CDN Business May Get Worse Before It Gets Better, Further Details On Pricing

With Limelight reporting earnings last night, it's now clear that the major players in the CDN space, the vendors that control the vast majority of the market share for video delivery, are all experiencing no growth. Akamai's M&E business was down and Limelight, Internap and Level 3 all reported no revenue growth for their CDN business. And with Q3 typically being a weak quarter for the CDNs and some of them setting guidance that shows no growth over Q2, we may have yet to see the bottom.

While Limelight was very optimistic that they will see growth in the second half of this year and that the CDN market as a whole will pick up, I'm not so sure that industry wide, that's going to happen in the next two quarters. While pricing still took a decline last quarter, I see the bigger impact being that traffic growth with current customer is no where near the levels it once was and many smaller content owners continue to go under. While Akamai and Limelight both talked about the future of HD and higher-quality video, more devices on the market, blu-ray streaming etc. none of that will take place any time soon on any kind of large scale to impact their revenue in the near-term.

I'm constantly asked when is the tipping point and what's the next killer app for video that allows the industry to once again grow at a rapid rate? Devices are great, but they are not enough of them yet to make any real impact on volume overall. That will change, but not in the next two quarters. Higher quality video is coming and we see more of it every day, but again, not enough volume yet by itself to push the industry forward with rapid growth. While I think we'll start to see some growth again next year, I don't think the CDN market really starts to grow again, at the rate we've seen in the past, until 2011. As an industry, we're still waiting on all these devices to penetrate the market along with higher bitrate video, over-the-top services, the broadcast of more live content and better ad integration. I don't see this taking place in earnest by next year. That's not to say they won't be any growth next year in the CDN market, but I'm afraid it's not until 2011 that we really start to see the surge in the space that we saw in 2006.

Clearly, a lot of questions still remain about the current decline in pricing and where that is headed. With that in mind, StreamingMedia.com will be rolling out our annual video CDN pricing survey that we do each year. (entire survey results, including raw data now priced at only $295)  Last time we had over 1,000 customers complete the survey and this year, I'm keeping it even shorter and to the point. The questions that will be asked include:

• 1. Which industry vertical does your company best fall under?
• 2. How are your videos being delivered across CDN(s)?
• 3. How many CDNs do you currently use for video delivery?
• 4. What is your current contract length with your CDN?
• 5. How much has your total video traffic grown so far this year, compared to last year?
• 6. On average, what do you pay per month to CDN(s) for video delivery?
• 7. How much has your video delivery pricing declined from last year or since your previous contract?
• 8. Are you paying CDN(s) on a per GB delivered model or per Mbps sustained model?
• 9. On average, what is your video delivery traffic per month?
• 10. How much do you pay per GB delivered or per Mbps sustained?
• 11. What bitrate(s) are your content encoded for?
• 12. What is the one thing you think CDNs need to do a better job of?

The last time we did this survey, I published a lot of the data for free on my blog here, here, here and here. I'll be doing the same thing again this time and hope to have all the data collected within the next few weeks.

If you think there is a question I'm missing from this list that I should add, you're welcome to make suggestions in the comments section but please note that I'm trying to keep the survey small and keep it focused only on video delivery and pricing.

Special Note: I am having the pricing changed on the 2008 CDN pricing survey from $495 to $295. That change will be reflected on the website shortly and I hope that by make it cheaper, it's now affordable for everyone that wants to have pricing data from 1,000 customers so they can see the trends.

Comments

[going the way of telecom]

This industry is going the way of the telecom business, no doubt in part thanks to the telecom guys at LLNW and L3. All they know is commodity selling and the've driven the value out of the market.

[Roadman]

The 2006 climate was the year of M&A, the 2009 climate is global defficency, at the momment geo-political events create the streams, so far we had the inaguration for 2009, everything else has been a yawn for the CDN's.

Are applications the growth drivers or consolidation? with the On2, freewheel, blip.tv thing it looks as if safety in numbers is the play of the day, the global downturn seems to dictate this strategy.

It would be good to wish the industry to move forward at a rapid pace as if there was a killer app(or a Hollywood production) that could solve the problem but market forces deem wishes as fishes.

Your prediction of 2011 growth for the CDN market feels right, untill then I believe it's about survival coupled with R&D.

Please ask Akamai if they consider Octoshape a viable partnership and if they will include support for the On2 codec in the http://www.openvideoplayer.com/ .initiative.

Many thanks...

[John]

Mr. Telecom,

Prices in the CDN industry have gotten commoditize because of companies like Amazon, AT&T getting into the CDN business. These guys have no place being there.

When everybody and their mother with 4 PoPs calls themselves a CDN and offers prices 75% cheaper than the real CDNs, it's hard for them to compete at higher prices.

In the last 18 months prices have dropped from around a $1/GB to $.50/GB (for the smallest deals).

At least LLNW innovates and offers a real CDN. Even L3 has a full service offering. Just having HTTP caching servers in a few data centers doesn't qualify as a CDN.

I think that these web sites with 10GB of video should just get a business class web host and deliver from their own server and not waste the CDNs time with that kind of traffic!

Take a look at all these video asset management systems out there now, most of them have a CDN behind them and you get into most for under $500/month.

- John

[Kieran]

@Roadman

Which On2 codec?

VP6 is already natively supported in Flash. Openvideoplayer is Flash/Silverlight based so any codec additions would have to be made in the base plugin.

[GreenField]

Just like the ISP world in the late 90's, CDN's will find themselves pigeon-holed based on thier abilities. This is already happening - Tier 1, Tier 2, Tier 3, etc.

In my opinion - the current Tier 1 players are: Akamai; Limelight, Level 3. This is today, could change quickly.

You might ask - why is my CDN not in the Tier 1 list? This is an opinion - my opinion - but it appears to be taken by many other large firms such as mine. This is not a slight against any of the other CDN names you may have heard or or may be purchasing from today. It is based on due-diligence performed during recent CDN purchases.

1.) Take the number of POP's out of the senario - they don't mean much if they don't talk to one another or are not active or replicate and deliver 'actual' content.
2.) Does your CDN offer a control panel where I can "manage" my content and get links, provision LIVE publishing points, open trouble tickets, download documentation, view reporting, download logs, etc.?
3.) Asset management - not just the storage of a video file but the actual meta data managing as well. Things like meta data creation/extraction, editing, tagging (really cool things like - speech to text).
4.) Multiformat support - Flash and Windows - sure. But can you do MPEG2 TS to a set top device? How about M4V for the iPhone? What about the next format???
5.) Self provisioning of LIVE, On-demand, audio or video, all the regular fare of static content... 'nuff said.
6.) Viability - Publicly traded or publicly available financial documentation.
7.) Advanced services such as software and application development, transcoding, colocation or managed hosting, player design/build, etc.
8.) Link security, user authentication, player authentiction, domain and IP filtering, geoblocking, etc.
9.) RSS, podcasting, etc.
10.) API's or an Application/Software Development Kit, player development kit.

Once buyers figure out that the vendors in the market are truly different we will all be better off. Keep in mind - it is not bad to be a Tier 2 CDN - look back at how many ISP won millions of dollars in DSx/OCx and collocation deals in the late 90's - despite the fact that they were not UUNet. There is always a market for CDN's that don't have these services, nor do they intend on building them in. Perhaps you only need simple, straight forward, cheap, domestic RTMP delivery and you are willing to write a little code here and there...

Its the bottom of the second inning in this market and there is/will be plenty of opportunities for all - as long as you know what you are good at (strengths) and where your CDN fits (niche).

[Todd Vierling]

Disclosure: I am an Internap employee.

There's also the hindrance of US broadband providers to consider. Regardless of the development of better compression (e.g., H.264), HD video ups the ante considerably when it comes to bandwidth demands.

As we well know, consumer broadband providers in the US are, in general, hostile to this kind of change, and have done only the bare minimum (if that) to upgrade their last-mile infrastructures over the past decade. And "mobile" is simply not a tangible market yet: US wireless providers on the whole have absurdly low caps, or absurdly high prices, on data plans -- leaving most users out of the picture when it comes to mobile video streaming.

These hurdles will likely prove to be difficult ones to cross when serious growth returns in 2011 (or later).

[Mobile man]

When are you going to differentiate and include mobile video streaming? Mobile data would be very interesting around pricing and usage: price to ingest (either in minutes or GB), amount streamed, mobile apps used, etc.

[Roadman]

@Kieran

Thanks for putting me right, I find this blog a great place to learn from as I negotiate the field.

[Chip Ruhnke]

Where will CDN growth come from?

Media companies have spent the past couple of years putting some content online with ad-supported models. Are these making huge profits? I think they are still evaluating these early efforts. Can they go deeper into their content catalogs? Not while advertisers insist on buying adjacencies on prime content, rather than demographics or viewers. But database developers are learning to match user demographics with inventory, so when advertisers are willing to buy a viewer instead of a program, a ton of long-tail content will come on line. Enterprises are also only begining to use their video and multimedia assets properly. IT vendors still don't promote asset management and still fear desktop video on the LAN, partly because they only know Cisco and Cisco doesn't make a cacheing video server. But internal corporate AV departments are delivering value, and sooner or later smart Cisco vendors will get behind those efforts, and realize that much of that content needs to be on an outside CDN. Level 3's drop in revenue is mostly wholesale to other network companies. It's not the tier-1 CDN's that fail to deliver value, it's the tier-2 and tier-3 providers that disappoint clients and the overall decline is felt on L3's wholesale side.

[Sandy]

We use CDN server to track the URL's which user tries to target on. What do CDN do with those data.
what is the advantage of those data