In this comprehensive podcast update on the CDN market, I separate facts from opinions, providing a detailed analysis of the latest trends. I cover everything from CDN pricing trends, market sizing (𝐚𝐛𝐨𝐮𝐭 $𝟓.𝟑𝐁 𝐢𝐧 𝟐𝟎𝟐𝟑), traffic and revenue growth rates, DIY initiatives (TikTok, Disney, Apple, Netflix), tech trends and demands across 4K, ULL, CMCD, bitrate optimization, multi-CDN, Open Caching deployments, changes in performance testing, to unique delivery challenges in regions like India.

Revenue numbers for CDN vendors, including Akamai Technologies, AWS, Cloudflare, CDN77, Edgio and Fastly, are broken out, along with where the numbers come from. Other CDNs discussed include Comcast Technology Solutions, Google Cloud Media CDN, Netskrt Systems Inc., Orange, Qwilt, Tata Communications and others.

I discuss the potential impact of market consolidation, Netflix streaming NFL games on Christmas, and the trend that will continue with customers looking to save on delivery costs. I also highlight bad data and red flags within reports in the market so you know what to avoid and data to stay away from. Listen here: 🎙 https://lnkd.in/epzXSMVA

I mentioned many numbers in the podcast, nearly all from public sources, and I believe I got them right. But if I misspoke, I welcome any corrections. If you have any questions, please put them in the comments section or contact me anytime. (dan@danrayburn.com)

My thanks to the nearly twenty CDN vendors, customers, sports leagues, OTT platforms, carriers and DIY content owners I spoke to in the week before the podcast.

The podcast transcript provided by my podcasting platform has many mistakes, so I will work on editing it and posting a transcript when I can.

Nothing even comes close to the viewership of NFL games. Below is a list of recent sports viewership numbers from the NHL, MLB, WNBA, NBA, Formula 1, U.S. Open, Indy 500, UFL and Premier League.

For comparison, for the 2022-23 NFL season, Monday Night Football on ESPN averaged 𝟏𝟕.𝟏 𝐦𝐢𝐥𝐥𝐢𝐨𝐧 𝐯𝐢𝐞𝐰𝐞𝐫𝐬; Thursday Night Football on Prime Video averaged 𝟏𝟏.𝟖 𝐦𝐢𝐥𝐥𝐢𝐨𝐧 𝐯𝐢𝐞𝐰𝐞𝐫𝐬; Sunday Night Football on NBC averaged 𝟐𝟏.𝟒 𝐦𝐢𝐥𝐥𝐢𝐨𝐧 𝐯𝐢𝐞𝐰𝐞𝐫𝐬; FOX averaged 𝟏𝟗 𝐦𝐢𝐥𝐥𝐢𝐨𝐧 𝐯𝐢𝐞𝐰𝐞𝐫𝐬; CBS averaged 𝟏𝟗.𝟑 𝐦𝐢𝐥𝐥𝐢𝐨𝐧 𝐯𝐢𝐞𝐰𝐞𝐫𝐬. Even the 2022 FIFA World Cup didn’t come close to the NFL.

🏀 ABC’s Game 5 coverage of the NBA Championship final averaged 12.2 million viewers and peaked with 13.2 million viewers. The average audience for 𝐍𝐁𝐀 𝐅𝐢𝐧𝐚𝐥𝐬 Game 3 on ABC and ESPN averaged 𝟏𝟏.𝟒 𝐦𝐢𝐥𝐥𝐢𝐨𝐧 𝐯𝐢𝐞𝐰𝐞𝐫𝐬, peaking at 13.9 million viewers.

🏎️ The 108th 𝐈𝐧𝐝𝐢𝐚𝐧𝐚𝐩𝐨𝐥𝐢𝐬 𝟓𝟎𝟎 averaged a Total Audience Delivery (TAD) of 𝟓.𝟑 𝐦𝐢𝐥𝐥𝐢𝐨𝐧 𝐯𝐢𝐞𝐰𝐞𝐫𝐬 across NBC, Peacock, and NBC Sports digital platforms and peaked at 6.46 million viewers.

🏒 The 2024 𝐍𝐇𝐋 𝐄𝐚𝐬𝐭𝐞𝐫𝐧 𝐂𝐨𝐧𝐟𝐞𝐫𝐞𝐧𝐜𝐞 𝐅𝐢𝐧𝐚𝐥 (Game 6) on ESPN averaged 𝟑 𝐦𝐢𝐥𝐥𝐢𝐨𝐧 𝐯𝐢𝐞𝐰𝐞𝐫𝐬 peaking at 3.7 million. The series averaged 2.3 million viewers; through 41 Stanley Cup Playoff games, viewership averaged 1.4 million.

⛳ The second round broadcast of the 𝐔.𝐒. 𝐎𝐩𝐞𝐧 on NBC and Peacock produced a Total Audience Delivery (TAD) of 𝟐.𝟑𝟖 𝐦𝐢𝐥𝐥𝐢𝐨𝐧 𝐯𝐢𝐞𝐰𝐞𝐫𝐬 and peaked with nearly 3 million viewers.

⚽ The UEFA European Championship Group Stage match on June 16 on FOX peaked at 2.2 million viewers.

⚾ ESPN’s Sunday Night 𝐁𝐚𝐬𝐞𝐛𝐚𝐥𝐥 coverage of the Boston Red Sox and New York Yankees on June 16 averaged 𝟐.𝟎𝟕 𝐯𝐢𝐞𝐰𝐞𝐫𝐬.

🏎️𝐅𝐨𝐫𝐦𝐮𝐥𝐚 𝟏 𝐂𝐚𝐧𝐚𝐝𝐢𝐚𝐧 𝐆𝐫𝐚𝐧𝐝 𝐏𝐫𝐢𝐱 on ABC had the largest live television audience on record, with an average of 𝟏.𝟖 𝐦𝐢𝐥𝐥𝐢𝐨𝐧 𝐯𝐢𝐞𝐰𝐞𝐫𝐬 and a peak of 1.97 million viewers. The Miami Grand Prix in early May set the all-time F1 record for a live telecast with 3.1 million average viewers.

🏀The 𝐖𝐍𝐁𝐀 game of the Indiana Fever versus the Chicago Sky averaged 𝟏.𝟓𝟑 𝐦𝐢𝐥𝐥𝐢𝐨𝐧 𝐯𝐢𝐞𝐰𝐞𝐫𝐬 on ESPN, with viewership peaking at 2.19 million.

🥎 The NCAA Division I 𝐒𝐨𝐟𝐭𝐛𝐚𝐥𝐥 𝐖𝐨𝐦𝐞𝐧’𝐬 𝐂𝐨𝐥𝐥𝐞𝐠𝐞 𝐖𝐨𝐫𝐥𝐝 𝐒𝐞𝐫𝐢𝐞𝐬 Finals on ESPN drew 𝟏.𝟗 𝐦𝐢𝐥𝐥𝐢𝐨𝐧 𝐯𝐢𝐞𝐰𝐞𝐫𝐬 and peaked at 2.5 million, making it the most-watched Game 1 on record.

⚽ The UEFA European Championship Group Stage match on June 16 on FOX peaked at 2.2 million viewers

🏀 Over Memorial Day Weekend, ION, CBS and NBA TV set viewership records for 𝐖𝐍𝐁𝐀 broadcasts, with 𝟕𝟐𝟒,𝟎𝟎𝟎 𝐯𝐢𝐞𝐰𝐞𝐫𝐬, peaking at 981,000.

🎾 The final round of the U.S. Women’s Open on NBC had 943,000 viewers.

🏀 CBS carried a 𝐖𝐍𝐁𝐀 game between the NY Liberty and the Minnesota Lynx, which drew 𝟕𝟎𝟒,𝟎𝟎𝟎 𝐯𝐢𝐞𝐰𝐞𝐫𝐬, the highest viewership for a WNBA game on CBS.

🏈 The 𝐔𝐅𝐋 game between the D.C. Defenders and the San Antonio Brahmas on FOX averaged 𝟔𝟖𝟒,𝟎𝟎𝟎 𝐯𝐢𝐞𝐰𝐞𝐫𝐬, peaking at 1.3 million viewers.

⚽ For the 𝐏𝐫𝐞𝐦𝐢𝐞𝐫 𝐋𝐞𝐚𝐠𝐮𝐞 2023-24 season, Peacock averaged a Total Audience Delivery of 𝟓𝟒𝟔,𝟎𝟎𝟎 𝐯𝐢𝐞𝐰𝐞𝐫𝐬 per TV match. The Championship averaged a Total Audience Delivery of 2.9 million viewers across all TV and streaming platforms in English and Spanish.

StackPath announced it is shutting down the company and will liquidate its assets. Employees posted comments on LinkedIn last week, and now the news is official. Company investors had been trying to offload the company for at least 18 months, with multiple vendors telling me they looked at the business but weren’t interested. Founded by the former chairman and CEO of SoftLayer Technologies, which was acquired by IBM in 2016, StackPath raised $180 million and acquired MaxCDN (SMB CDN), Fireblade (WAF), Cloak (VPN) and Staminus (DDoS) to launch the company. Read my launch blog post here.

In 2017, the company acquired CDN Highwinds, which had already acquired BandCon and in 2017, StackPath merged with Server Density to enter the monitoring business. In 2020, StackPath raised $216 million in a Series B round from Juniper and Cox to get into the edge computing business, suggesting that IoT, 5G and OTT video would be the driving demand. Last year, StackPath exited the CDN business, with Akamai acquiring “approximately” 100 enterprise CDN contracts from the company.

StackPath’s downfall was a constant change in management, no core product focus, no clear go-to-market strategy, no understanding of customer needs and calling themselves the “world’s first platform” for providing compute and services at the edge, which they weren’t. At one time, StackPath’s technology was the enhanced security option for Eero’s mesh Wi-Fi system, which StackPath was banking on to grow its revenue. However, that option disappeared when Amazon bought Eero in 2019.

There was also a lot of in-fighting between management and investors and far too many egos at the top. Between 2017 and 2019, I hosted multiple meetings with some of StackPath’s management team and was not impressed. The company was fast and loose with its numbers, overinflated revenue, customer deployments and other aspects of the business. By the time new management took over in 2020, I believed that the damage had already been done and that StackPath wasn’t saveable. For clarity, I last talked to StackPath’s management team in 2021, so my comments don’t pertain to all the new employees running the company with whom I have not interacted.

StackPath sent out its last invoices to customers on June 12th and is no longer offering technical support, which is a crummy way to close down a business when some customers are going to have technical questions.

Private Equity firm Clearhaven Partners has acquired IP video transport vendor Zixi and will invest in the company, along with Zixi’s executive leadership team, which also invested in the transaction. Terms of the deal were not disclosed. Clearhaven Partners continues to add to their portfolio in the video space, having made a significant investment in Wowza Media Systems in 2021 and investing more than $100 million in SundaySky in 2022.

I got to spend some time with Clearhaven Partners at the NAB Show and like their core focus on the software business, emphasizing data, video and content. At the NAB Streaming Summit in April, Clearhaven’s Founder & Managing Partner, Michelle Noon, spoke about vendor valuations and the M&A outlook for streamers and streaming tech companies. You can watch that discussion here.

Words matter. What is TV? Our industry continues to use words interchangeably as if they all have the same meaning when they don’t. TV is a device; it’s not a type of content. Cable, satellite, and streaming are different means of distribution and video services. Netflix is not in the “streaming” business; they are in the content business and use streaming to distribute their content. “Streaming” is not TV; it’s a technology.

Today, many view the definition of TV as the type of content or viewing, and when used in that context, the word will not have a single definition. When a 30-second video clip on YouTube is being compared to a 90-minute movie on Netflix, as Nielsen does, under their definition of “TV viewing,” the term no longer has a meaning. What is classified as watching TV varies amongst users. Many younger users think of their phone as their TV, and they would be right. For many, the TV is no longer considered a device but the act of viewing a particular type of content or specific user experience, no matter the screen size it is being watched on.

For some, that is content on YouTube; for others, like myself, I only consider it a TV viewing experience if the content is long-form and professionally produced. But it’s not a right or wrong answer; it’s simply a matter of viewer preference. What’s being compared is the act of viewing video content, and what one person defines as content suitable for “TV” viewing will vary. There will never be an agreed-upon definition, and that’s ok. What’s not ok is when companies like Nielsen put out video viewership data and use many of these terms without defining them. They intentionally confuse the market, which is part of their business strategy. The more complicated they make it, the more content owners and advertisers need to use Nielsen to attempt to figure it all out.

Nielsen’s methodology for the Guage report comes from what they say are “TV households” and “linear TV sources,” the latter defined as “broadcast and cable.” A household is a location, not a type of viewing. Cable is a method of distribution. Broadcast is a type of distribution, and today, one can rightfully argue that broadcast would be a channel like ABC and a live stream on YouTube from Red Bull TV — two different types of distribution and business models.

When Nielsen measures viewership on Netflix, we know users watch long-form, professionally produced content. Thanks to Netflix’s data, we know what that content is and what’s most popular. Yet, when Nielsen measures viewership on YouTube, we have no idea what type of content is being watched. A short user-generated instructional video on YouTube is being compared to a professionally produced long-form piece of content on Netflix. That’s not comparable unless Nielsen was simply looking at the total hours spent with video services over other means of content consumption. Content owners of every type compete for our time and eyeballs, but that’s not what Nielsen is comparing.

Within the other category, Nielsen doesn’t count “high-bandwidth video streaming,” which has no definition. The bitrate of a video has never been one of the factors Niselen uses in measuring, and they don’t even collect bitrate data across services. Yet they use the bitrate of a video to define why they don’t include vMVPD services in the total viewing time, which makes no sense.

Nielsen doesn’t define viewing TV viewing by the average length of a title or type of content, which would be far more valuable. We have no idea what kind of content is being viewed on YouTube or even what percentage of it is being monetized. Netflix isn’t free. All content on YouTube is. What would be very helpful to know and track is what percentage of content being viewed on YouTube is also being delivered with ads and how that trends over time.

As an industry, no one wants to agree upon even a core set of definitions and methodology when comparing all forms of video. This is a problem and where we need to start. Content type, length, genre, distribution type and business models are all variables of any video service. But we should start with the basics regarding the length of content, type of content (UGC versus professionally produced) and average viewing time. Simply calling everything TV isn’t helpful. Words matter.

For the 2024 NFL season, games will be broadcast and streamed across ten different networks and platforms. That’s great for the NFL’s revenue, but the fragmented distribution is bad for fans. The NFL can get away with this since, in 2023, NFL programming accounted for 93 of the 100 highest-rated shows on TV. Adding two games on Netflix this year allows the NFL to grow its audience, with the Netflix deal being the first that allows for global distribution with no blackout restrictions. Here’s a breakdown of how many games are on each platform:

• Amazon Prime Video: 16 games (all streaming exclusive)
• Peacock: 1 game (streaming exclusive)
• ESPN+: 1 game (streaming exclusive)
• Netflix: 2 games (streaming exclusive)
• NFL Network: 4 games (all games also on NFL+)
• NBC: 17 games (all games are also on Peacock)
• FOX: At least 100 games (week 18 schedule still TBD, all games streaming via FOX Sports)
• CBS: At least 100 games (week 18 schedule still TBD, all games streaming via Paramount+)
• ESPN/ABC: 21 games (all games streaming via ESPN+ or ABC)
• * One could suggest that YouTube TV could also be added to the list since that deal is direct with the NFL, whereas games on Fubo and Hulu+ Live TV are deals that vMVPDs do with the broadcast networks.
• * On the B2B side, Everpass has a deal with the NFL for viewing in bars and restaurants
• * Also on the B2B side, Reach TV has a deal with the NFL for viewing in airports

In a CNBC interview, Brian Rolapp, NFL’s chief media and business officer, was asked how he addresses concerns that the market is too fragmented and that it’s too hard to find games. As expected, he said he “doesn’t think it’s fragmenting too much” for the NFL since “every game we have is on broadcast television in the local markets.” While this is technically accurate, it requires an antenna to get an Over-The-Air (OTA) signal, and you can’t pause or rewind an OTA signal without paying for a DVR. Suggesting there is no fragmentation due to OTA availability is a poor argument. Fans have to go out and buy additional equipment to get the broadcast, and watching an NFL game via OTA is not the same experience as viewing the game via a cable or satellite network.

Unless viewership declines or stays flat over the next year or two, the NFL can continue to put revenue over the fan experience. As frustrated as fans are with today’s sports viewing experience, the NFL has no incentive to change its fragmented distribution strategy as long as fans continue to watch games, no matter how complex the NFL makes it.

A new patent lawsuit has been filed against Google regarding caching and content delivery methods dating back to Sandpiper Networks. The lawsuit relates to how Google delivers content across its network for YouTube content and third-party content across Google’s Cloud CDN platform.

The patents in question were owned by Level 3 and were sold off by Lumen in March of this year to Sandpiper CDN, LLC. I don’t know who is behind the newly formed entity, which appears to have been incorporated solely to acquire the patents. Andrew Swart and David Farber are mentioned in the suit regarding the history of the patents, but I don’t have confirmation of their involvement in Sandpiper CDN, LLC. and the suit. I contacted them asking for clarification and will update this post if I get a response.

For those who don’t know the name Sandpiper Networks, the company is considered by most to have built the first public content delivery network. [See my vendor list at cdnlist.com]

They first offered their “Footprint” CDN services in September 1998, and their ‘598 patent was issued in February 2001. Sandpiper was acquired in a stock deal by Digital Island in October 1999, valuing the company at the time of the news at $630 million. After the news, Digital Island’s shares soared, and by the close of trade, Sandpiper was worth a billion dollars on paper. I don’t recall the final deal price and terms.

In the suit, Sandpiper CDN, LLC. claims that in 2023, Level 3 “decided to exit the CDN market and began selling off its CDN assets” due to the “rampant infringement of its patents, which depressed its revenue and profit.” That assertion is not accurate. Level 3 exited the market as some of its largest customers, including Apple and Disney, moved to other solutions, with Apple building its own CDN and Disney consolidating its traffic to just two third-party CDNs.

Lumen’s CDN revenue did not decline due to a lack of companies licensing their patents. Level 3/Lumen’s CDN business declined before 2023 due to less overall traffic growth in the market, lower pricing across the industry, and customers optimizing their video to deliver fewer bits. We’ve seen these trends across other CDNs, which are well documented with public data.

The patents in question are 478,903; 8,595,778; 8,645,517; 8,719,886; 9,021,112; and 10,924,573.

Disclaimer: Over the past twenty years, I have been paid to work on many patent suits tied to CDN patents, including those owned by Progressive Networks, Burst, Move Networks, Microsoft, Viatech and others. As of the publication of this post, I am not working on the Sandpiper CDN, LLC. case.

Note: The Sandpiper logo shown is the company logo from Sandpiper Networks and not of Sandpiper CDN LLC.