Dan Rayburn: 0:02

Welcome to this week's edition of the Dan Rayburn Podcast, the show that curates the streaming media industry news that matters most, unvarnished, unscripted and providing you with the factual data you need to know, without any of the hype, the pulse of the streaming media industry.

Speaker 2: 0:24

Welcome to the Dan Rayburn Podcast. I am Dan Rayburn. We're on in solo today for this special podcast where I'm going to be talking about the content delivery network market just this entire podcast. I'm going to focus on CDN. I've gotten a lot of questions over the last couple months specific to vendors in the market. Actually, I get questions all the time about cdn's which I'm going to go through, but there's just so much to talk about. It's incredible that three decades since streaming technology first first came about, how much interest there still is in content delivery networks and how we deliver video, naturally just because there's so many, so many different ways that it's taking place live and on demand. It's different in regions around the world, so there's probably something to talk about in the CDN topic, which I just love talking about as well. I collect a lot of data on. So what I'm going to do today is I'm going to go through an outline of what I plan to cover.

Speaker 2: 1:23

I've spent a couple of days prepping all the content for this podcast, looking up numbers and one of the key things I'm going to do in this podcast, and I think it's needed more in our industry overall, as many of you heard me talk about, but, in particular when it comes to talking about things tied to video delivery, you have to separate facts from opinions. Many of these CDNs are public. We have numbers to work off of and yet I continue to see posts in particular on LinkedIn and some of these reports which I'm going to talk about that don't even get the factual basics right when it comes to public companies. It's unexcusable, and if you're an analyst in this space selling a report on the CDN market and you don't even know the numbers from public companies, you should really quit your job. You're charging thousands of dollars for a report where you could get more information if you just look up a public filing from one of the CDNs. To me, that's unexcusable. So I'm going to go through some of the problems with some of the data we have out on the market, and part of the purpose of this podcast today is really to separate that fact from opinions, arguments and debates that I see taking place on LinkedIn over bad data and things not actually taking place.

Speaker 2: 2:43

You know where someone wants to talk about. You know Amazon's delivering all this video for Netflix. They're not. Netflix delivers all their own video, but they rely on AWS for all kinds of other pieces of their entire video stack. The issue here is they don't have relationships with customers and vendors. The largest customers in the world are not talking to them. The vendors don't talk to them off the record. They're guessing, or they talk to one customer if they're lucky and they suggest it represents the entire industry. So I'm going to talk about how I'm getting my information.

Speaker 2: 3:23

In preparation for this podcast, I had almost 20 calls with some of the largest customers of CDNs in the world who I talk to regularly but wanted to make sure I talked to them just before this podcast. I talked to almost every CDN vendor out there. I talked to some telcos and carriers, so I collected a lot of additional information. A lot of it is on background. There's all kinds of data and information I have on profitability and revenue size of customers contracts. A lot of that I can't give out. Some I can and will today, but some I cannot. But I'm going to give out as much as I can. The only thing I have in this space, in this industry, is my reputation. That's it. The relationships I've built up over 25 years with a lot of the vendors and content owners and broadcasters and sports leagues. That's all I have. So I'm very careful as well in terms of the information I give out. You'll never see me quote any sources anywhere on my blog, nor will I in today's podcast. I am under NDA as well, with not only vendors, but some vendors but also some of the largest content owners out there or sports leagues who allowed me to see what they're working on early, or they want me to have this information on background. So, legally, there is some information I can't give out just because of NDAs, and that's how I'm happy to work in the industry.

Speaker 2: 4:48

I'd say another important point before I go through the outline here is there are many ways to deliver content with great quality, with a great user experience. There's not only one way to do it, and when you see these arguments on LinkedIn of people saying, well, you have to use this protocol, or you know it has to be quick, it has to be um, you know HLS, it has to be one thing over another. There's a combination of video technologies in the entire video ecosystem. Where it comes to picking and choosing the best technology based on the type of content, you have the business model. You have the user. You're delivering it to their location, their device. The quality. My point is there's no one way to deliver video, in particular, with great quality, and that's it. There's no standard, not HD, not 4k. Quality is subjective to the streaming service and the consumer. That's the bottom line. There are many ways to deliver content with great performance.

Speaker 2: 5:55

Also, I'm not going to be using lots of terms today that you see in our industry, like edge, right. Who cares about the edge right Now? The reason I say that is the edge is a location in the network, it's not a service. And yet you have people out there saying we sell edge services. No, you don't. You happen to sell a service that might be at the edge, but that's very different. What you're actually selling is a delivery technology that happens to have a network topology that's distributed at an edge. But what is the edge? Many don't define that. I think of the edge as two ways. The edge is really the device that the consumer is using to consume that video or piece of content TV, phone, whatever. That's the edge device. But the edge, then, is the last mile. It's the ISP. It could be wireless, it could be fixed. That's the edge. But it doesn't mean just because you're putting content closer to end users. It's a guaranteed better quality of experience. That's not guaranteed. So my point is I'm not going to be covering a lot today around or using a lot of terminology today that is undefined.

Speaker 2: 7:08

Someone asked me why I'm giving out all my information for free. A lot of people know I've been prepping this podcast and you know, oh, you could charge for it. Put it behind a firewall? Yeah, I probably could. But you know my job, as I see it, is to inform, educate and empower others. I say that all the time. This isn't about money. For me, money isn't power, information's power. So I'm not in this for the money. I'm not always trying to monetize my content. I'm happy giving away almost everything I do for free. I do charge Wall Street, institutional money managers and others tied to Wall Street for a lot of the raw data from a CDM pricing report. Vendors buy it as well. Content owners buy it. I did give it away for 15 years for free, but about five years ago I started charging for that to give away information.

Speaker 2: 8:01

I don't believe that one's self-importance is based on what one knows, but, frankly, what one can share with others. So that's my goal. I also am going to go on record here. I think it's really important. I've done this on my blog multiple times. I've done it on LinkedIn. I do it on CNBC. Over the years someone suggested that I'm making money from the stock of CDN vendors or that's why I'm writing something in particular. So let me make this really clear I've never bought, sold or traded a single share of stock in any public CDN company ever. Even in my managed portfolios that are managed by a third party Akamai, fastly, egeo, cloudflare are all removed from my portfolios.

Speaker 2: 8:49

When I speak about CDNs, I'm completely free of any financial stock bias. Next, people ask me all the time how I track the CDN market, how I know what's happening with vendors and customers. So I think it's important to point out, as I know what's happening with vendors and customers. So I think it's important to point out, as I did 10 years ago on my blog, I am not the smartest person when it comes to the CDN market. I'm not. Some people say all the time Dan knows everything about CDN. I don't, but you know what I get to speak to the smartest people in the CDN market who are building, running the largest private and public CDN networks around at a telco and carrier level, isp level, cdn level. There's far more people out there that are smarter than I am on the topic of CDN. I eat, sleep and breathe that topic, but people give me access to their time for which I appreciate and their data so that I can help educate the industry help educate, sometimes, regulators as well and have the background information to actually set proper expectations in the market. I'm simply the gatekeeper of information that is coming from people far smarter than I am.

Speaker 2: 10:03

So what I'm going to cover today for the agenda is the following If I don't get to something you want to hear about, email me, call me, put it in the comments section on LinkedIn or I'm going to post a link to this podcast. So the first thing is I'm going to talk about my definition of what is CDN, because there's not an agreed upon definition in the market. Second, I'm going to talk about the size of the market compound annual growth rate. I'm going to talk about some of the bad data in the market and why it's bad. I'm going to give out some trends around bit rate optimization, multi-cdn customers, cutting costs, reducing volume. I'm going to give out a list of the largest customers by name of CDN services. It's incredible. No one's ever done that, but a small portion of customers make up the largest share of the market. All you have to do is look at some of the public data from previously Limelight, now Egeo and Akamai and you can see that.

Speaker 2: 11:00

I'm going to talk about DIY CDN trends. I'm going to talk about some specific ones across TikTok and Disney. I'm going to talk about DIY CDN trends. I'm going to talk about some specific ones across TikTok and Disney. I'm going to talk about some demands across topics for 4K, ultra-low latency, cmcd. I'll also cover the latest on CDN pricing from my survey. Talk about open caching. I'm going to call out some unique things regarding capacity in India and LATAM. Also, some ways that performance testing has changed in the market.

Speaker 2: 11:31

I'm going to discuss Netflix and NFL on Christmas Day, because that's going to be the largest live streamed event in the US ever come this Christmas. I'm going to talk about some of the sports viewership numbers and then I'm going to talk about some of the sports viewership numbers and then I'm going to break down vendors in the market. I'm going to go through their recent public numbers. I'm going to talk about their earnings. I'm going to talk about things that we know publicly. And then, finally, I'm going to talk about where are these companies investing, what is the tie into security compute Naturally that's where the margins are in the business outside of CDN and also just talk what I think in terms of consolidation and what's expected across the industry.

Speaker 2: 12:14

So let me start with my definition of CDN. So it's very important here that when we're talking about a particular sub segment of our industry, that whoever's talking about it or writing it defines it. It's very important. So my definition of CDN for what I'm talking about today, the numbers that I'm giving out is tied to the delivery of video small objects and large objects. That's it. Not security, not things under security umbrella, waf, ddos, bot mitigation, credential stuffing, none of that. Nothing around compute, nothing around content that has acceleration needs, like what Akamai calls DSA, dynamic site acceleration. I'm not covering any of that. So this is just a delivery of video. I don't care if it's live, on demand, streaming download doesn't matter to me. Small object delivery, large object delivery. That is my definition of CDN.

Speaker 2: 13:16

Now, the reason I give that out is because none of the CDNs that are public even define, when they're giving out revenue numbers, how they bucket bit delivery the same. So, for instance, fastly, their CDN revenue falls under what they call network services. They break out their revenue for the first time ever last quarter under three buckets network services, security and other. Akamai breaks it out under three buckets of delivery, security and compute. Now you'll notice Akamai didn't say CDN. So for all the times you see someone on LinkedIn writing Akamai had X number of revenue in CDN. Well, no, that's not accurate. It's not what the company said. Even worse, I see a lot of people take these numbers and they throw the word video after it. So-and-so, cdn had X percentage or X number of video revenue. None of these companies not a single company breaks out what percentage of their revenue under delivery or CDN is tied to video. Akamai I believe it was two years ago did break out in their investor relations day the total percentage of bits going over their network that were video. It was over 50%, but they didn't tie that back to revenue. So anyone who puts that number out there, it's not a real number.

Speaker 2: 14:43

Let's go into the size of the market this is pretty wild what I see out there all the time and the compound annual growth rate. So, first and foremost, the market is not as big as many suggest, full stop. It's very important to understand that those that are writing about the size of the CDN space, they don't define it. They include non-CDN services. You look in these reports or these outlines of the reports all of a sudden includes hardware hosting, storage, transit, transcoding. That's not delivery, the market.

Speaker 2: 15:26

So every day through Google News Alerts I get some sort of news alert that some company who sells research that I've never heard of has put out a report saying the size of the CDM market today is somewhere between $20 to $40 billion and growing anywhere between 10 to 30% compound annual growth rate. I've never heard of the vast majority of any of these folks. The other thing is, you'll notice there's almost never an author's name on the report. Pound annual growth rate I've never heard of the vast majority of any of these folks. The other thing is, you'll notice there's almost never an author's name on the report. That's a red flag. But also even from some of the companies you have heard of who I'd like to call out today, but I'm not going to because I don't even want to give them visibility who are doing a terrible job in the market. They are still listing out a company like Limelight Networks in their report. Limelight Networks changed their name two years ago, this month, two years ago. So you're putting out a report in 2024 and you don't know that Limelight Networks changed its name two years ago Tells you, right there, you shouldn't be buying the report.

Speaker 2: 16:25

Verizon's gone. I still see reports listing Verizon Imperva that's not a CDN. Centurylink changed its name to Lumen four years ago. They're still listed in reports. Mac CDN this is the one that's even worse was acquired eight years ago. I still see them listed in reports.

Speaker 2: 16:45

At&t and Rackspace are not CDNs. They don't have one. They resell Akamai and other third-party CDNs in the market. So these reports are absolute scams. Do not buy them. If you want to know the size of the market or what's going on, I will gladly give it out to you free of charge and I will show you the numbers and where it came up with them and what the methodology is. And in many cases I'm going to point you directly to multiple resources where, even if you don't know the exact number, you have enough background information to come up with a very realistic one.

Speaker 2: 17:22

Everyone says, well, amazon breaks out no revenue tied specifically to CloudFront. That is true, but you know Amazon used to put out press releases on a number of customers that were paying and using CloudFront as a service. We have base numbers to work off of. So, very important, you don't buy reports in the market Now. The market by the market again, I mean CDM market, excluding China. Everything I'm referencing today does not include China delivery in China. There's a long list of vendors specifically in China right Alibaba, baidu, baixing Cloud, byteplus, cd Networks, china Cash, china Net Center, china Telecom Edge. Next, which is owned by Baixing Cloud Kingsoft. There's quite a few. You can see them at cdnlistcom where I track all the CDNs. So everything I'm talking about today numbers included, size of market pricing does not include China at all.

Speaker 2: 18:33

The market last year for these services grew about 1%. That's it. 1% Now, that's not a guess. We have the numbers. We have enough companies that are public of a certain size that combined make up the largest share of the industry from a revenue standpoint. We have their numbers, we've seen them. So these reports suggesting the market's growing 10% to 30% ridiculous. This year. We think the market will stay pretty flat. It'll grow one or two percent, that's it.

Speaker 2: 19:10

I've also seen a report recently where they said the market this year is $30 billion, growing to 60, and they're including commercial CDNs and DIY CDNs. What? Why would you include Netflix's internal CDN? They don't charge money for that to any third party. They're not making revenue from it. So how can you include a size of the market with DIY CDNs and commercial CDNs together? Of course you can't. Makes absolutely no sense. So you have to look at what you're buying. If you're looking at any of these reports Also, this was a report that came out two days ago.

Speaker 2: 19:51

This report is talking about CDNs and it's talking about how the trend is. Bitcoin token powered, layer one blockchain, decentralized content delivery network. That's all their terminology, right, which is garbage. It's throwing a whole bunch of marketing terms in. This report says the rising 5G penetration is driving market expansion due to the proliferation of smartphones and mobile devices. Of course, that makes no sense either In certain countries and regions. Yes, we're seeing more smartphone adoption and changing in pricing for packages and data, like we've seen in India, but if you add up the revenue of all the CDNs, which I'll break out later, excluding China, the market was just over $5 billion in 2023. Just over $5 billion. I put it at between $5 billion to $5.3 billion and I'll break out the numbers for you. Separately from that, the CDN operator platform market those that are selling software to operators to deploy and manage their own CDN. That was under 100 million last year. That's the numbers. Anyone who wants to argue and say, no, no, the market's much larger. They have no numbers to stand on. We have enough of these numbers and those that I do see debating the numbers and saying they're larger. Then I look and it's like well, they're also including hosting in it. I literally sometimes see people where they're including Squarespace and Wix. Those are not CDNs. They're actually using third-party CDNs to do a lot of that. So that's the size of the market. I expect it to grow one to 2% this year, pretty much flat with last year.

Speaker 2: 21:45

Now a small number of customers make up the largest share of revenue. Again, this is a fact. If you look at Limelight and then Egeo's latest earnings report, which is from a couple quarters ago, they discussed how many customers make up what percentage of their revenue. Many customers make up what percentage of their revenue, and it was under 20 customers made up more than 70% of their overall revenue. We have data points in the market to show that a small subset of customers actually make up the largest percentage of revenue overall.

Speaker 2: 22:19

Now, who are these customers? Well, I'm going to tell them to you right now. So these are customers. I had to pick a threshold on how much they're spending, because just the word large what does that mean? It means something different to me and something different to you. So pick a threshold More than $100 million. These are customers spending more than $100 million on CDN this year with third-party CDN services.

Speaker 2: 22:43

Now, note I did not say video, I said CDN. So, small object, large object software downloads, streaming, live and on demand. They're spending over $100 million in CDN. These are the following customers this is not a complete list. This is a large portion of the list though Prime Video, which is not a company, it's Amazon, but it's for Prime Video. Disney, comcast, paramount Global, warner Bros, discovery, tiktok, microsoft, apple, sony Entertainment Network, nintendo, riot Games, valve, ea, activision Blizzard, valve, ea, activision Blizzard, take-two, interactive, roblox, ubisoft, roku, nfl, major League Baseball, esl, spotify, reliance, which is Geo, tv18, viacom18, snap, the Times Group, reddit.

Speaker 2: 23:42

Now some are saying well, dan, how do you know that? Well, for starters, some of these customers are listed in the public filings of third-party CDNs because if they make up more than 10% of the revenue. They have to list them by name. Second, I talk to all of them pretty much so I know what they're spending individually. Again, I won't go into that. But also, third, again, public sources are out there. When Reddit planned to go public inside their filing to go public, they said that they plan to spend at least you ready $385 million on cloud and delivery delivery services by September of 2026.

Speaker 2: 24:24

We have data points out there. So that's a list. A lot of the largest customers out there. Again, it's not a complete list. That's not everybody I know. I'm missing some more gaming ones, some software ones. There's some others out there, maybe LATAM as well. I am going to put up a list on my blog in a separate post when I get to it. I am going to put up a list on my blog and a separate post when I get to it, but that's the largest majority of them for sure. Also, keep in mind that a lot of these are there's a lot of different brands underneath it, right? Disney has a lot of different brands under it, so does someone like Fox are part of a much larger contract that these parent companies might have with CDN providers, and many of these contracts just so listeners are aware, in case you don't know, they're revenue commits. A company will go to a CDN and say, okay, we're going to agree to spend at least $200 million, whatever it is this year, with your company across CDN and say, okay, we're going to agree to spend at least $200 million, whatever it is this year, with your company, across CDN security compute, whatever the services are that they offer. These are largest customers do revenue commits? Many times it's not bandwidth commits, so the contracts do change.

Speaker 2: 25:43

Next let's go into DIY. So there's been some talk lately I've seen of oh you know, diy is coming up again and there's folks working on DIY. I don't know of any company right now globally working on building its own CDN. The only one that's been continuing to build out right now is TikTok. Tiktok was the latest. Before that it was Disney. Tiktok has a DIY CDN Now. To date they've only deployed it in APAC and it is expanding slowly. Now Akamai actually talked about this on their public earnings call, on their last quarterly earnings. They didn't say TikTok by name, they couldn't mention the customer, but they were referencing TikTok and what they said was to date we haven't seen this customer's DIY CDN impact the percentage of traffic share we have, but we do expect that to impact us at some point going forward. So TikTok is expanding slowly.

Speaker 2: 26:41

That said, it is not a replacement for third-party CDNs. It's a complement to them. It will not replace the third-party CDNs that TikTok is using around the world. It's just like Apple. Apple has their own Apple Edge Cash program that they're deploying out there. I cannot go into details. I'm under NDA with Apple so I can't go into details around what they're doing. But Apple views their CDN as a complement to the other third-party CDNs that TikTok Apple, many of these DIY CDNs. Yes, they're looking to bring more traffic to their network, but in specific regions where it makes sense.

Speaker 2: 27:19

This idea of one size fits all there's no such thing. All these companies are global now. It's not in the old days, even 10 years ago, where it was mostly North America, north America. These companies have gone international. You're delivering content all over the world. There are certain regions you can't go into for data privacy reasons. That's another issue where you may not put out caches as a result. So there's all kinds of reasons to deploy your own CDN in certain locations but not in others. But you have to be to a certain size and scale. So that is the way to think about DIY.

Speaker 2: 27:57

I don't know anybody else right now building out DIY. Disney is going to increase their investment in DIY, but again, it's not going to replace the CDNs they have. I'm not going to go into numbers, but it's only a small percentage of their overall traffic and, from what I understand, their DIY is starting to show a profit on paper, which means, okay, if you're seeing those positive results from the CapEx and OpEx that you spent, okay, why not expand it out a bit? But it's not a replacement and that's important for people to understand. Now some thought because of Akamai and others talking about how they lost a large chunk of traffic and again they didn't say TikTok, but that's who it was from a customer it was because, okay, the customer took all this traffic in-house. No, akamai actually came out and said that was not the case. So we have the data, tiktok in particular. They optimized their encoding, they changed how they were doing prefetching and downloading content from the origin and, as a result, the volume of bits going over the third-party CDNs declined. Akamai actually came out and discussed prefetch during the earnings call. So if you're not listening to earnings calls from CDNs. That's a mistake. You need to because they actually talk about a lot of good information not in every single one, but in a lot of them.

Speaker 2: 29:20

So I don't know of any new DIY CDN underway, not even at GEO for the cricket stuff. Some people say like, oh well, you know, the sports leagues will do it, like NFL or Major League Baseball. Well, no, the NFL doesn't deliver their own stuff. The NFL outsources their content from a licensing standpoint to all the different avenues. The only thing they actually control is NFL Plus. That's not that much traffic and certainly not enough to build out your own CDN. Mlb is a great brand but, believe it or not, mlb doesn't have a lot of traffic. It's not that big. Believe it or not, mlb doesn't have a lot of traffic. It's not that big. So I do not know of any new DIY CDN underway right now.

Speaker 2: 30:13

Let's talk about what's impacting CDNs. Almost two years ago, I wrote in detail that many streaming services were focusing on doing a better job of optimizing their bid rates and, in some cases, reducing their bid rate ladders to save money. Disney, hotstar we saw it from BBC. We saw multiple companies on their own blogs talk about what they did to optimize and how much money they received, and naturally that's going to impact CDNs a couple quarters down the line. So we knew this was coming. We saw it I mean, I was documenting it. So were some of the actual content owners, and yet many are only just now seeing that show up in some of the revenue numbers of public CDNs and they seem surprised. They're shocked by it. Now, that doesn't surprise me.

Speaker 2: 31:02

As a general rule, wall Street has very little understanding of what's going on with the CDNs. They really do not take a lot of time or effort to understand what's taking place and they think if they talk to one customer or do one customer check, that represents the industry. That doesn't count. That's not everybody. There's a few people out there who are good, but as a general rule, wall Street has very little understanding of what's going on with CDM vendors. Now, given all the cost cutting everyone is doing in anticipation of current recession, macroeconomics, everything taking place, streaming services are just cutting back on the number of bits they're delivering in many cases and in many instances they're not growing the number of subs they have or they decline. Not everyone's growing the number of subs. Sling TV hasn't, hulu hasn't. It's gone up and down but it hasn't grown over time. So CDNs I've spoken to and customers are taking the approach of we need to do more with less. We need to optimize everything.

Speaker 2: 32:10

Now, in addition to CDNs, I talked to a lot of ISPs, especially during live events Peacock, wildcard game, super Bowl During those events I've got ISPs messaging me, texting me in real time about what percentage of traffic on their network is coming from where live. Is it transit? Is it peering? How are they doing this? So I talked to a lot of ISPs. So I talked to a lot of ISPs. You know, one ISP in the US didn't want to be named, of course told me that since the beginning of the year their combined peak ingress traffic so that's IP, transit, public and private peering, cdn, caching, all of that. But is that the time of the year? It could be that not new content coming out, not sports, right now in the summer, and they said it's not unusual to be down or flat this time of the year, but even with the subscriber growth they're down for the full year and they're thinking their year-end traffic will be lower this year as we're still negative for the year going into the summer. It's interesting that's not just one ISP, so they're saying the same thing, which is overall rate of fewer video bits across our network this year. Not surprising, and tying into that.

Speaker 2: 33:28

Next topic is 4K 4K there is very little bit delivery taking place at 4K in any volume. Multiple CDNs have told me of all the bits they'll deliver this year, 4k UHD will make up 5%. That's it. Now. What you really want to ask is okay, but was it 1% last year, going to 5% this year? That's growth. No, most are saying it's just flat. Ask is okay, but was it 1% last year, growing to 5% this year? That's growth. No, most are saying it's just flat.

Speaker 2: 34:04

Multiple OTT services have told me which I reported on the blog last year that the 2% or less of their total viewing time was in 4K UHD content and that was from one of the largest OTT platforms in the world. Macs and others charge more for 4K. Yeah, it makes sense. So from my point of view, the lack of any large-scale 4K UHD adoption by customers is not a bad thing. This is about economics of scale. But any streaming tech should only be used when it's driven by a clear business purpose. Now, for some, 4k doesn't matter. They can charge more for it. Some consumers are willing to pay more money for it. The vast majority are not, and even though streaming services don't break out what percentage of their users are buying what tier for OTT services, in which case we would know how many users they have.

Speaker 2: 34:53

For 4K, all you have to do is talk to the CDNs that are delivering the content for them. That gives you an indication across the entire industry. So 4K I don't see anything growing there. Obviously, what we're seeing is we're seeing HD with HDR. Right, that's the biggest thing that we're seeing, just in terms of better quality, but not going to 4K.

Speaker 2: 35:17

Let me cover multi-CDN real quick. I see some saying multi-CDN is a trend. It's not a trend. Multi-cdn was something I looked up that started about nine years ago, from when I started reporting in the blog. It's not a trend, it's the norm. Mid-tier customers have been doing it for years as well, in addition to large customers. If you're a very small customer in SMB, you're spending $1,000, $2,000, $3,000 a month. No, you're not doing a multi-CDN strategy. In most cases, it makes no sense. So it's not a trend. Nothing's changed with multi-CDN. Now, the reason some people say that is because they see some changes at a particular company. So, for instance, disney went from at least six CDNs down to three. I documented this in the blog last year. Prime Video still uses more than a dozen. Paramount at the time, if I remember correctly, was using two because their service was only in the US. The moment they went internationally, they expanded to a lot more. So that's not a trend.

Speaker 2: 36:16

You're changing the number of CDNs you're using based on where you need to deliver traffic and in what regions, and nobody would debate hopefully not that in certain regions of the world, depending on the performance you want, the capacity you need, certain CDNs are better than others. That's just a fact. Akamai dominates India full stop. Nobody else has capacity even close to what Akamai has in India. I saw what Gio was requesting last year for cricket from all the CDNs Nobody could even come close to what Akamai already has in the country. Right, that's just the reality of it. That's okay. Certain CDNs are stronger in certain regions. That's why you are shifting traffic in a multi-CDN strategy based on not just price but performance and the coverage you want to have. So it's not a trend, that's just. That's the way it is.

Speaker 2: 37:17

The vast majority of everyone you can think of out. There is already in a multi-CDN strategy. Now Major League Baseball is actually only on one. You can trace that and see it's coming from Fastly. They recently got off DSS. Whether or not they stay single source CDN don't know. I would imagine over time they would go to more than one, but that's very rare. It's a very rare case.

Speaker 2: 37:40

Ultra-low latency I'm not going to spend a lot of time here because I've done a lot online. I know people love to argue about ultra-low latency. We don't see many use cases. Aws did put up an interesting use case on their website a week ago two weeks ago where it was talking about a particular mobile application for auctions in Europe. Okay, makes sense. Low latency, it's auctions, you're taking bids in real time. You better have a solution there. Not a lot of traffic, that's okay, it's a great use case. But other than that, we don't have a lot of use cases.

Speaker 2: 38:14

This is about business and everyone's running around saying, oh well, customers demand ultra low latency or low latency. No, they don't. They absolutely do not. They might like it, but you don't see customers saying, well, I'm not going to stream that piece of content because it's not at low latency. That doesn't happen. So we're talking about a time period in our industry where everybody's looking to cut costs continually still this year. Don't think that that's over. So doing Dash because you're going to save 500 milliseconds, okay, but what's the business value? It costs more to do that, to replicate content in all these different formats. See, you have to think about the business impact. Last time I checked and I won't say this is 100% still, because I haven't checked there was a reason Disney was all HLS Cost. Think about it. Cost saves a lot of money. So again, it's about the right technology based on the business model.

Speaker 2: 39:12

Cmcd interesting, but I've seen some suggest well, it's going to drive new traffic in the market. It's not driving traffic. It's not increasing revenue. Cdns give away CMC data right. Cmcd for those that don't know is simple. It helps bridge QOE and QS data for improved playback so you can correlate metrics to enhance CDN performance. That's all it is, but it's not driving new demand.

Speaker 2: 39:40

I've heard that mentioned multiple times. It's not the case. Don't believe me. Go talk to any CDN, all right. So let's go to CDN pricing, something I know a lot of people want to hear about. So in Q1, I ran my CDN pricing. Again, this was the 17th year that I've collected it Over 500 customers. I do it twice a year Q1 and Q4, almost every year those are the two quarters. Sometimes I adjust it a little bit, something like COVID hits, based on what we see for demand in the market. If you go to cdnpricingcom, that's where you'll see what I've talked about with CDM pricing the last time I did it. That's going to be old and outdated. I will get a new post up at cdmpricingcom Now.

Speaker 2: 40:21

I used to give all this away for free the data sheet, the Excel data sheet that has all the raw data. There's no customer information in it. But I stopped doing that because I noticed everybody was using it, especially in Wall Street and others, and frankly, making money from it. So I thought, okay, I like giving out free information, but this is a little too much. So I do not give it out for free. I do charge for it. It's a couple thousand dollars to get the raw Excel sheet. It's raw data. It's not a report. I'm not analyzing it for you. Cdn vendors buy it, wall Street folks buy it, wall Street folks buy it. So there's a lot of people who buy the data, sometimes once a year, sometimes twice, sometimes every three years. So contact me if you'd like more details on that. I break it out based on size of customer, how much they're spending. These are customers that are spending at least $50,000 a year. If you're spending less than that, I do a separate survey once a year just for SMB customers.

Speaker 2: 41:13

I talk about the traffic growth. Is it going up or down? By what percentage? What are you paying per gig delivered? If it's on a per gig delivered, some are per meg sustained, naturally are you bundling this with other services? Sometimes in the surveys I even ask which CDNs they're using so I can track. Well, okay, I've noticed a lot of customers over a certain period of time have expanded the number of multi-CDNs that are actually in their decision-making and where they route traffic. So the survey questions change a little bit at times, but it's all around price, traffic volume and what the trends are. So to give you an idea in terms of pricing that I'm seeing, you have some of the largest customers in the world who have blended global traffic pricing at triple zero eight and triple zero nine. It's a fair price because you're delivering traffic in other countries Japan, south Korea, africa, india where it's expensive, it's not cheap to deliver content there. You have others who maybe are delivering traffic just in the US where you're at triple zero four, triple zero five. So pricing is dependent based on so many different factors.

Speaker 2: 42:30

Now, for anyone listening, please do not go to your CDN and say, because I get this all the time, well, dan says that I should be paying triple zero four. It's not what I'm saying. I don't know what your needs and requirements are. If you want help, call me. It's free. Just call me.

Speaker 2: 42:46

I'll tell you what the going rate is for the amount of volume you have, with the commit level that you have and the region you have time of day you're delivering it. Is it live? Are you paying an RSVP fee? There's lots of variables. When people say to me what's it cost to deliver video, the way I answer that question is I ask them a question back and I say what's a new car cost? Where are you driving it? Is it just going to be in a place where it's always good weather? Do you need to handle it in the winter? Do you have kids? Do you want to top on it, top down? You have to take all these things into account, so that's not necessarily the pricing you should be paying for. Lower pricing is okay if traffic bits are growing, but they're not.

Speaker 2: 43:32

And I'll talk about that later, because on the last earnings call from Fastly they actually specifically called that out. Pricing was going down but they weren't getting large commits. So again, cdns are giving out a lot of this information. You just you have to listen to what they're saying. Lowest rate I've seen on pricing is around triple zero three. Triple zero three two. Now that doesn't mean CDNs are giving this pricing either. 00032. Now that doesn't mean CDNs are giving this pricing either. Cdns realize they can't give this stuff away. There was a rumor that somebody put out I don't even want to call them an analyst because I don't think they are, but they put out this LinkedIn thing to create turmoil that, oh, I checked with a customer and Akamai is giving away CDN delivery for free. No, that is not true. Actually, akamai went on record to say that it was completely inaccurate and that no CDN can afford to give away bid delivery, not even Akamai.

Speaker 2: 44:21

So many want to say the race to the bottom on pricing. What I can tell you is there is no race. There is no race. You have CDN vendors that are telling customers you want below this price, thank you for the opportunity to bid on this business, but we're going to walk away because they can't make money on it. And if you don't believe me, again, listen to the earnings calls of even someone as large as Akamai, the largest CDN in the space on revenue, who tells Wall Street we've got some contracts expiring tied to delivery. We've got some contracts expiring tied to delivery which we've decided not to renew. It's not strategic to our business, it's not good traffic. So we know this, this idea that everyone's racing to the bottom, to the lowest price. They're not, I can tell you they're not. And especially with Lumen and StackPath out of the market. They're not because those two CDN vendors were more willing to cut pricing. That made no sense. Now they're gone.

Speaker 2: 45:22

You have some CDNs in the market today where their cost, just on a bit delivery from the network side is triple zero three. That's their cost and that's not technically their full cost, because they're not factoring in everything to actually support that, but just from a pure network infrastructure standpoint triple zero, three. So if a customer wants triple zero three or lower, you're not making any money. Even at triple zero four, you're not making money. So most CDNs are trying to stick at a triple zero five price point and not go below that. Some are depending on the size of the customer. Yes, there are instances where some CDNs will give a lower pricing because oh, by the way, you'll then give us millions of dollars in cloud security business that we have high margins on, fine, we're willing to discount it, but they're not giving it away. That is absolutely not true.

Speaker 2: 46:15

Clients have also. Customers have been asking for a low single global rate. The shift has been in the regional quality they expect. Previously, clients really expected and requested global delivery at grade in the US and Europe and good enough everywhere else, but they weren't ready to pay a premium for excellent global delivery. Now we're seeing across the industry demand for the same level of quality across a much larger part of the different regions in the world, but they don't want to pay a premium. So it's an increased cost for vendors. You're seeing this shift from from, as one vendor said to me, good enough to excellent in these regions, and it requires significant time and investments.

Speaker 2: 47:02

So if you really want to understand this market, talk to customers in different regions of the world, talk to CDNs in different regions of the world. It's very important. One CDN here two I thought this was important they said that we've acquired a few small to mid deals that were formerly served by Akamai. In these cases we learned that Akamai refused to go any lower at the price because they saw no upsell potential to other products Exactly. So that doesn't mean Akamai lost the business or whoever the CDN is. It means it's just not strategic to their companies, not the type of traffic they want to have on their network. So also, sometimes when you hear someone say, oh, the CDN lost that business, they didn't lose it, they no longer want it. It's not profitable for them. It can be profitable for other CDNs, depending on the size and scale, especially if it's a small to mid deal.

Speaker 2: 47:57

So you have to look at the details and particulars. Let's talk about open caching. I got a lot of people emailing me in before this. Hey, dan, open caching, open caching. What's going on? I don't have any numbers. I don't either. I'll be honest. I have no idea what's going on with open caching in the sense that we've been given no numbers by anybody to actually judge what is taking place.

Speaker 2: 48:20

Verizon has been talking about open caching for years with Disney, but they've given out no details. They've given out these high level details that they continue to present year after year, saying, well, we compared a third party CDN we won't mention to open caching, verizon, and you know. Here's how much better the performance was based on these metrics. Yeah, but at what capacity and at what cost? And were you testing live and on demand? And what was the CDN you were testing against? There's no information, none. So we don't actually know that it's a better solution. And also the data that I see them still talking about. That data looks to be at least 18 months old and the CDN that they're comparing it to they don't mention. I know who it is, I won't say who it is, but their network has changed drastically in the last 18 to 24 months, so we don't have any details.

Speaker 2: 49:14

On the SVTA's website, they list 32 companies tied to open caching, but only two of them are content owners Disney and Sky. So to date, I don't see any deployment of open caching. I have no data to say otherwise and I'll get separately into quilt later. Uh, when I go through vendors, but just somebody show me data that tells me something different, I'm willing to look at it, but verizon will not answer any questions about their deployment in any detail. They just they won't answer them. Even when I put them on LinkedIn, they won't reply. They don't do product briefings right. They don't want to talk about. So we have no numbers to show it. To show anything, people multiple sources tell me that Verizon's never pushed more than two terabits of capacity with open're. They're open caching, but we've never heard them say that. So just no idea there. Um, I don't see it being deployed. Someone who has other details with actual data, not high level generic stuff. Yeah, happy to look at it. Uh, regional traffic let's talk about that for a minute.

Speaker 2: 50:27

Delivering India is at a premium. Almost no other CDNs can really get into India because just man Infocom Limited and Airtel they charge a lot for ports in their network. I mean a lot. So no one has made much of an impact on Akamai's business in that region from a capacity standpoint, because Akamai invested early and often in India and some other CDNs who actually went in there are actually shutting down in India. They're pulling out because they just don't have enough business, they don't have enough sales. The cost there is so high. So that's still a region of the world that Akamai is absolutely dominating in.

Speaker 2: 51:08

Tata just real quickly has a CDN in the country right here with very little capacity. I'll go through Tata later when I do some of the vendor stuff. Airtel is also in the region and if you combine Airtel with Reliance Geo Infocom, well let's see it's Reliance Geo Infocom Limited Technically that's the name that and Airtel. They're 85% of the broadband market in India, two ISPs really. So India is an interesting region. You know some places APAC, japan, south Korea much more expensive as well. So regional traffic still matters. North America, europe pretty much the same, but there are some countries in Europe here and there LATAM can be hit or miss depending where you're going, but there's still not a huge demand in Latin America from a viewership standpoint.

Speaker 2: 52:04

Let's do performance testing as a topic for a minute. So previously you know a lot of clients that I saw evaluated traffic metrics, usually on a quarterly basis or maybe yearly, and they allocated traffic at a regional level. But in many cases some are evaluating in, you know, near real time. They're doing traffic relocations within multi-CDN stack more dynamically. It's happening in a country or even in a single ISP level. Now we've seen that for live events, naturally, but you're seeing that now a lot with VOD. So you have more customers that are no longer willing to make a year-long or high-volume commitment because they're changing how they deliver that content and we're going to continue to see more of that. That shouldn't really surprise anybody.

Speaker 2: 52:59

Coming up in the last few things here that we're going to discuss, let me real quickly highlight Netflix, nfl and Christmas. So I did not realize until I looked up the numbers how big NFL is on Christmas. On pay TV For the 2022-2023 NFL season, just overall, espn averaged over 17 million viewers, prime almost 12, nbc almost 21, fox over 19,. Cbs over 19. The numbers that you see for Christmas of what the NFL did was absolutely incredible. So in 2023, the Raiders and Kansas City Chiefs game did more than 28 million viewers on CBS. It was the most watched Christmas Day game in 34 years. The Eagles-Giants game was 29 million. In 2022, the Broncos-Rams game was 22.5. In 2021, the Browns-Pockers was 28.5. Now that was across Fox and NFL. These are big numbers. Now, this was very interesting. I didn't know this. I'd never seen Netflix disclose this before.

Speaker 2: 54:19

In the press release with the NFL, it was mentioned that traditionally on Netflix's network, christmas is their largest day of streaming for the whole year on pay TV, which is now going to be two games on Netflix on top of it being the largest, sounds like day of viewership on Netflix's platform to begin with. This is going to be something. This is really going to be something to watch. From everything I understand, netflix is still going to deliver a hundred percent of that video themselves. Even if they get a 75% 60% of the fraction of what they get on pay TV, it's still going to be the largest single live streaming event in US history inside the US. Now keep in mind, too, that Fox, cbs, nbc, when they had these games on pay TV blackout restrictions Well, sorry, I shouldn't really say blackout restrictions it's only available regionally in North America. Well, netflix has no blackout restrictions. It's available globally.

Speaker 2: 55:35

Do we think a lot of people in other countries care about the NFL? No, but there's some for sure, and there's definitely people who are going to be traveling outside the US for the holidays. So my point is, as an industry man, we got to keep an eye on that one, because that is going to be traveling outside the US for the holidays. So my point is, as an industry man, we got to keep an eye on that one because that is going to be really interesting to see what Netflix pulls off from a number standpoint. I do think they'll give out numbers afterwards because they're working with the NFL, so I think they'll have to. I really hope they execute flawlessly because it would be bad for the industry if they didn't.

Speaker 2: 56:03

Netflix is, I would argue, the most knowledgeable, experienced company, delivering the type of content they are, the scale they are and the quality they are than anybody else out there. Sorry, youtube, but to me it rivals YouTube because we're talking about long-form content Now obviously we know live is a different animal. Long form content Now obviously we know live is a different animal. And then supporting live on older devices much harder as well. So I don't know what the stack is going to be for Netflix. I'm sure we'll hear more, but keep an eye on that one Very, very interesting.

Speaker 2: 56:38

So let me go through vendors here. I'm going to end with vendors, just some numbers and whatnot. As far as people writing in saying Dane, talk about edge compute security, I'm not going to cover that, but what I will say is we all know that CDN vendors as a general rule have migrated and did this many years ago over to services with higher margins. Right, we saw Akamai start this in the market. Wow, I can't even remember how long it was when they acquired. I believe the first one was NetWeek. They realized that at some point down the line, cdn would have lower margins and they needed to expand into more value-add services, as the industry calls it, with higher margins. Very good at that. But that's where CDN vendors are investing the majority of their time and their effort for the most part, unless they're really just dedicated to CDN, which very few are. If you look at again in Akamai, look at all the acquisitions they're making, it's tied to security. It's not tied to CDN. Yes, they bought some contracts from Stackpath and Lumen where it made sense, but they only took select contracts and they gave out numbers as far as revenue goes. We know exactly what they were worth. So that's what we're going to see as far as CDNs going forward investment.

Speaker 2: 57:55

As far as consolidation, what to expect? I don't expect consolidation. To be honest, I really don't. I think there's some we still have to keep an eye on. Right, obviously, egeo, because they're still going to do some regulatory filings around financials which we're waiting on, which I'll cover in a minute. Google's Media CDN. They've never to date, really put a lot of effort behind that, so they have to decide what they're really doing there. What's the strategy? Telecodes and carriers I don't see a lot of change there. They're still deploying CDN to reduce transit costs, but I don't expect a lot of consolidation in the market Nothing anytime soon.

Speaker 2: 58:34

So let me break out finally here.

Speaker 2: 58:35

Last couple of vendors.

Speaker 2: 58:36

I'm going to go through numbers.

Speaker 2: 58:37

Let me start with AWS.

Speaker 2: 58:39

So this is one where I don't have a number. That is fact. Okay, this is Dan Rayburn's estimate I'm not going to go into my sources that CloudFront revenue in 2023 was about a billion dollars. Now, it's important, when you're looking at these numbers and breaking them out in further detail, that you're truly comparing apples to apples, because we know that AWS for CloudFront and also CloudFlare, have a lot of customers that are developers, smbs. At one point, cloudflare actually gave out a number that they had more than a million customers using their service. Well, those million customers are not paying $50,000 each. So AWS and Cloudflare in particular, have a lot of customers that are very, very small. Many years ago I don't know how many people know this Amazon actually put out a release on the number of customers they have paying. They use the word paying for CloudFront and at that time this is years ago the number was over 40,000.

Speaker 2: 59:50

So think about that. You have a large portion of CloudFront's revenue that is tied to revenue and customers that Akamai would never go after because they're not targeting a $500 a month customer. It's not better or worse, it's a different strategy. So I put AWS's revenue to a billion dollars total for CloudFront. Now what portion of that is actually comparable to what Akamai and AGO and Fastly and some of the larger CDNs are going after? I think it's probably about half 40% to 50% of that, but that I don't know. That number I'm giving out does not include AWS media services. Yeah, I expect that to be about a little over $400 million business this year.

Speaker 2: 1:00:40

Aws still dominates the market when it comes to a video platform outside of just bit delivery. It's just what they have end to end. Nobody's matching that right now at that scale. Also, multiple CDNs are not trying to get into that business. Yes, egeo is because they have Uplink, but Macamai is not in that business. Fastly is not in that business. That's not what they're trying to do. So you have to understand the products and services these companies offer. It's very hard to just say, well, this company compares to this company. Well, they actually look very different all these CDNs, in most cases with different products and services.

Speaker 2: 1:01:18

If you want to hear more about how Prime Video uses multiple CDNs, go to nebstreamingsummitcom. Ba Winston from the Prime team was one of my keynotes this year in April. He talked in great detail about how Prime Video uses multiple CDNs, including CloudFront, how they do multi-CDN decisioning. He talked about ultra-low latency. He was great. He didn't give out numbers of revenue, of course, but some really good technical details.

Speaker 2: 1:01:43

If you want to listen to that, let me go on to Akamai. So I see people write all the time Akamai had video revenue of this, akamai's CDN revenue was that? It's wrong. Akamai calls it delivery revenue. Their delivery revenue for 2023 was 1.54 billion. I see people multiple times write things like Akamai has exited the media delivery business or Akamai is no longer in the CDN game, and I just laugh. That's all I can do. Right, because this is a public company.

Speaker 2: 1:02:19

Here's their number. They break it out for us. They're larger than anyone else, but you want to claim they exited the market. Now, I did have somebody write in recently that was asking me Dan, these numbers you're using from the public companies, do you go in and certify their numbers? And I was like what, am I an accountant at Accenture? No, I'm not authenticating any of these numbers. If they're a public company putting numbers out there, those are the numbers. They're not making those up. If they were, they would have a very large regulatory problem. But no, I'm taking the numbers from the public companies exactly as they're presenting them. I'm not, then, certifying the numbers in any way.

Speaker 2: 1:03:05

Now, in Akamai's Q1 earnings call they gave out some interesting numbers. Here, other numbers they expect to see $3 to $4 million in revenue from the Olympics. And notice they didn't say Peacock, that's Olympics, that's for all the business tied to the Olympics across all the different broadcasters that you're doing work for. So is Olympics big from a branding standpoint? Yes. Is it large in terms of revenue? No, don't believe me. Here's the number from Akamai and kudos to Akamai, because they've now been doing this for multiple years in a row where they give out the revenue number.

Speaker 2: 1:03:38

They also said due to social media customer who has optimized their platform to save money. Again, they didn't say TikTok, they couldn't, but it's TikTok, akamai said it would take a hit of $40, $60 million in revenue for the year due to less traffic. They specifically, as I mentioned before, talked about how this customer is prefetching less content, optimizing their entire infrastructure stack for the past few quarters. Now Akamai also said outside of TikTok, the large customer, that due to quote slowing traffic growth across the industry something they highlighted gaming in particular Everyone of course thought video. Well, they actually highlighted gaming Slowing traffic across the industry, but the use case is gaming they expect to see 20 million to 30 million less delivery revenue for the year.

Speaker 2: 1:04:28

Now they also said by the end of the quarter this was Q2, they were talking Q1, but reporting Q2, that they will reprice five of their seven largest delivery customers and expect the remaining two to reprice in Q3. Now, regarding the pricing, akamai said the pricing they were seeing was in line with what they expected, so suggesting that they're just dropping pricing to win business. Not the case. Also, akamai confirmed that there is delivery traffic that Akamai is not taking on as it's not quote profitable or strategic. Traffic to Akamai is not taking on as it's not quote profitable or strategic. They did also mention that their delivery business is highly profitable, that they've reduced their delivery capex cost to low single digits, which is a 50% reduction from a few years ago. So for those that wonder, is Akamai's CDM business profitable? Well, they're saying their delivery business is profitable.

Speaker 2: 1:05:29

For those wondering overall traffic growth, why is it slowing? Why is the rate going down? Akamai even called out some examples of some OTT platforms that lost subs. So subscriber growth across the industry isn't always strong every quarter and every year. Let's go to Egeo. So the last we heard from Egeo was a little bit of news in March saying that they were guiding in range for their Q4 guidance on revenue and EBITDA. So their accountant Ernest and Young resigned with a week to go in Q4. So new accountant BDO was selected and they're working to audit 2023 and check everything. They have to go back and check everything Ernest and Young did to 2022 as well. So we're talking about thousands of hours of work here.

Speaker 2: 1:06:21

So I don't have an ETA when AGO is going to refile the numbers. I don't expect the numbers to change in any way. They've already talked about how they've taken north of $60 million in costs out of the business just in the last nine months, and that includes network optimization, some layoffs, offshoring, vendor consolidations. They're still cleaning up that business. They've done a good job with it. A lot has changed. They also are very clear that they're holding to a floor on CDN and restructuring comp planes around those floors. They're not giving this stuff away to win market share. You can't become profitable that way. They've also said you know, come out and publicly said this. They're active, increasing pricing or removing customers who are not profitable. Hmm, it's the exact same thing Akamai's talking about. These are the trends that we're seeing from the CDNs.

Speaker 2: 1:07:19

You know their focus also Uplink security. So that's you know Uplink is a large revenue business for them. You can see the numbers if you look at when they acquired that business. So that's important. Also, the app side is important to them as well. And then, finally, they took out 68 million in debt financing in October and they extended their term debt by three years. So they have runway. You see, some people sometimes, oh sorry, egeo will just go under. They're not going under, they're continuing to restructure their business, but they've already restructured in terms of their debt. They took 68 million in financing. So they're one that we have to watch and we have to look at the numbers when they come out. We're still waiting on more information.

Speaker 2: 1:08:10

Fastly did not have a great earnings call. Unfortunately, there was a lot of uncertainty in Fastly's earnings call and here's what you need to know. So this is important because, for the first time ever, fastly broke out revenue into three product lines Network services, which they define Thank you Fastly, you define something Awesome. More vendors need to do this. So network services they define it as solutions designed to improve performance of websites, apps, apis and digital media. Cool. Next bucket security Products designed to protect website apps, apis and users. Good. Third one is other Defined as emerging products offering. Sorry, emerging products offering, which includes compute and observability.

Speaker 2: 1:09:05

So in Q1, network services compromised comprised sorry 79% of revenue. Security was 18. Other was 2.7%. What we found out on the call after was compute is about 1% of revenue. That's it. Compute. For all the talk, compute is 1%. Now many have asked me why would Fastly put these numbers out in these buckets when it's not great that the compute revenue is so small and they're talking about being this compute platform? I don't know, I have no idea, but 79% of revenue is tied to network services.

Speaker 2: 1:09:39

Now they discussed challenges in the CDM business and they said they saw a slight uptick from the typical level of re-rates, which are largest customers. But we have not seen the commensary traffic expansion usually associated with this motion. So in other words, pressure on CDM pricing is not great, with the largest customers demanding lower pricing. But even with that lower pricing, they're not guaranteeing larger commits. Now, not surprising. I've been seeing this from customers for the last 12 months. What was surprising was that Fastly is saying they were only just seeing that recently. So that certainly wasn't good, because they did come out and say that when it came to the largest accounts quote they saw more pricing pressure than they are used to. So I don't know what they were used to, but the larger pricing pressures, you know that's definitely not good. You know that's definitely not good. So as a result here, if we look at some of the numbers again, when I was talking about largest customers in the space, fastly came out and said their top 10 customers represented 38% of total revenue in Q1. So what that means is any further impact on pricing for the largest 10 customers is going to have a negative impact on revenue growth. There's just no way around that. But again, it's another data point that shows you what is really taking place across the CDNs. It's clear we don't have to guess on a lot of this.

Speaker 2: 1:11:28

Additional things from the earnings call they talked about how quote volatile the market was. What they're seeing in the market regarding traffic customers adding multiple CDNs in their traffic data delivery mix. In other words, more CDN customers being sorry, more customers adding additional CDNs in some cases. Again, all things that we've seen in the market. This seemed to really confuse the company. I'm not quite sure they did say that because of what they're seeing with these customers or they're now quote changing their engagement model, uh, which is good, but you know, you gotta, you gotta change your engagement model before this happens, not after. I think part of the reason, too, is Fastly hasn't had a chief revenue officer for a while. They did recently hire one since the earnings call, which is good, but they definitely have work to do.

Speaker 2: 1:12:26

Cloudflare 2023 revenue was just about $1.3 billion. Now, cloudflare is one of those companies where it's really hard to determine what percentage of their overall revenue is tied to delivery, because they do so much in terms of DNS and DDoS and I don't put that in the delivery bucket. So I'm guessing here. Guessing that about 500 million of their 1.3 billion would be tied to my definition of CDN. But then under that 500 million, the vast majority of that is SMB. Small Cloudflare is not doing any live events, right, that scale that you see. Right, they're not one of the CDNs in the mix for all the big events you see out there. So much of it is small object delivery for SMB customers. But you have to put them in there. So, the 1.3 billion I put the number at about 500 million. That goes under CDN.

Speaker 2: 1:13:22

Cdn 77, this is a private company, but I forget when, but many months ago they allowed me to put out revenue publicly for the first time. 2023 revenue was in the 140 to 150 million range. Two thirds of that came from content delivery. The vast majority of it, they did say, came from video specifically. So that's good. 2023 revenue was up 30% year over year. They gave me this number before the year ended, so they thought it could be closer to 40% year over year. Companies currently cashflow positive. They get the majority of their delivery revenue from video content outside the US. They get the majority of their delivery revenue from video content outside the US. They've upgraded their network from 105 terabits per second to 150 in terms of egress capacity. They're quick and they're nimble, so very focused on what they're doing. Hence they can really focus on the margins. They're not as price sensitive to US-based delivery pricing and super large customers who are spending tens of millions of dollars a year who want the lowest price. There's a profit to be made in the mid-tier market for vendors that specialize, and that's exactly what they're doing, so I've been impressed with them.

Speaker 2: 1:14:39

Let me get to Quilt. So I got a lot of questions. I actually I get questions on Quilt all the time and I wish I could answer them, but the problem is I don't have data, I don't have information. You know Quilt. They talk to me all the time, which I appreciate, but they just as a company, they don't put out data, and we've seen them announce deals with Airtel, bt, stc, telefonica, vodafone, cisco led a $70 million investment in 2021.

Speaker 2: 1:15:11

But what we don't know about Quilt is, even though they have deals in terms of deploying their caches around, what is their capacity? What are they pushing in any given day? What is their peak? What is their revenue? How has it trended over the last few years? They talk about rev sharing with some service providers, but we've never heard what rev. Sharing is or does it work? We've yet to hear from content owners or network operators sorry, network operators that they're making money from this. So unfortunately I really don't have any insight into what quilt is working on, other than you do see them announce these deals with these, with these large telcos and carriers and there's a couple more I'm not going to mention that they're closing deals with right now. But then what we don't know is inside a telephonic or, like I mentioned before, verizon, what does the deployment look like? What's the size of it? How does it compare from a performance standpoint? Carriers would say, well, it performs better, but okay, but at what scale? For what type of content, live on demand? We just really don't know. So wish I had more information. But the company's made a decision at this point where they don't give out more information on what they're deploying. Hopefully that changes down the line. I'd like to highlight it more, but I just don't have the data.

Speaker 2: 1:16:31

Google's media CDN I'm not going to go into a lot of details here. It's very small, it's a bare bones CDN. They haven't added a lot of functionality, they don't break out revenue, but it's small. It was interesting to see a few months ago, a patent lawsuit was filed against Google regarding caching and content delivery methods and it dates back to Sandpiper Networks. For any of you OGs listening Sandpiper Networks, you can find more about that on my blog. Let's go to CTS, comcast Technology Solutions. They're very small, I'm not going to go into numbers again, but very small. They have a few large customers. Peacock is one of them. Small footprint right, it's Comcast. They're not all over the place, so small, but having Peacock in-house as one of your customers certainly helps them.

Speaker 2: 1:17:28

Tata Communications this is a strange one. I did actually hear back from Tata today, finally, not in enough time to talk to him before this podcast. But what's odd with Tata is everyone I know there is gone, uh. What's what's odd with todd is everyone I know there's gone, uh. Links in their blog, on their, on their website or blog post from 2015, data sheets are from november of 2020. Broken links everywhere. Um, you try and sign up for their newsletter, you get timeout errors. Their website loads so slowly, which is crazy because they're saying they, you know, have the fastest network out there. I get links to CDN stuff that take me to a page about mobile CDNs in China. Their site's an absolute mess. So as far as I'm concerned, I wrote Tata off.

Speaker 2: 1:18:13

But then a few months ago they announced a partnership with I'm probably going to say this wrong CIRON C-I-R-I-O-N in LATAM, where they're going to deploy Tata Communications CDN services. So I don't understand what's going on there. I'll do a blog post update once I have a briefing with them. What I hear is they have very little capacity in India and yet that's their backyard. So it doesn't sound like they're really in the CDM market anymore. But I'll get some more details on that.

Speaker 2: 1:18:48

So I got two left here, a new company called Netskirt, canadian based. They've raised about 7.2 million series A round that came out just a few months ago and they're focusing on delivering video to remote and rural communities, specific video applications like rail operators and airlines. So their goal is to help create a CDN footprint to help smaller ISPs manage live event traffic. Underserved urban areas are many times too far away from the video source or CDN nodes for the customers to really realize the benefits of the CDN. So that's the problem they want to solve. They have a total of 3.5 terabits deployed today across 50 ISPs in 31 states. They plan to add another 15 terabits to its footprint. They do want to expand to Brazil, uk and Italy. So here's an example of not trying to go too big, too fast, trying to solve a very specific problem in a very specific location. Could they charge more per bit delivery? Absolutely right, it's specialized. However, are they going to get the level of traffic that major CDNs are getting? Well, no, because that's not what they're looking to do. So, again, a unique use case of a company that's very focused.

Speaker 2: 1:20:07

I'm also going to highlight here Orange, real quickly, only because I got a couple of questions about them. If you haven't seen, I did a blog post a couple of months ago. So they're offering a commercial CDN offering in France and they're selling content delivery services for live and on demand plus software downloads, dynamic content and very few, uh, orange carrier, telco, isp, network operator, whatever buck you want to put them in Very few are doing that, uh, actually selling it to third parties, uh. So, uh, they went through everything they were doing with me, which was great. Their underlying CDN software is from G-Core 12 ops in France. Now they have a total capacity of five terabits.

Speaker 2: 1:20:47

So when they announced this, I saw some goofball analysts in the space being, like you know, akamai feels threatened by orange CDN. It's like guys, just stop Right, it's public information, they got five terabits Right and it's in one place, france. So this is the reality of it. Nothing wrong with that. So also they call their service a premium service. Their value proposition is it's a guaranteed level of capacity in a network they own and operate. So they're focusing on customers who are willing to pay more and want a higher level of QoS in the country. So they're not going to compete on price and they're not going to undercut other CDNs to win business. That's directly from them. Smart Now, phase two they want to roll out into Africa, middle East.

Speaker 2: 1:21:35

They've suggested they could grow total capacity in France to 10 terabits by the end of this year. You know possible. So the way to think about them, it's just another potential CDN option for customers using a multi-CDN approach in a particular region of the world that want a certain level of QoS and are willing to pay a premium price for that. Some customers will need that. But there's also a limit in terms of the size of the customer you can go after with and that's when that's the capacity you have. But it's another option in the market. So that's what I got today.

Speaker 2: 1:22:12

I'm sorry this took wow, this took an hour and 20 minutes. I was trying to condense everything down but as you can see, there's a lot over here. I did get through a lot of information. I will provide a transcript of this blog post so you can search throughout it. It's too much to put sorry. Of the podcast, it's too much to put in a blog post. I will do an additional blog post that gives out a chart on the size of the CDM market, broken out by vendors, like a pie chart. It's something everybody can use and I will also do something additionally on CDN pricing.

Speaker 2: 1:22:45

I am always updating cdnlistcom so as new vendors enter or exit the market, you can see that. If you really want to understand where the CDN market's gone over 25 years at cdnlistcom, at the bottom of that blog post I keep a running list of every CDN that's ever been in the market ever. What happened to them when they left the market? Did they go under? Did they get acquired by whom? Last I looked there was about 80 vendors on there. I'm always updating that.

Speaker 2: 1:23:14

But if you have any questions, reach out to me, email me, call me on my cell phone, 917-523-4562. I take calls all day long, free of charge. Largest customers, smallest customers, whomever yes, again, I am charging for the CDM pricing raw data. If you're interested in that, reach out to me, but I'm here as a resource, so if there's something you're looking for, reach out to me. For all those that provided information to me before, I sincerely thank you. I'm sitting on a lot of data I couldn't even get to on this podcast because there's so much. I appreciate the trust that others have in me in terms of talking to me about what they're doing, knowing that I won't put it out there if you don't want it out there, but the background information is extremely helpful. So that's what I have today. If you have any questions, reach out. Appreciate everyone listening.

Speaker 2: 1:24:06

I'll add up some more blog posts on this topic over the next couple months. I'll be watching what Netflix does some additional new services launching in the market. I'll be looking at those as well. I've got some data on latency. You'll see that capacity You'll be seeing. I'll be putting out some information on a recent survey that there as well. So check out the blog. If you want to get to all of the posts on my blog just about content delivery, go to contentdeliveryblogcom. That'll take you right to the post on my blog just about CDN. That doesn't include anything else in the industry. I've tried to make it real easy. Thanks very much for listening. Appreciate everyone's support. You have any questions? Reach out to me. Talk to you next week.

Dan Rayburn: 1:24:50

If you enjoyed the show, send it to a friend, have questions for Dan or Mark, connect with them on LinkedIn at any time, and be sure to check out Dan's blog at streamingmediablogcom.