The Dan Rayburn Podcast

Episode 61: How Users Are Consuming Long-Form Content on TikTok; OTT Platforms Aggressively Cutting Workflow Costs; Navigating AI Hype in the Video Industry

Dan Rayburn Episode 61

What do TikTok's video lengths and user engagement tell us about the future of content consumption? Join us this week as we uncover the surprising data behind this social media giant and its innovative audience. We'll also dive into the growing trend of long-form content including movies and TV shows being chopped up into bite-sized pieces for a more interactive and engaging viewing experience.

As the demand to get to profitability with DTC services grows, content owners are slashing millions in infrastructure costs through optimization of their video workflows, especially around encoding, storage and delivery. We discuss how advanced codecs and content adaptive technologies are enabling reduced bit rates without sacrificing quality. You will also learn about Netflix's multi-year technology and workflow transformation, and the increasing importance of purpose-built silicon in the race for 4K streaming.

Lastly, we tackle the issues of media inaccuracies and AI hype within the industry. From the sensationalized reporting surrounding Disney's CFO's leave of absence to Microsoft's questionable AI claims, we emphasize the importance of accurate information and transparency. Discover the potential impact of AI on the streaming industry and Microsoft's relationship with Open AI – all in this eye-opening episode.

Articles mentioned in the podcast:

TikTok's becoming a TV platform. One pirated clip at a time
https://www-cbc-ca.cdn.ampproject.org/c/s/www.cbc.ca/amp/1.6874102

With Streaming Services Cutting Bitrates to Save Money, Vendors That Bet on 4K for Revenue Growth Have Lost
https://www.streamingmediablog.com/2023/06/cutting-bitrates-4k.html

MSVP: Meta’s first ASIC for video transcoding
https://ai.facebook.com/blog/meta-scalable-video-processor-MSVP/

Questions or feedback? Contact: dan@danrayburn.com

Podcast produced by Security Halt Media

Speaker 1:

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:

Welcome to the Dan Rayburn podcast. I'm Dan Rayburn, along with co-host Mark Donaghan, back for another podcast this week where, mark fortunately we didn't really have a lot of news come out.

Speaker 3:

Yeah, that's nice. Now we can get to these topics that you and I are always saying we're going to talk about that at some point.

Speaker 2:

I have so many of those that we've never gotten to. I know the list is so long, but at least we have some interesting data. This week we're going to talk about some TikTok some I wouldn't say new numbers, but there's been some numbers about TikTok here in terms of what's working and what hasn't been working, but also how some of the younger generation is using the platform. We're going to talk a little bit about encoding, bit rates and blog post I put up this week that seemed to stir up a lot of people, which I thought was kind of odd.

Speaker 2:

Yeah, to a degree, i mean I don't get it, because all I did was put out facts, but we're also going to talk a little bit about AI and then a little Disney and Netflix news. We've got Netflix earnings coming up in less than a month, or about a month exactly, so that should be fascinating. But let's jump into TikTok. So there was an interesting story put up this week, pretty in depth, which I really liked, and it wasn't just a quick blog post And it was really in depth in terms of giving out data details, and the title was TikToks becoming a TV platform one pirate clip at a time. Yeah, and it was really interesting to hear how the younger generation is spending hours watching the long form content on TikTok that they're trapping up into 10 minute pieces.

Speaker 2:

So a little background history here for listeners. So when TikTok first began, the videos on the app were just 15 seconds. That's as long as they allow. Then, in 2017, they bumped the video length up to 60 seconds. In 2018, they maxed out the video length of three minutes. In 2022, tiktok expanded the maximum length of videos to 10 minutes And then in 2023, they quietly removed the ability to record videos up to 10 minutes long And it returned to three minutes, but that's not the end of the video. It returned to three minutes, but that's only recording, so it's still possible to make longer videos, because you could upload videos that were up to 10 minutes long. So what the article talks about is how consumers are using the platform and also getting past some of the issues when it comes to pirating content by chopping up TV shows that are 30 minutes long into three chunks, three parts Movies into six or nine. Yeah, and they're watching it in parts. It's crazy.

Speaker 2:

So two key takeaways I took from the article. One I thought it was interesting how TikTok has experimented with video lengths over the years And on their site there are some numbers fairly recently that I noticed when investigating this topic where they talk about the length of videos that work best on TikTok, and I thought it was really cool how they gave out a lot around engagement. As you and I know, mark and everyone in the industry, we talk about engagement video. Very few to find it.

Speaker 2:

Tiktok was saying for any user who, immediately within the first three seconds of their video, calls out what their product is, what they're pitching, what they're selling or just what they're trying to get across, especially like life hacks, that there's a 67% chance that someone watches that video all the way to the end. So some really interesting stats. You can check those out just by googling latest TikTok engagement stats. But the second thing is that I took away from this the moment you think you know how platform is going to be used by consumers. They do something different. That's right. We would never thought of TikTok as a long form platform. Now, of course, that's not what the majority of the people are doing.

Speaker 3:

Yeah.

Speaker 2:

But I think it's a good reminder to us as an industry that consumers dictate how they want to consume video. We, meaning the industry studios, broadcasters, ott platforms we don't dictate how consumers interact with video.

Speaker 3:

That's right, and they're going to find ways that either suit their style and their preferences or just are kind of innovative. I agree, dan, this article I was really blown away. It is just really well written, there's a lot of detail and it's thought provoking. It was really not full of just opinions and hyperbole but presenting, i think, actual user interviews, because there's a fair amount of quotes in there. But what really struck me was that, on one hand now, for me personally, i think what a terrible experience to watch even a 30 minute show that's cut up into three segments and who knows if the quality is even the same. I haven't even done it, so I don't even have TikTok, but anyway, terrible experience. But it seems like, from reading through the article and the interviews that were done with TikTok users who are doing this, is it part of it? is the experience of being able to watch kind of a short segment and then comment on it Like these feedback loops and people.

Speaker 3:

Yeah, it's the comment piece And yeah, i think somewhere it said something. it's like now you can talk in the theater. That's the comment thread is you're watching it?

Speaker 3:

Yeah, it's the interactivity And that got my wheels spinning like, okay, so how much of this is really about quote piracy? Now, it's a total quickbait title. But you know, okay, obviously, if I owned the rights to this content, you know I'm going to be concerned and you know I would be right, correct, that is, to go pursue my, you know, perhaps legal remedies, you know, for these people who are uploading illegally this content. But I didn't get the impression that they're doing it because somehow they're trying to like get free movies, or that's not it. It's a new experience that they want to have. That's exactly right. You know the same thing. I took away Yeah.

Speaker 2:

The experience, the interactivity, and also it was fascinating to hear. And so, mark, before the podcast, i went and I looked at some of this content and TikTok First, the quality is really good. Oh, okay, it's not a quality problem. Yeah, so you're right, the experience, even for me, i'm like really, i mean an hour and a half on the phone. I just I wouldn't be able to do that unless I really wanted to see the content. But looking at the comments, it was interesting how many were saying, now that I've been exposed to this Oh man, i really got to go watch that movie in full. Yeah, yeah, that's right, it's driving usage, it's driving.

Speaker 3:

Exactly, exactly, yeah, yeah, i don't know That's So. I really like that. We're focusing this, not on, you know. Oh, isn't this terrible? And look, you know, you and I are, you know, in full agreement that if you own the rights, you know if and something's being used illegally, you know by all means you need to protect that That's, you know. So we're not like, we're not trying to advocate for, for, for content theft At the same time. What this really opens up, though, is is just the innovation that the consumer is driving here. You know, now, does it mean everybody wants to watch in 10 minute increments or five minute increments? but probably not. But the point is, they want to be able to interact with the content. Talk, have it be. You know, like, how cool would it be if, while you're watching Netflix, there was, there was a comment thread, there was a comment feed.

Speaker 2:

Now, I wouldn't, i don't know. But yeah, we either.

Speaker 3:

But, but, but clearly these users would, or at least a segment of them would, or you know, he probably so.

Speaker 2:

Also they're on. the difference there, too, is they're on their phone, that's right.

Speaker 1:

If you're now on your computer watching Netflix please, yeah, yeah, yeah, or if you're on, a TV like boy.

Speaker 3:

that'd be a terrible experience to be like one character at a time typing in Oh God, i drive you crazy, just setting up passwords alone is bad enough, but to your point.

Speaker 2:

I also think the content. it's incredible to see how many TV shows were chopped up. Now, if you think about a 30 minute show, Mark, I forget the exact number, What is it? Six minutes of commercials, seven minutes of commercials. So I noticed a lot of shows that are 30 minutes long are chopped up into two. two, That's right.

Speaker 3:

That's only like 10 minutes each. Yeah, yeah, it's like. It's like 21 minutes of broadcast time.

Speaker 2:

Yeah, An entire episode really only in two clips and 20 minutes That's not a bad format to be doing it.

Speaker 3:

Yeah.

Speaker 2:

Now longer movie, two hours long.

Speaker 1:

I think that's very different. Yeah, yeah.

Speaker 2:

Fascinating, though, to see what's happening there, and I just I love the fact that the users are deciding how they're going to use the platform. To your point, and with TikTok's numbers, we know that's not what the majority of people are using it for. Yeah, but it's enough that you're seeing more and more stories about it. Now, flipside, are they really going to cut down on piracy more now? Don't know. Yeah, hard to know, but you mentioned Mark in terms of rights holders and what is the real impact.

Speaker 3:

Yeah.

Speaker 2:

Today we're recording this. Friday, june 16th, extraction 2 came out on Netflix. It only came out in the last 12 hours, or even less than that. I checked out all the torrents just before the podcast And it's there. There's already 40 different versions of extraction 2 up And the rips are all I mean. some of these are with Atmos. You know the quality here. it's everything from 600 megs that's ready for a mobile device.

Speaker 3:

Yeah.

Speaker 2:

All the way up to the largest one here is 27 gigs in size. You got 1080, you got 720. You have all kinds of different audio that you want here, different tracks. So incredible to see just the point of if consumers want to get content, they're going to get it And they're going to figure out a way to use platforms to their advantage.

Speaker 3:

Yeah.

Speaker 2:

So just a reminder that we as an industry do not dictate a lot of that. The consumer is dictated, unless the platform makes it impossible. Now, if TikTok were to go back to the max limit you can do is three minutes in a video, that might kill that overnight. Yeah, i don't see them doing that. Yeah, but possible they could, fascinating. So let's move on. Mark to.

Speaker 2:

I did a blog post this week. A lot of traffic to it, a lot of traffic That's. By the time people listen to this, this is going to be near 50,000 views on just on LinkedIn. Amazing, so interesting. Some of the comments too, that people were posting, frankly, as a little surprised by some of them.

Speaker 2:

So a year ago I wrote a blog post saying that I'm starting to see this new trend. I'm hearing it from CDNs directly And from content owners that content owners are starting to optimize their encoding bit rates and reducing some of their top tier rungs in their bit rate ladders, in addition to using better HEVC compression and other other tactics, just in an effort to save money. Fast forward a year later I wrote that August of last year. Everyone's doing this Right. This is no longer a trend. This has taken hold. We have seen Netflix, we have seen Amazon Prime Video, i've seen it in BBC, we've seen it with some of the largest content owners out there. Some will also talk and talked publicly about it, and the reason they're doing this is one simple thing, and that is to just reduce their infrastructure costs. Now, at the same time, i haven't seen people in the industry or consumer saying they've noticed this And obviously, if you do this right, if you do this well, which I expect these companies to do consumers won't notice.

Speaker 2:

Yeah, but to give you an idea, i've got some companies you all know my name that are public companies that tell me they're saving tens of millions of dollars this year alone Just by optimizing their encoding. Yeah, so my, my post mark is really twofold One, it's just to talk more about what's going on, but second, it's really to let vendors know that, for all the vendors out there that for so many years talked about growing their business on 4k and more 4k adoption and higher bit rates, that didn't come. And if you're still betting on pushing out more bits whether you're a cloud encoding company, storage company, edge compute company, content delivery network you've lost that bet. Yeah, cause these companies are also talking about optimizing it even more in 2024. This is all about justifying every dollar you're spending. Yeah.

Speaker 3:

A hundred percent Years ago. I think a lot of people know my background, the companies I've been associated with, when, when we were initially back in 2000,. You know 13, 14, 15, trying to evangelize this whole concept of reducing bit rate but retaining quality, and yeah, i don't have to go into the details of how that was done, but, um, there was a lot of resistance from, uh, you know from I'll call it the ecosystem, because the CDNs are like now, wait a second. Now, how is this good for us? You know, like, you're saying, you can deliver, and you know, look, i mean, it definitely set us back, because, you know, we realized that we had to even change in some environments, completely change our value prop. You know, because people are like, wait a second, i don't think this is good for us. Now, part of our story, though, which is what is playing out today, was that, look, bit rates are going down. So, whether you play an active part in that and then you're able to participate and you know there's various ways you can monetize that or, um, it's just going to happen. They're going to find ways to reduce bit rate, and that is exactly what's happened, and for some people, it's adopting HEVC more widely.

Speaker 3:

You know AV1 is not a significant enough, you know, for most of the industry to say AV1, but it's coming.

Speaker 3:

So there's advanced codecs. There's some just incredible work being done around um, how we're doing, um, how you know you look at Netflix as kind of the most advanced but you look at how they do content adaptive so that you know one title maybe an optimal bit rate for a certain quality target and resolution is, you know, i'm just going to say two and a half megabits at 1080p, um, but another title it's 3.3 megabits, and then another one it's, you know, back to three and it fluctuates around. And you know these really sophisticated schemes are allowing them to overall get these 20, 30, 40% reduction in bit rates while, frankly, delivering largely the same quality, because the previous techniques and methods, in a lot of cases you were either over encoding, meaning you were spending more bits than was really needed for the quality target. You know that was being delivered Or you know there were titles that were that were not so good, because you know you had ratcheted it down. So, um, i mean, this is just the fact, this is like you say you're presenting the facts, yeah.

Speaker 2:

And the comments were interesting. So you know a couple of people are like well, people are doing this because they want to help the environment. No, they're not Trust me, yeah.

Speaker 3:

None of these are, just not a single company I've talked to have.

Speaker 2:

They mentioned being greener in the environment. It's nice to think that, but come on, let's be realistic. That's not the case. The other thing is you had some people say, well, with all the streaming services losing money, they have to know Netflix projects 3.5 billion dollars, a free cash flow this year. Yeah, they are still doing it. Now, interesting to note, mark on Netflix specifically, they just opened a job for what they're calling senior engineering manager financial planning and revenue infrastructure technology. Wow, netflix says the company is getting started on a multi-year technology and workflow transformation. What they're telling you is we want someone who is experience in infrastructure planning financially. Yep, yep, yep, as any good company does As any good company.

Speaker 2:

Yeah, the days Even a profitable one.

Speaker 3:

Yeah, that's right. Yeah, the days of zero percent interest rate allowing companies to just kind of go do whatever they fancy because money's free, you know, it's flowing down the street They are over and we're not going back to it. And I can absolutely affirm you know, dan, and my work in the industry that these trends, this whole cost reduction, this is not a little blip, this isn't oh well, 2023 is a pullback, it's a slowdown. We need to save money. This is going to continue. Just like you know, regretfully, there's companies that have reduced their workforces. I see, for many of those positions, they're basically not returning, because what's happening is, you know, whether it's depending on the role, whether it's AI or, you know, generative AI, that is, you know, or just simply becoming organizationally more efficient, like they are not going to just begin to bulk up again.

Speaker 2:

Yeah, yes, i think efficiency optimization is key. Doing more with less, that's right. But also look at how Wall Street is valuing these companies. They want you to be efficient. They don't want you wasting dollars. So, no matter how much free cash flow you have, you still can't waste any money. Yeah, that's right.

Speaker 2:

But you know the comments mark on the post started to get okay so maybe I'll just go into in the post if you haven't read it. I also talked about the demand for 4K is extremely little. Yeah Right, and people were like, no, it's not the case. 4k is still the future and whatnot. Yeah, so I listed out every just example top of mind Friday night major league baseball Apple Plus it's not in 4K. Youtube has already said NFL Sunday ticket, when it comes out on YouTube TV will not be in 4K. That's right. Espn Monday night baseball It's not in 4K.

Speaker 1:

It's not in 4K.

Speaker 2:

It's not in 4K. ESPN Plus even when you pay $80 for a UFC fight, it's not in 4K. I could go on and on and on now.

Speaker 3:

The Super Bowl was not in 4K this year, now last year.

Speaker 2:

Right, if they upscale.

Speaker 3:

Yeah.

Speaker 1:

But even when you look at other countries.

Speaker 2:

Obviously, ipl and cricket wasn't in 4K because the average bit rates a mega or less, because so much is on the phone. The point is, 4k serves its purpose in some use cases for select customers, many of whom have to pay more for it. Max charges more for 4K. Netflix charges more for 4K. Yeah, youtube TV charges more for 4K. Now we also know that those aren't the most popular packages If you look at what the ARPA is of those companies. Now, youtube TV, we don't know, but the point is you can argue with it, but this is the data. Now also, akamai and other CDNs went on record for the blog post and actually gave out numbers of what percentage of traffic across their network is in 4K. For some it was 2% of total traffic. Now, 2% of a big number, though, could still be a lot of traffic. Fair point. However, what I also asked them was how much did it grow from the year before? None, it was flat. So that's the reality of 4K in the market. It serves its purpose in some limited cases. But going back to what we talked about before, mark, as an industry we've always pushed 4K. This is what consumers want. What do we just learn from TikTok. It's all about what consumers want, not what we want them to want, and I think of the number one example out there is curved TVs. These manufacturers pushed curved TVs and I don't know of a single person who bought one and I always thought what the hell are you pushing curved TVs on us for 3D TVs? right, those didn't last. So 4K is no different. Will there be use cases for it? Yes, will there be times where it makes sense? There will. But it's also incredible how many executives tell me all the time that there is quote finishing returns off of 4K. Many cases, even for sports. That's an issue as well for sports. So this is the reality in the market.

Speaker 2:

You can argue all you want, but go talk to the CDNs, go talk to the content owners, go talk to the folks who are not using 4K. Ask them why they aren't doing it. It's very clear. It's really not debatable. Is there some value in 4K? They're absolutely. Yes. Sure, will some people pay more for it? They definitely will.

Speaker 2:

But the other thing I'm going to point out here if you ever hear an executive in the streaming space still pits you that quote consumers expect 4K video, or the other thing we've heard oh, if you don't have 4K, users are going to churn, run far away from them because it means they have no concept of what's going on in the market. That is not the reality in the market today, and I'm just going to end with the final point, mark of Netflix. With Netflix's $10 a month plan, the default is 720p. It's not even 1080. And yet, well, we don't have the data from Netflix yet.

Speaker 2:

They keep talking about how they think their basic plan is going to be one of their most popular. So consumers are voting with their wallet. That's the bottom line. So, if you're going into this, i also thought. Talking about the topic of compression, you know Meta put out a blog post recently about their video processor, their scalable video processor that they developed in-house And great blog post, super detailed, and they talked about how they designed it for the needs of video and live streaming workloads specifically at Meta.

Speaker 2:

Also, they talked about how it can be configured to efficiently support, they say, both the high quality transcoding need for VOD as well as the low latency and faster processing times that live streaming requires. So I thought just another interesting example when we're talking about ways companies are optimizing the experience. Now, meta obviously has a different size and scale than most companies out there. They can go out and develop their own ASIC solution, but just another example of where companies are looking at how do they compress video better and do it with fewer bits While not decreasing quality. Yeah, yeah, so great blog post. I am going to put wherever you get my podcast from. I am going to put in the description for the show. I will put up the post to the TikTok and I'll put up the post to Meta, because they're extremely detailed and the Meta one okay. Even that one goes a bit above my head, right, they're drilling down into chipset things, but just great details.

Speaker 3:

Yeah, highly, highly recommend that people really pay close attention to this shift that is happening. And again, you know, there's always the exception proves the rule. Right, it doesn't mean that everybody's moving to Silicon, but there is a very pronounced trend in the market towards dedicated Silicon. And you know again, not to. You know again, point to generative AI. But that is a bellwether And it's an extreme one, because just the volume and the incredible compute capacity required for these large language models. But this, again, this is not a trend, this isn't a little blip in time, this isn't.

Speaker 3:

You know, we're just going to all go back to software running on CPUs. We're going to see more and more purpose silicon, purpose built silicon, asics as they're commonly known, application specific integrated circuits in the network, and the efficiencies are just so profound. Getting back to that word, efficiency, the cost reduction, the op-ex that it's, it's almost too good to be true, you know, if you really dig into it. Google, for example, has been encoding on YouTube, so all YouTube encoding on asics, on Silicon, and one really well respected analyst in the space estimated that they have been able to remove up to 10 million CPUs. Now, that doesn't mean 10 million servers, but that's CPUs because you know their configuration. The point is, think about the energy savings, think about just the capital cost, you know, of deploying an asic. Now, google, of course, spent, you know, hundreds of millions of dollars to produce this right, but at their scale that makes sense.

Speaker 2:

And it probably made that money back the first year.

Speaker 3:

Oh, it's, it's, it's, it's profound, And so this is also a trend that traditionally because a lot of times silicon in years past was adopted from consumer applications And so the quality was not very good, It wasn't very flexible And people kind of, you know, did their best to optimize it for the needs of of real, true broadcast, high quality streaming applications. But this is really changing. There's a whole new generation that's coming, that's in the market, that's going to be in the market And it's something that you know, listeners should definitely be paying attention to.

Speaker 2:

I didn't realize the 10 million number. Yeah, that's big. And and tying into the larger conversation of cost savings, we'll do a podcast soon. Mark on infrastructure, because I'm going to put out an update on the CDM market. But one of the things the CDN The largest CDN see, guess who that is told me the other day was that their power costs in Europe have gone up 30% this year. But what they're being charged in co-location facilities, that's a big number. Yeah, that is a direct impact on the margins of delivering bits over the internet, which are already the margins are so slim. Yeah, yeah, yeah, i same with transcoding.

Speaker 3:

Yeah, bring it in how you're saving 10 million.

Speaker 3:

PUs huge hundred percent. I was Interesting how all these, how the timing is just earlier today is well, on Friday June 16th, i was in a conversation with also a very large and very well-known, very important CDN talking about various things, and they made the comment that in some of their pops, power is the constraint more than rack space. It's, it's just, it's power, and then it doesn't matter, fits in the video industry. This actually is, you know, very well known and spend this case for some time. The point is is that we have to again doing more with less, which is a theme here today Efficiency, and more with less. We have to figure out how to do more video encoding, more Processing. You know all the operations we have to do to stream, you know, high quality video without adding machines. And as long as we're stuck with software and CPU, there's a little bit of optimization that can be done, but it's not enough, only so much.

Speaker 2:

Yeah, well, we've got time for a couple more things, Mark Let's. First and foremost, i want to start with some some neph, so let me just look up here. So the Wall Street Journal and some of their articles are really starting to a no, amy, you know, they put out an article talking about Netflix might do a celebrity golf tournament. Yeah, a live event. Right, it's their first entry. They're getting into live sports.

Speaker 2:

Sports was the headline sports, yeah right, but others picked up on it and here were some of the the headlines Following Netflix's celebrity golf tournament announcement. Netflix has not made an announcement, people none. I don't get why the media just can't get this right. There's been no announcement of any kind. We don't know if they're doing event Maybe they are We just don't know. It's not a definitive. What we do have is a definitive is Netflix did announce That they will do their second quarter earnings on July 19th. So we've got just about a month from today. I Think, mark, we're gonna get some good information on that around the avod, the password sharing, because they gave out details and avod Not even on a you know an earnings call. They're willing to give that out during their upfront presentation, which I thought was was great, and that is not in years past. On, netflix gave out information. So I'm hoping we get more information on password. I think we will, especially with what we've seen with their stock. I think they're gonna have to address password sharing, just what they've seen so far, and then, really quick, only to some more media nonsense. So Disney CFO Was announced yesterday is taking a leave of absence due to a sick family member.

Speaker 2:

Headlines I read today was things go from bad to worse for Disney. What Another one was. Disney CFO leaves after tense relations with executives. It again Wall Street Journal Said things like a person familiar with the CTO situation said there have been no dramatic changes in her life recently That would require her to step back. Then they say the abrupt exit caught some colleagues and associates by surprise. Okay, so now it's making it sound like wow, this is just so, so strange. No one knows what's going on. Then in the next sentence they say the CFO has an ailing husband who's been in health care facilities since the start of the year. Well, this person's been at Disney 23 years. They've been the CFO for more than eight years. This idea that they're leaving for personal medical reasons, that all of a sudden this is a bad thing for the company Or they're leaving because they're fighting with the executives It's just absurd.

Speaker 3:

Yeah, by the way, she's staying on as an advisor. Right, so normally yeah, exactly, normally, if there's you know something there, you don't keep the person around as an advisor Normally, especially a CFO and They're obviously gonna be okay.

Speaker 2:

They mean they, they have an interim CEO who's been at the company. This happens at public companies. But for these titles of just things go from bad for worse, or you know, this person has no reason for leaving. First of all, it's their own reason. They want to leave. It's personal Right. No one needs an explanation. Yeah, you can leave your job and just say I need to move on, like you don't owe anyone an explanation, even in a public company. So the way some, some of these and Wall Street Journal wasn't the only one to write that tone, which personally I didn't like They have the right to do whatever they want. So Disney is going to be fine, but interesting stuff again with the media, right, it's just. It's clearly it's for headlines. I mean, one headline was just things go from bad for worse to Disney. That is, it's clickbait, it's all the clickbait.

Speaker 2:

Let's talk AI mark. in the last few minutes we have yeah, you mentioned it a few times, so So an interesting thing happened over the last few days. So Microsoft's finance chief Said that she expects physical fourth quarter growth for Azure cloud of 26 to 27 percent year over year in constant currency And then said expects one percentage point of it Coming from AI services. so very small number. Then, in a public discussion That the finance chief had with Microsoft's technology chief, kevin Scott, they provided more specific saying quote the next generation AI business will be the fastest growing 10 billion business in our history. And, of course, what do you think happened to this dark price? It's all boost, yeah, it got upgrades, and yet they didn't say anything else. They just said that'll be the fastest growing Business in their history. they didn't say over what period of time, they didn't say from what product. So someone asked an additional question, which I love, right, great question it. and their response to the question was Because it's a really a very general platform, we have lots of different ways that 10 billion dollars Of average recurring revenue is going to show up. first What? You didn't answer the question And yet the news of that day. it was like a day or two days ago. before recording this, all over was Microsoft and AI and they're going to see a boost in revenue And maybe it is their fastest growing 10 billion business in history. but maybe that takes five years, seven years, 10 years, i don't know. But the AI stuff to me is just starting to get a little out of control.

Speaker 2:

Adobe came out, adobe at earnings, great earnings for this quarter, although next quarter. Wall Street was kind of tempered. Adobe gave out guidance, just mid-range. Wall Street was expecting higher. But Adobe said quote Adobe's groundbreaking innovation positions us to lead the new era of AI, given our rich data sets, foundation models and ubiquitous product interfaces. Yeah, okay, great. What does that mean? Adobe? I don't know. Now Adobe has rolled out in beta some interesting tools I tried them out today Having to do with, obviously, creating photos, editing photos Interesting, no doubt.

Speaker 2:

But my point here, mark, is and then I'll let you go and what you think of AI. It's too early, we don't know how it's going to impact the streaming industry, the video industry, and I think companies and people have to keep their eye on the ball of what's on today. Don't be developing for something that's not here, it's unproven, we don't know that it works at scale And we don't know the best applications where it's going to be used. And then, finally, most importantly, we don't know that content owners, broadcasters and others who are using a video stack of some kind are willing to pay more for it. They might expect it as part of functionality of platforms are already using. So if you think AI is going to grow your business from a revenue standpoint, it may not.

Speaker 3:

Yeah, that's actually. That's actually an interesting point. There's a lot to unpack there. But going back to Microsoft's, you know this Amy Hood who is? she is a CFO right?

Speaker 2:

Or is she just, like she should, be CFO Anyway?

Speaker 3:

anyway, she's obviously senior enough to be speaking on behalf of a, you know, trillion dollar market cap company, so she's qualified to be making comments. Yes, she's a CFO. She's CFO, okay. So you know, i think of Microsoft's relationship with open AI and their investment, and you know. The only thing I just wonder is if some of this vagueness is that they need to be a little bit careful with forecasting too much.

Speaker 2:

I would think so The business of their major partner, open AI you know what do you give out a 10 billion number, Because now you put a number out there.

Speaker 3:

Well, because, because it could be the open AI is 9 billion of that is going to be is on the path. I don't know. That's just what comes to mind. I you know. I won't comment further, but that is. that is an interesting reality there And it is very clear, like if you go to Azure, there's banners all over the Azure website about webinars they're hosting and interviews about how you know it's a heat.

Speaker 3:

I mean, it is very clear that Microsoft is doubling, tripling, quadrupling down on AI, So you know, in terms of Azure, and then, of course, look, they're introducing co-pilot across the office suite. I mean, so it's very clear, So I'm so. None of it surprises me, And I my I guess what I'm really saying is the vagueness that you picked up on. I just wonder if it's like they just have to be very careful because I would get one customer, It's open AI and then and then themselves, they have two customers They have.

Speaker 3:

Microsoft and then they have open AI.

Speaker 2:

So yeah, I just hope it doesn't start a trend mark where, you know, back in the days, everyone through blockchain enabled into the video business.

Speaker 3:

Everything was like hey, that's exposure.

Speaker 2:

Right, it'll boost our stock price, and then it was like okay, that didn't work. I just hope vendors in our space Yeah, that's the ones I care about the most Don't start throwing in AI into everything, because that doesn't generate currently any additional revenue or interest in. Everybody says everything with AI. It doesn't mean anything. At this point. You have to define it, yeah.

Speaker 3:

Yeah, it's true, and, and to that point, you know, you bring up a really interesting perspective and and I tend to believe that this is largely what's going to happen Is it AI enabled or AI? you know, supercharged is just a it's required, it's not something that, oh, I can charge more because of it, it's just required, and part of that's going to come because, you know, look, you can go to Bard today. You know which is Google's response to open AI It's totally free. Yeah, you can argue about whether it's as good and all that, but set that aside, it still is Google and it's built on a serious, large language model models. So, but it's totally free. You know, even open AI is free. You pay for the pro. It's 20 bucks a month, and if someone's really yeah, it's cheap.

Speaker 3:

The point is is that now, you know, are these going to get monetized more deeply? And you know, yes, and I believe that where the monetization can happen is where there's where there's real specialization. So I mean, i'm just making something up on the fly, but if I'm a lawyer and there's something that normally I have to pay a law clerk to do, that, now I don't have to pay that law clerk to do. you know, that's going to have a different level of value than you know. occasionally, i want to summarize a blog post, you know, and so I copy and paste text. It's just, it's it's different, right?

Speaker 2:

So yeah, it's different use case you know.

Speaker 3:

and then if we talk about our, our business, you know, yeah, if there's something that's so transformational in how I'm I'm encoding video, how I'm streaming video, how I'm, you know, subtitling my video, how I'm, you know, whatever the primary functions are, yeah, there's probably value there. People will pay for that. But just saying, oh, it's an AI, fill in the blank, you know, ai encoder, and that's why it's 50% more Like. I don't think so.

Speaker 2:

Yeah, i don't think that'll work. I almost think of AI to some degree, depending on the application use cases like ultra low latency or even just low latency. You know, when I surveyed hundreds of customers, 90% said I expect that as part of my CDN service, i'm not paying more Yeah. And we saw CDNs come out after years of trying more saying we actually couldn't get more per bit Yeah. Now I think AI has a possibility to be very different in the sense that, to your point, you could use it for transcriptions, closed captioning, transcoding There's a lot of different ways. You could provide intelligence in a video stack Sure. And if you can really show a difference there, where it helps monetize or save money or whatnot, sure, I think the right customer for the right application would pay an additional fee for it. But I think we're a long way away from an industry knowing what those best applications are, how they can be utilized. Also, just a reminder that AI is not proven. So you're implementing any new feature functionality into a workflow that has been around 10, 12, 15 years at scale. Well, that means it has to work at scale. It has to be proven, right. So there's a lot of unknowns with this.

Speaker 2:

So I think it's interesting. I mean, we're obviously going to keep an eye in it as an industry, but the fact I just to me, mark, i thought it was so fascinating that Microsoft was the first company I've seen, of the ones that I'm looking at, to actually come out and give a revenue number which I had not seen before. Maybe others I mean obviously in the video, but that's, i look at that as a very different business. Yeah, that's very different. Yeah, so interesting. A lot of stuff this week. It's incredible, you know, i got to say I'm thankful to wake up every day and still be in this industry 26 years later, because even when no news comes out, there's always so much taking place.

Speaker 2:

Yeah, Yeah, there's always so much to talk about and think about and feedback I get from customers and what they're actually doing and what they like and what they pay and which vendors they're using and what's working, what isn't, what they're working on the future, what their problems are right now that keep them up at night. It's just, it's fascinating the amount of information that's out there. So, mark, coming up on a podcast soon, i'm working on doing a long blog post on a wrap, wrapping it up right now, of just the current state of the CDM market. We're going to go through some competitive landscape where some vendors are go through breakdown some of their revenue. I'm also going to talk about some of the pricing in the market. You have companies out there who are not growing revenue year over year. There's reasons for that. It's not a quality problem, it's just the state of the market. The market's not as big as something. From a TAM standpoint, compound annual growth rate I'm projecting this year is between 3% to 5% at most for CDM market. That's not a lot. We've got some good, interesting figures from some of the public CDNs. Fascly has an investor relations day next week, so I don't know what information will come out there, but there'll be something I'm sure that we can use as well. So in the next week or two not next week, maybe the one after we'll do an entire podcast I think we'll just call it the latest on infrastructure. Yeah Right, let's talk cost pricing, some competitive landscape, what we're seeing in the market, stock prices as well. I'm sure we'll get a lot of comments on that Once I post the CDN stuff to the blog I always do, But that's what we have coming up.

Speaker 2:

And then, mark, you and I are going to take a one week break at the end of the week the end of this month, since we're traveling. So there will be a one week break in the podcast, then we'll be back. And the Netflix earnings on the 19th. That entire week there's going to be a couple different earnings coming out. Netflix is always the first, but get ready for a bunch of data from Mark and I to break down at the end of July. There's definitely going to be a lot. That'd be great. So with that, mark, thank you, we're out of time. You have any questions? Hit Mark and I up All the stuff we talked about today. I already have it up on LinkedIn. You can reach out to us directly or just through LinkedIn You have any questions. We're happy to help you out. We appreciate everyone listening and we'll be back next week. Have a great rest of your week.

Speaker 1:

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.