the last mile, but rather a business problem due to the cost associated with the extra bits going from SD to HD.
Yesterday, Nielsen released numbers on the adoption of HD quality broadcast video to the TV and found that only 19% of viewing on broadcast television was actually in true HD, even though 56% of all U.S. households have an HD capable set. With 44% of U.S. homes not having HD service or an HDTV, these numbers also directly affect all of the streaming device manufactures who are bringing new devices to the living room. You hear a lot about HD quality video and how streaming media is suppose to be a replacement to TV, but rarely does anyone talk about the quality differences between the two.
If only 19% of all TV viewing is HD quality video, the number for online is even less. In 2008 we did a survey and found that of the 1,000+ content owners we surveyed, less than 2% were encoding video for 3Mbps or more. The same survey in 2009 found that of the 812 content owners we surveyed, only 3.8% were encoding video for 3Mbps or more. I'll be completing a survey for 2010 shortly, but based on the initial information I have collected so far, I would expect that less than 5% of all content owners today are encoding their content in HD quality video. HD quality video simply is not growing as fast as some folks think and it's important to keep that in perspective as I hear a lot of industry people talk about "the growth of HD video" as being one of the catalysts for this industry.
The Nielsen report is also a good reminder that it does not matter what the install rate of a device, technology or service is, but rather the adoption rate. That's all that matters.
At the Streaming Media West show in LA last week, a lot of companies put out news on new product announcements, customer wins and service offerings. Here's a quick roundup of all the announcements I came across and I'll be expanding on some of these over the next week or so.
By all accounts, last week's Streaming Media West show in LA was a great success thanks to all of the speakers, exhibitors and attendees. We had the best lineup of speakers to date with over 130 executives across more than 40 sessions talking about the business, content and technology of online video. We had packed session rooms and each of our keynotes by Google and Adobe had more than 500 attendees. The exhibit hall consisted of a great lineup of vendors showcasing their products and services and many of them really helped to bring the industry together with networking events and round-table dinners. Many companies also released a lot of news around the show which I have compiled here and we had some great media coverage by the Huffington Post, GigaOm, Variety, CNET, Tech Crunch, LA Times and others. Over the next few weeks I'll be writing up on my blog quite a few in-depth posts from the show and we'll have all of the sessions archived in video for viewing later this month.
The number of paid conference attendees jumped by the largest number we've seen to date and part of that clearly had to do with moving the show to LA, which allowed for a lot more participation by local executives in the area from media, entertainment, broadcast and enterprise companies. I can confirm that the show will be back in LA next year, at the same venue, and we'll announce those November 2011 dates shortly.
For those that came out to the show and supported it, we thank you. We got a lot of positive feedback from speakers, attendees and exhibitors. My only regret is that I simply didn't have enough time to speak to every single one of the thousands of attendees and I wasn't able to really walk the exhibit floor, since I have to spend almost all my time managing speakers and venue logistics.
But with the show now over, I'm happy to get on the phone and follow up with anyone who wanted to discuss any aspect of the industry with me. If you'd like to chat or want an introduction to someone who was at the show, shoot me an email or call me at 917-523-4562 and I'll be happy to assist. Just because the show is over does not mean the networking has to stop. I also welcome any and all comments about the show and any ideas you have on how we can make it even better next year.
We look forward to seeing you at one of our next shows, the Content Delivery Summit on May 9th and the Streaming Media East show taking place May 10-11th in NYC.