Harmonic Inc. (HLIT) Earnings Call Transcript & Summary

Good morning, good afternoon. I'm Sanjay Kalra, Chief Financial Officer of Harmonic. Welcome to Harmonic's 2021 Virtual Video Investor Event. It's great to have you joining us today. Before we continue, I must mention that today's presentation will include forward-looking statements. Actual results may differ materially from these forward-looking statements. I refer you to our most recent forms 10-K, 10-Q and 8-Ks filed with SEC for more information regarding our risk factors. During today's event, we'll provide details on video segment's strategic vision, growth drivers, investment initiatives and long-term financial model. Also speaking today will be Senior Vice President Video Products and Corporate Development, Shahar Bar; and our President and CEO, Patrick Harshman. After the presentation, in the interest of time, we will take questions from sell-side analysts. I will now hand it over to Patrick. Patrick?

Okay. Well, thanks very much, Sanjay. And let me add my welcome to all of you, and thank you. Thanks for joining us today. As Sanjay outlined, today's event will be focused on the video part of our business. And before we dive into the really the detail and the substance of that, simply, for those of you who are maybe newer to Harmonic, begin with a high level at-a-glance overview of the company. We operate 2 related, but distinct business units, Cable Access, which will be the subject of a companion investor event a week from today; and our video business, which is indeed the subject of today's conversation and meeting. We're a Silicon-Valley-headquartered company. The 2021 guidance -- revenue guidance, excuse me, that we pointed to is a little over $450 million, up 20% or a little north of 20% year-over-year. We do our business around the globe with leading media companies, broadcasters as well as service providers. And our market capitalization is somewhat above $700 million. Very importantly, and really underlying the substance of today's meeting, the company, although we've been around for a while, several years ago, we undertook very significant new waves of innovation all around taking advantage of cloud native technology, in particular, targeted cloud native technology, enabling new transformations for the industry around both broadband access and video streaming. And we think that these investments that we've made have put us in a -- really a position that we haven't been in for some time. We're tremendously excited about the future. And we're very excited, both today and tomorrow just to explain to you why we think that these investments and the corresponding market development that we've undertaken over the last couple of years, have really put us in a place where we have a platform for a new wave of compelling growth. So let's look at the video business and key takeaways that we want to convey over the course of our meeting today next. Really, the fame is enabling premium video streaming. And it really is a streaming part of the market that constitutes the next-generation opportunity for this business, the incremental growth opportunity for the business. Over the course of today, we hope to show you that: number one, we're targeting a rapidly growing streaming infrastructure market that, by itself, represents greater than $1 billion of additional addressable market by 2024; that in targeting this market, we have truly differentiated technologically leading solution, in particular, our SaaS streaming platform, where we're already beginning to see rapid customer adoption. Combination of this compelling market and our compelling solution means that we're quite confident that we will be able to drive over 30% annual growth rates, leading to greater than $100 million in streaming revenue by 2024, the majority of which will be recurring SaaS revenue. Taking this SaaS, together with the recurring revenue portion of our existing broadcast in media business that we already have in place, we anticipate that, by 2024, over 50% of our total revenue will be recurring. And last, but certainly not least, all of this is really in addition to our existing broadcast market, which -- where we continue to be a leader. And our business plan, as Shahar will describe in some more detail, we continue to take advantage of the opportunity to capitalize on our very strong brand, technology and presence to continue to drive that as a very important revenue and gross profit foundation for us. So with that overview, let me now turn it over to Shahar to talk to you about the markets and Harmonic's unique position in the market, and the reason for our optimism about our future prospects. Shahar?

Thank you, Patrick. As Patrick mentioned, we're going to talk about the streaming infrastructure market, which is a compelling growth opportunity for us as well as the broadcast infrastructure market, which we operate in today. And as a starting point, we're going to take a look at the video delivery infrastructure ecosystem and see how has been impacted and changed by the recent trends. So the video ecosystem is composed of 2 main players, originally, the content programmers as well as the pay-TV providers, also known as MVPDs, multichannel video programming distributors. The content programmers were assembling content from a range of sources from their own studios, from news events, from stadiums. They would put them, stitch them together, assemble them, insert advertisements, creating a linear channel, which they then would send over satellite to the pay-TV providers. Those pay-TV providers would take content and linear channels from a range of content programmers around the globe, bundle them together into different packages and then sell them to their consumers over their broadcast networks. Both the content programmers and the pay-TV providers deployed all their solutions in their own facilities historically with proprietary or dedicated hardware and software systems. This has been the ecosystem in place for many years and profitable for both of these entities. What's changed right now and is the next -- the ecosystem is the introduction of a few disruptions. The first and foremost is the introduction of streaming. We now have a new player in the ecosystem, the streaming player, or also known as the virtual MVPD. These are new entities that basically license content or created content on their own, and then decided to stream it directly over for the Internet or what's known as CDNs, content delivery networks, directly to consumers. That has completely upended the ecosystem. It also provided an opportunity for the content programmers themselves to stream directly to consumers and circumvent the traditional distributors. The second major trend that has upended the traditional ecosystem is the introduction of cloud, and that's more from a technology perspective. Over the recent years, the evolution of the cloud and its maturity has allowed all these 3 players, the programmers, the traditional pay-TV and the streaming virtual MVPD players, to utilize the cloud for a range of video workflows, and that has shifted the workload from traditional facilities on to the cloud. And finally, the government has also had an impact on this traditional ecosystem. Recently, as an example, the FCC has taken away some of the infrastructure, the satellite frequency available to content programmers to distribute their content and has taken that into the 5G wireless broadband world. That has resulted in the content programmers having to innovate and find new and more efficient ways to distribute their content. The 2 major trends that are going to be with us for the long haul from this ecosystem evolution is really the streaming, which is going to become the dominant form of video and the rapid adoption of cloud in the infrastructure space of video. Looking at the streaming side first, a recent report by Omidia estimates that by 2024, there will be 400 million additional streaming subscribers above and beyond the global pay-TV subscribers. And as we look at the outer years of the decade, that gap is only going to get bigger. The simplicity, the ease, the price points, the convenience of the streaming application just makes it such a compelling proposition for consumers. On the technology side, the innovation, the investment and the capital that the 3 major cloud vendors are pouring into their infrastructure, makes media infrastructure very, very valid for the cloud, especially for variable workflows, which streaming is exactly that. So we're now going to pivot understanding the main trends, we're now going to understand how these changes affect companies such as ourselves in the infrastructure market. The first we're going to look at is the broadcast infrastructure market and see how those changes have impacted us. First, to understand about the broadcast infrastructure market, which we've operated in for many years, the foundation is the linear channel. As mentioned earlier, content programmers would assemble a linear channel from a range of sources, pre-recorded programs, new studios, live sporting events, insert advertising and create a linear channel. That technologically is known as playout. They would then compress that linear channel, also known technologically as transcoding, and sent it to the MVPDs, the pay-TV providers, who, in turn, would aggregate all the channels they could get, and they would also compress them, or transcode them and serve them as bundles to their consumers. The infrastructure was sold as appliances, software or hardware appliances. They will be deployed on-prem at their facilities of the programmer, or the MVPD. They were sold in a CapEx model with additional recurring SLA, and the total addressable market was the total number of channels, linear channels, that the programmers were creating around the globe and distributing to the pay-TV providers. And then the total number of channels the pay-TV providers would take off these satellites, recompress, retranscode and send it to their consumers. Historically, that was a very profitable and growing business as more every year, we would see a growing number of linear broadcast channels in the market. That growth has essentially plateaued, and that leads us into this slide, which is the key trends that are impacting the broadcast market. The linear channel growth that we saw for tens of years is no longer prevalent. We're now seeing it plateau and slightly decline. The focus has shifted into the second area, which is efficiency. A lot of the content programmers and pay-TV providers continue to generate very healthy revenue and, more importantly, profits from the traditional broadcast ecosystem, so much so that they're using those profits to fund some of their next-generation endeavors such as direct-to-consumer streaming. Therefore, the importance of squeezing profits out of the broadcast ecosystem is very strong. And what we're seeing right now is an extensive refresh cycle of the traditional broadcast infrastructure underway to refresh a multibillion-dollar installed base and just generate as much profit as possible from that installed base. So many of the broadcasters and pay-TV providers around the world are looking at their facilities, how do I take this traditional legacy equipment, repurpose it and deploy new equipment that has a fraction of the footprint, is easier to operate, requires a lot less OpEx and overhead and can generate more profit for me for my critical future endeavors. And finally, we move to the third trend, which is similar to the second, but it is basically through the government intervention, we're seeing, again, a push forward or a pull forward of the refresh cycle. By the government pulling away certain infrastructure such as satellite frequencies or terrestrial frequencies, they are forcing the traditional players, programmers and pay-TV providers to refresh the equipment and find innovative ways to deliver their traditional broadcast services. So overall, in the broadcast infrastructure market, you have a headwind, which is the plateau and slight decline of the investment in linear broadcast channels. However, you've got a couple of tailwinds, which is the massive refresh cycle underway, which is also fueled by the government intervention to accelerate that refresh. Pivoting over to the streaming market, we see a completely different system. We still have the traditional playout in transcoding, which is here on the screen on the slide in the light blue colors, the playout and the transcoding. That's still relevant. You're still creating channels, you're still creating VOD libraries in the streaming world. However, there is the next part, which is that target about the actual stream, which is known in the industry as a unicast stream, the specific stream that each and every consumer gets, which is differentiated. Each one of us, as subscribers, we each have our own unicast stream. The first thing that needs to happen in the infrastructure side for streaming is we need to take that channel that was created, or that event, and we need to package it towards the end consumer device. Every end consumer device, be it a Roku, a tablet, an android or an iOS, needs a different packaging. And then we need to serve it and hand it off to the content delivery network, the CDN. That is known as the origin function. Both of those are in that dark blue square on the screen. But before we hand it off to the CDN, we are now inserting video advertisements, targeted personal advertisements into the stream. And then we hand it over to the CDN for delivery to the consumer. These are not overlay banners. These are actual adverts that are inserted at the right place in the video signal for the consumers to watch and to be targeted with those ads. So now the ecosystem for the infrastructure of streaming is really composed of 3 parts. We've got the media processing, we've got the packaging and origin, and we've got the targeted ads. What is also different about the streaming ecosystem is the majority of the streaming platforms are porting, if not already, to the public cloud. And in addition to that, and as a result of that, the business model is no longer CapEx. It's actually usage-based. It's metered, it's variable and it's a SaaS -- all based on SaaS-based platform. As a result, the traditional media processing, we, Harmonic, sold as CapEx in the broadcast world, we're now selling it as $1 per hour of compute. The same thing regarding the traffic, we're actually selling that, the packaging and the origin based on gigabytes. How many streams are we processing, packaging and serving off to the CDNs? Those gigabytes are charged to the customer. And finally, the targeted adds are charged per CPM, which is the cost per 1,000 impressions. And impression is essentially a verification that the targeted personal ad that was served to that consumer was viewed by that consumer. Once that confirmation is set, the content owner gets compensated for that ad for inserting into the ad for the stream. And we, as an infrastructure provider, we get compensated for facilitating that. All these changes have dramatically shifted the addressable market. We still have the linear channels, that's still relevant and still there. But now we actually have a little bit more media processing. There's a lot of live events that are now occurring in the streaming world that did not make it into the traditional broadcast ecosystem. There are larger VOD libraries. All those are in the media processing area. But of course, the bigger ones are the packaging and the origin, all the gigabytes and gigabytes of streaming that we are serving off to the CDNs. And the rapidly growing market of targeted ads and streaming, so the CPMs and ad impressions. Those have dramatically changed the overall market. The key trends in streaming are quite different from the ones in broadcast. And if I have to summarize in one sentence, in the world of streaming, we have a lot more streaming platforms than we ever had programmers around the world, watched by a lot more subscribers, being fed a lot more targeted ads. Let's start with the first one, the streaming platforms. In the traditional broadcast world, if you were a new programmer, you would have to spend an enormous amount of capital to set up your facility, to build the teams, to buy the equipment, to get regulatory approval just to launch 1 or 2 new linear channels. In the world of streaming, that barrier to entry is dramatically lower. If you have content, you can contact a company such as Harmonic, we'll take your live or file or VOD assets, we'll place them on the cloud, and we'll be able to stream them and deploy them around the globe within minutes. So the barrier to entry is so much lower, which creates a very significant and long tail of streaming platforms. They could be regional, they couldn't be niche to specific areas. But overall, there are thousands and thousands of more streaming platforms available around the globe versus what was traditional programmers. There's going to be more subscribers watching those. First of all, there's a lot more people interested in these niche and global streaming video content out there, plus what we spoke about earlier, the simplicity, the price, the mobile adoption has made just streaming just a rapid growth from a subscriber perspective. As you have more subscribers as a content owner, you now have an opportunity to monetize that content by serving targeted ads to your subscriber base. The advertisers around the world are eager to find more and more attractive mechanisms to deliver targeted ads. And the content owners are eager to monetize their content. That serves a very unique combination, which is feeding into the world of streaming where consumers are watching the content, they're glued to the screens. They're very interested in the content, and they're being served the ad. So it creates a great opportunity for both the advertisers and the content owners to monetize. And that leads us to the rapidly changing TAM and addressable TAM for us, Harmonic, as an infrastructure supplier. What we're showing here in the chart are both of the TAMs of the broadcast in light blue, infrastructure market, as well as the dark blue, the streaming infrastructure market TAM. If we look first at the broadcast TAM, we can see that between 2020 and 2024, we're estimating a single-digit decline percentage-wise annually from -- and around a $200 million decline in the TAM between 2020 and 2024. However, on the streaming market, we see a completely different trend. We see a rapid growth, nearly doubling between those years and reaching close to $1.1 billion. The streaming infrastructure TAM is driven by everything I just talked about, more streaming platforms watched by more subscribers and fed more targeted ads. So let's look at the 2024 year specifically to understand how that TAM is composed of those 3 parts. What we see here is those same 3 parts I spoke about in the previous slide. The media processing we estimate in 2024 to be represent about 40% of that $1.1 billion TAM, around $430 million. A relatively similar amount will go to the traffic, which means the packaging of the streams and the origin essentially serving them and feeding them to the global content delivery network, CDN. And then we also see this new yellow part, the targeted ads. As systems become more mature, more ads are being inserted, and that part of the TAM is growing over time. So we see how the TAM of streaming is completely different from the TAM of broadcast infrastructure, which was mainly focused on just the media processing. The TAM of the streaming infrastructure market is quite more diverse, and it has very significant growth trends. To really illustrate this, we're talking about here a specific case study and a recent Tier 1 streaming project that Harmonic launched. So first, we launched it on our VOS360, which is our brand name for our streaming platform, which Patrick mentioned earlier that we started development on several years ago. We launched it on 2 or dual Azure cloud instances, which essentially means that we're launching it on 2 separate Azure clouds. We're deploying the same exact service and software, and we're keeping those clouds in sync, which results in exceptional reliability as consumers can get the feeds from one or the other, and they're essentially resulting in no downtime. From the media processing part, there's a few of tens of live channels, there's hundreds of premium sporting events, and there's a relatively large video-on-demand library with over 60,000 assets. On the traffic side, the packaging and origin, there are several hundred thousand subscribers that join the system through their apps on connected TVs mostly. On peak hours during the peak sporting events, there was around 0.25 million simultaneous users on the platform, spending somewhere between 2 to 3 hours on the platform, representing the duration of the sporting event. And during the first month of launch, Harmonic's VOS360 injected -- inserted into those streams that we handed over to the CDNs around a total of 175 million advertisements or ad impressions. If we look at the invoice that we recently sent to this customer in May -- at the end of May of this year, we can see that the media processing portion represented just around 14% of the total invoice. The traffic represented about 1/3 of our invoice to the customer. However, the ads represented a little bit over half, so around 53% of the invoice. This pie chart is quite different from the pie chart I showed in the TAM where the ads was around 20% of the total TAM. And that illustrates that this customer has what I consider to be a mature system in the sense that they have very attractive content, advertisers are aware of it and are very eager to advertise on their platform. As a result, this content owner is able to monetize the content effectively well and inject billions essentially of ads per year to their subscribers. And that goes to the evolution of all streaming platforms. If a greenfield streaming platform starts out, typically, the first month of invoice will be almost entirely media processing, that light blue. They have to set that up. Those are the linear channels, the VOD libraries, the events, you have to process the video. So the first month's invoice may be 90%, 95% just media processing. As they grow, as they get more subscribers, the dark blue, the traffic one, will start to grow, and over time, we'll start seeing that take probably have maybe a little more than the invoice. And at the right time, and in line with that content owner's business model, they may and very likely will introduce targeted advertising, at which point, we will start to see this yellow, the ad part of the invoice, grow. And the more popular the content, the more ads are being inserted, and therefore, that will become a bigger part of the monthly invoice. And that is -- essentially shows how the TAM is evolving in this really attractive streaming infrastructure market. So now that we understood how the streaming infrastructure market works, how the TAM works, let's look at our products and why, specifically, we are quite confident in our ability to do very well in this market? So the product, as I mentioned earlier, is known as VOS360. We started development on VOS360, which was an extensive investment around 2015. Actually, coincidentally, it's around the same time that we started development on the CableOS side. Our first customers -- product was ready around 2017, and we onboarded our first large customers in 2017. And you can see those logos here, Fox and AT&T TV were our first large customers on the platform. Every year since then, we've started to add more and more customers on the platform and with an accelerating pace. You can see some of the logos here that we've placed on the chart. This isn't the full customer base. It's just a subset of our total customers. Some of the logos, you'll see here are very well-known brands, maybe one of the world's top brands. Some of the logos are second-tier brands, which you're probably familiar with. And some are brand -- are names that probably you've never heard of, and that goes to my point about the amount of streaming platforms out there. You're going to have the Tier 1s, you're going to have Tier 2s, and then you're going to have a very, very extensive long tail of regional niche-focused brands, and we go after all of them with our VOS360 platform. So the VOS360 platform is a fully managed, cloud-based streaming SaaS infrastructure platform. And there's 3 main differentiators to this platform versus anything else out there in the market. The first and foremost is the investment we've put into it. To date, we've put in over 1,500 engineering years into the platform. That has made it an extensively reliable, very feature-rich and a very advanced platform, often much more advanced than anything out there in the market. The second and quite an important differentiator that we decided to invest on early is cloud neutrality, which turns out to be quite important as time progresses. What does that mean, cloud neutrality? It means that we are able to deploy our VOS360 platform on Azure, Google Cloud platform, also known as GCP, and AWS, all 3. That is quite important for many customers. Some have their own long-term contracts with one of the cloud vendors, and we are deploying for their systems, the VOS360 on their cloud accounts. Some actually want to run on dual clouds, and they want to have the flexibility to be able to run the VOS360 on one or the other. Some may only run on one, but they won't know that they will have to shift over time as the company that they work for, the media company will pick different clouds every few years to optimize their costs. For those customers who don't have a preference, then we decide, which cloud we'll deploy on based on geography, location and the application. The third area that we've differentiated ourselves in the market is DevOps or operations. So we don't only provide the software on the public cloud, we actually operate the infrastructure we provide on the cloud. We provide a full-blown software platform. And so that means that we ensure that our software operates well. We also ensure that the cloud infrastructure operates well and everything that connects it to. So the compute nodes, the load balancers, the storage systems as well as all our software that utilizes those cloud components. And as a result, we provide essentially a service with a service level agreement to the customers. We have uptime commitments for every one of our deployments. If we don't meet them, we pay penalties. The benefit for us is that a lot of the customers are very eager to utilize our DevOps and essentially move away from the old model in broadcast infrastructure where they would buy products and have to spend a fortune operating them just for the infrastructure. Now they are able to utilize their engineers to focus on their core competence, the apps, the recommendation engines, the consumer-facing side, while the infrastructure is essentially being outsourced to our SaaS platform. A little bit of stats on the SaaS platform from 2020. Overall, the usage on the platform grew by over 40% in 2020. That represents the total amount of invoicing that we set between 2020 versus 2019, so it grew by over 40%. Our egress, which is another technological term for the traffic or how much traffic gigabytes we sent out of the public cloud, grew by 50% in 2020. Our customers nearly doubled during that year. The number of linear channels we placed on the platform went up by over 3x or 345%, which shows that the linear channel world is not over, it's simply pivoting from broadcast to streaming. The number of locations we deployed out went up by 66%. Those are the locations of either Azure, GCP or AWS. And based on our DevOps capabilities, generically, globally around the world, our uptime was well above 99.99%. And for those customers as the case study I showed earlier, the one with a geo-redundant cloud instance, our uptimes are significantly higher than that, well above 99.9999%. So very rapid growth we've seen in 2020, which has continued in 2021. Our go-to-market is composed of 3 main pillars. First and foremost, we've been in the media industry for a long time. And as a result, we have a very strong relationship with many of the media companies around the world, and we have a direct sales force to address those media companies. And we approach those same media companies, whether pay-TV providers or programmers or content owners directly, and we pitch them and sell them our streaming SaaS platform. The second pillar of our go-to-market are the cloud platforms, specifically Azure and, more recently, a little bit of Google GCP. So when these companies sign up a very large contract, say, with a content program or a pay-TV provider, they are eager to showcase that there are many SaaS platforms available on their clouds for applications that they can use. And we are one of the premier Tier 1s that they showcase to the customers that have signed up large contracts with them. We also work very closely with Microsoft Azure as well as GCP to compete against AWS' native media services, which I will talk about in detail on the next slide. And then for the long tail, we are extensively building a partner network. These are a mix of regional partners, resellers, service integrators, OVPs, which means online video platforms. These are end-to-end providers, which go in their region or their focus area and provide an end-to-end streaming service to the customer. That could be including the app or that could include the recommendation, the billing, payments, e-commerce, a lot of components. What we do for them is simply reduce the headache of the infrastructure. They deploy the VOS360 on their cloud accounts or on ours. They connect to us through APIs, and then they're able to focus solely on their core competence, the application, while we do the infrastructure. That relationship enables us to go after a lot bigger part of the market, a lot of the logos, the smaller logos out there around the world that are growing rapidly. Moving to the competition side. There's 3, again, major groups. We'll talk about the first one, AWS Media Services. It's very important for me to point out that is not AWS Cloud. AWS Cloud is actually a strong partner. Actually, there is a media summit that they're initiating in the next couple of weeks, and they've invited me personally to speak as a guest speaker on this summit. What we do compete with is the AWS Media Services, which is a separate sales force and a separate sort of subsidiary that was part of an acquisition they did in 2017, that resulted in native media services available on the AWS for streaming purposes, and we do compete against that subgroup within the Amazon Corporation. How do we compete against them? First and foremost, it's important to point out that if our customers have selected Azure or GCP, AWS Media Services is not relevant. They only provide their services as native technology on AWS. So therefore, if the customer has selected those other clouds, we don't compete against the AWS Media Services group. The same goes for if the customer selected a range of private data centers, where, again, AWS Media Services do not apply. If the customer has selected an AWS as a cloud infrastructure, then yes, we will compete against the media services group, and we do so with our differentiation, our feature richness, our operations, DevOps, which AWS Media Services does not offer as an example, and our cloud portability. If it's a greenfield customer, which has yet to pick a specific cloud, we will partner very closely with Azure as well as GCP to combat AWS coming in with their infrastructure and the media services group. This model of competition and cooperation with AWS has been in place for 2 years, and we performed quite well in the market. The other competitors that you see here, the legacy broadcast infrastructure competitors, I put some of the logos here. I would say there are not very strong competitors in this new area of cloud, SaaS and streaming. And the best analogy is comparing it to the other division that we have at Harmonic, the CMTS, where we introduced the CableOS around 2017, which was a virtual CMTS technology, while the traditional CMTS players were in the hardware model integrated CMTS. That approach of moving first allowed the CableOS team to be several years ahead and retain that lead moving forward. I would -- versus the traditional competitors, I would compare that to exactly what we've done on the VOS360. We started early. We put in a huge amount of R&D investment, and we have a multiyear lead ahead of the traditional competitors. Lastly, a small part of the market, but what's important to note it is to do it yourself. Originally, a lot of the streaming platforms that started out 7, 8, 9 years ago, had to do a lot of the technology themselves. Very few of them have kept up with that development. It requires a lot of investment, it requires DevOps, it requires a lot of engineer -- software engineers to develop the technology and to constantly improve it and adapt to the new formats and recommendations, and technology -- just a rapid evolution on the Android and iOS ecosystem. So most of outsources 2 companies such as ourselves, the ones who continue to focus on their own in-house development and their own DevOps are really the largest of the largest, the Tier 0s, what I consider the Netflix, the Disney+ of the world. There are few and far apart, and we don't believe the majority of the market will follow them. So that was about streaming, and I want to pivot over a little bit about broadcast. Patrick's comment in the opening slides, we continue to believe the broadcast infrastructure market will perform very well for us, and I hope to convey that in the next few slides. So first and foremost, why are we so confident in our ability to perform well in broadcast. The first is a really our tech. While in the streaming world, we -- I mentioned the product, the VOS or the VOS360 product, in the appliance kind of more traditional world, our product is named XOS. Not by surprise, this is actually the same exact product. Our engineers developed the software on the public cloud. And the majority of our staff develops that product as the core, single, unified software stack. We then have dedicated teams, which take that core stack and make it available to a specific environment, say, Azure, or, in this case, a commodity off-the-shelf server by HPE. So we have a team, which takes that core cloud stack and then places it on a specific 1RU server made by someone as HPE, turn off all the cloud-related components, make it available for 1 or 2 of our targeted applications such as transcoding or such as playout and then ship it as an appliance to our customers, a software appliance. This creates, for us, an exceptionally efficient R&D team, where we don't develop the same technology twice. We have one core tech, and then we simply adapt it to the different environments. It's not that different from a company such as Toyota, which uses the same engine across multiple cars; or Tesla, which may use the same platform and the same engine across all their different models. The fact that our appliances are all software gives us enormous flexibility to adapt, to innovate and also to respond very quickly. And it's important to note that our products are world leading. We typically have the world's #1 market share in the key functions of broadcast, playout, encoding and transcoding, and the world's biggest logos utilize us as we are known as the premier vendor with the highest quality, the highest reliability and the best performance. And that leads us to the second part of our broadcast business, which is the recurring revenue part, and that becomes related to our support contracts. Today, we have over $1.5 billion of broadcast infrastructure installed base under SLA contract. We have a 24/7 team of experts around the globe that support them, and we are very well-known as a premium product and also premium support. As mentioned earlier, the broadcast business for many of the companies around the world is where they generate the majority of their profits. They want to make sure that they continue to generate those profits to invest in their next-generation endeavors. So ensuring that the systems perform well, that everything is top notch, that they don't have outages is very critical for them. And therefore, we believe that we have a strong and solid base of service recurring revenue for years to come based on our broadcast business. On the competition side, we really see those logos that I mentioned earlier, the traditional broadcast infrastructure competitors. We have been competing against these companies for probably 15 years or more. But it's important to note that it's not exactly the same companies. Four or 5 years ago, we competed against much stronger companies with better capitalized, better sales reach and much stronger cloud in the market. Those were Ericsson, those were Cisco, those were Harris and Belden, much, much bigger companies. All those companies have divested their historic broadcast infrastructure divisions and sold them off to private equity hands. That transition, as a result, of course, of new management, new focus, new strategy, new investments, has caused a lack of focus from some of them. In addition, under private equity leadership, the focus is often more about financial metrics, cash flow, EBITDA, debt-to-EBITDA ratios, et cetera. It's not always about innovating, and it's not that easy to adapt and to put the huge investments into both the streaming as well as the broadcast market. As a result, what we are noticing in the recent quarters is we are taking a lot of market share from these weakened competitors. Most of our 5G wins are basically taking market share from incumbents, which are the logos on the slide here. When the 5G initiative wave came in, we were ready with our XOS product. It was -- had all the feature sets and the capabilities and was much more attractive proposition than just buying from the legacy appliance supplier. We continue to believe that we will perform very well in the broadcast market considering these dynamics. And with that, I'll hand it over to Sanjay for the financials for a long-term view on both the broadcast and the streaming market.

Thanks, Shahar. Corresponding to the industry and business transformations that Shahar just outlined, we also see transforming video business financial model. The baseline of this discussion is the year 2020, when our video segment reported $243 million in revenue, which included $212 million in broadcast revenue and $31 million in streaming revenue. Looking ahead to 2024, we are targeting video revenue of over $300 million. For the reasons Shahar explained, we expect broadcast revenue will remain relatively stable over $200 million in 2024. And as we continue to gain market share during the industry's transition. Looking at streaming revenue, taking advantage of both steady market growth and Harmonic's differentiated and expanded solution set, we are targeting a revenue CAGR of over 33%, leading to over $100 million in 2024. Importantly, of this $100 million, we think at least 2/3 will come from our SaaS platform, implying substantially stronger recurring revenue. Total recurring revenue for our video business comprised of revenue from both of our SaaS platform and appliance support contracts, is expected to be over $150 million by 2024, representing a CAGR of over 12%. To put it another way, we expect about half of our 2024 revenue will be recurring, a fundamental business model transformation that will provide substantially enhanced video business financial visibility, consistency and, ultimately, profitability. As streaming revenues continue to scale, with SaaS growth resulting in higher-margin revenues, for 2024, we are targeting a non-GAAP video business gross margin of over 60% compared to 54.5% we reported in 2020. Our video research and development, marketing and G&A costs are expected to remain relatively flat, thereby generating higher operating leverage. We expect video non-GAAP EBITDA margins to reach over 12% by 2024 compared to 3.5% we delivered in 2020. Looking beyond 2024, we expect continued gross margin and EBITDA margin expansion as streaming revenues continue to scale. In summary, we are targeting a compelling financial transformation to a higher-margin and more recurring-revenue-based model over the next 3 years, mirroring the market and technology transformations we've reviewed with you today. Our organization is committed and excited to deliver on these objectives. I'll now turn it back over to Patrick for closing remarks.

All right. Well, thank you very much, Shahar and Sanjay. So I'd like to just finish where we started, which is key takeaways from today's discussion and key takeaways regarding where our video business is headed. Hopefully, over the course of the last 45 minutes, we've demonstrated our conviction that the streaming infrastructure market is indeed a rapidly growing and attractive one. I think it's just common sense that it's -- we all see it as consumers. It's where the market is headed. And it's clear to us that it's a greater than $1 billion opportunity. I hope we've also convinced you that we have high confidence, and we're seeing great customer traction around our SaaS streaming platform. We've -- as Shahar outlined, we've made tremendous investment in it. It truly is leading in a number of technological dimensions. Putting those 2 facts together, we're quite confident in our baseline model here of greater than 30% annual growth rate of our streaming business and exceeding $100 million by 2024. As Sanjay just emphasized, putting this together with the continuing service, recurring service element of our what we believe will be steady broadcast business, we see over 50% of the business unit revenue being recurring by 2024. And again, as Shahar outlined in some more detail, underlying this continuing investment in growth around streaming is a very solid -- increasingly solid foundation associated with the traditional broadcast market and the very strong brand and market position and customer relationships we have there. Putting it all together, we think this business is in a very strong position, stronger than it has been for some time, both in terms of the momentum and in terms of the near-term opportunity ahead. And as Sanjay just said, our entire company is committed to and excited about growing this part of our business. Okay. With that summary, we would now like to open it up to take a few questions.

[Operator Instructions] Our first question comes from the line of Simon Leopold.

I got just a couple of things I wanted to check on. One of which is, in your opening comments, Patrick, you mentioned the 5G reclamation activity, which you've seen some initial revenue from. And I understand there are some other opportunities that come out to 2023. How do you weave that particular opportunity into the outlook you've offered for broadcast and streaming? And then I've got a follow-up.

Okay. Thank you, Simon, for the question. You know what, let's take advantage of Shahar being here with us. Shahar, if you wouldn't mind, I suggest that you take that for a little bit additional color that Simon is looking for.

Sure. So you're correct that basically the 5G initiative has prided some revenue for us already in 2020. We expect to continue to win more projects in 2021 and see that going into 2022. We see that as a specific initiative in the U.S., the 5G, but it will span over 3 years of impact into our revenue. But the bigger trend is that the governments all across the globe are making infrastructure changes, taking away various spectrums, frequencies and so on and essentially pulling forward refresh cycle. So the way I look at the 5G initiative is more about as a tailwind to that whole massive refresh cycle of converting legacy infrastructure broadcast equipment to more efficient equipment. And maybe it's driven by companies driving for higher profits, but in this case, the driver is the government looking a way to take away some infrastructure. So it's just a specific case study of a much, much bigger trend.

Great. And then just as a follow-up, I wanted to see if you could help us maybe bridge the pandemic effect. If I think back your -- you exited 2019, your fourth quarter '19 revenues for the video segment were roughly $79 million, and you troughed in June of 2020 during the height of pandemic at a quarter with about $47.5 million. And so, to some degree, I'm presuming we've observed some recovery as well as the benefit from satellite reclamation business. But maybe if you could just level set us to where are we in the recovery from what you experienced during the part of the pandemic during 2020?

Well, I think I would say that there was definitely a decline in that second quarter or the end of the first quarter during the pandemic, specifically related to lockdowns and deployment of on-premise equipment. I would say that post-pandemic, what I see is less impact. Definitely, the business in appliances and in broadcast infrastructure has recovered, and we see that as relatively steady. What I do see as the bigger impact is the acceleration we saw in our SaaS business and specifically in streaming during -- because of the pandemic. We're definitely seeing a much larger percentage of the customer base around the globe eager to take advantage of cloud, eager to deploy streaming services, which the pandemic really hit home as the future of the industry. And that's giving us a lot of the confidence that we're projecting into the numbers of our streaming business. So it's true that we did see a little bit of up and down during the pandemic. We think that's behind us. We think the trends are very positive right now on the broadcast infrastructure market and much more stable and based on a lot of the refresh cycles we talked about. But the real benefit of the pandemic was the acceleration we see on the streaming and the cloud and the SaaS side.

Simon, I would actually like to add -- I would like to add that when we exited last year, 2020, Q4 was a record backlog and record bookings quarter for us. Yes, we saw to your -- to Shahar's earlier comment during the middle of 2020, we did experience the slowdown due to COVID. But Q4 kind of showed us a recovery. Q4 backlog, bookings and pipeline, they were all very solid numbers, and in fact, broadcast and streaming both. So while we don't break out that data, but we have seen the trends significantly growing, especially since Q3 last year, we saw the trends continue in Q4. Q1 also this year was a solid bookings and exit backlog quarter. It's still near record backlog we carry. So I think the trends are indicating that streaming and broadcast, both are doing really well.

[Operator Instructions] Your next question comes from the line of George Notter.

I guess, I wanted to -- I think you mentioned that SaaS invoicing was up 40% in 2020. You said that channels were up -- number of channels supported in that SaaS model, I think, were up 345%. So I guess I'm wondering if there's some deflationary sort of impact going on here in terms of traffic growth relative to revenue recognition or invoicing? Can you walk us through that dynamic on pricing versus traffic and growth?

Okay. I'll take that in a couple of ways. First, it's important to note that even if we increase the channels from the beginning of the year to the end of the year by a certain percentage, it didn't mean that they were up the entire year. So invoicing spans the entire year, and you could have added 100 channels in December, but you've only invoiced them for 1 month. So I think, just on that perspective, you can't really find an exact formula between those 2 metrics. The second is, the streaming TAM, as we showed it, is composed of the media processing, the linear channel side as well as the traffic, as well as the targeted ads. And as you can see, in the TAM that we're projecting, the media processing is smaller than the other 2 combined. And so even if you have channel growth, what you really see is that those are kind of the launches of new systems. And then, over time, you're going to see much higher traffic. And then, over time, after that, you're going to see a much more rapid adoption of ads. So I would expect that to have higher channel growth capacity and then, later on, to see the invoices trickle down towards the traffic and later on into the advertising.

I'd just like to add -- I'd just like to add, Shahar, that regarding usage. Usage, although, George, is a very good metric for revenue recognition. But some of our customers' usage in the early days do not meet the minimum monthly commits. But as these systems grow and then usage grows as time passes, they then meet and exceed the monthly commits, at which point, usage and revenue get more aligned. So I'll say, while, it's a good metric, but over a longer period of time, it makes more sense than initial periods of the customer.

Got it. Okay. That's great. And then maybe just another one. I think, historically, one of the strengths of the company on transcoding was really in real time. And as more of the world goes to streaming models, it feels like more of the content is pre-recorded, preprogrammed, the need for real-time encoding and processing is not quite the same. Is that a dynamic here that you guys see it play in your business, or the other trends sort of mask that?

It's not. I mean, I agree with you that the original -- the original streaming platforms out there were all video-on-demand. But what we're seeing is a lot of interest in linear and live events. So sporting events, live news events, concerts and other components, basically, are driving a lot of live transcoding and also just the traditional kind of linear channels being time shifted and captured. So we don't see a decline in that. What we do see is basically kind of a growth in both VOD as well as linear. You may not have exactly the same number of linear channels when all is said and done. But overall, you have a major boost just on the live events. Historically, because the programmer only had 1 or 2 or a handful of channels, there was only a small amount of sporting and live events that they were able to introduce into their schedules. But in the world of streaming, there's almost unlimited amount. And so you do have a lot more of kind of live event transcoding just simply because you can create a lot more content and make it available on the streaming platforms than whatever was available in the world of broadcast.

And, Shahar, if I could add, I think it's also important to note that the traffic component and the ad insertion component, from a Harmonic monetization point of view, that is kind of indifferent to live and on-demand, right? And we believe that our technology put differently is equally differentiated for both on-demand as well as live, which, I think, to George's point, is perhaps a little bit different than our historic core transcoding strength, which, as George correctly notes, where we've been known for a long time as we're the world leader in live transcoding.

Our next question comes come from the line of Tim Savageaux.

Thanks for the event as well, and the targets and disclosure, all very helpful. I mean, I wanted to ask about one part of that, which is the kind of streaming target where you're looking for $100 million of revenue out of $1.1 billion TAM. So arguably fairly modest market share especially given the kind of anecdotal descriptions around competitive success in the marketplace and up some 6% to 9%, but throughout the forecast period, and also relative to a 30% share in broadcast, right? So how do we explain that relatively single-digit share target in context. And does that suggest that what really needs to happen here is further consolidation among this group of streaming and broadcast competitors, some of which overlap? Or is there some other dynamic at play here?

I think there's multiple dynamics at play that are going on here. First, it's important to note that when we started out, we were mainly focused on the media processing side of the streaming. And recently, through the cloud platform, we're now targeting the bigger market opportunities of the traffic and the advertisement. And so that is essentially one of the major engines of fueling us with this rapid growth. I do think we will reach around 10% market share, and that's going to be, I would say, the first step, but longer term, our goal, past 2024, is to reach well above that and to get back to what we are historically known for, for the, I don't know, 20%-plus market share. It's also important to note that a lot of the original streaming platforms were kind of done with do-it-yourself and preliminary systems. And now as the systems are becoming more mature, more complex, there's a lot of interest to move to much higher-end platforms. And we're seeing a lot of the logos that we showed on that chart, are actually customers that were already streaming, and they simply moved to a much higher-end platform as the complexity grew. So in the short term, I think we will reach that 10%. Longer term, I expect us to continue to grow that consistently.

Okay. Just to follow up there. I mean, whether it's currently or in your target, I mean -- well I assume the DIY portion, you wouldn't include in your TAM, right, out in 2024. But maybe that's wrong. Maybe that's a SAM number. So if that -- if your biggest competitor now and in the future remains that would be interesting to know. And both now and currently in the '24 time period, who would you expect to have market leadership and at share level would they be if it's not our money?

I would expect the major players -- there are several players in the market. But the major infrastructure player that we will compete with mostly and specifically on AWS clouds will be the AWS Media Services. That will be our #1 competitor. We will have secondary competitors out there. We will have a lot of regional players that build their own solutions, some second-tier solutions. But as a pure infrastructure player, when we compete on AWS, it will be AWS Media Services. They will probably have also a material market share component. In the Azure and GCP systems, which are growing, obviously, they're not going to be relevant.

Our next question comes from the line of [ Ryan Kunz ].

Just following up on that last theme here on the competitive landscape on streaming. Do you expect new entrants in this space? I know there's a lot of these kind of video platform companies, big guys like Zoom, they've been talking about becoming a video platform company. Do you see them at all starting to encroach on this at all? Or you feel like the barriers to entry are so high that this is just a specialized segment?

We think there's a lot of players in the market, but not in the professional media space. So I if you're doing education services or maybe corporate events or those types, you may have more competitors available, but to really do a full-blown sports streaming system of high scale or Tier 1 streaming platform, we think there's not that many competitors that can do that type of scale very well and so reliably. And we are one of the bigger ones. So I would say, in the long tail, you may have a little bit more competition, but definitely not in the top end of the market.

And besides AWS, do you expect any of the other big in-house players to bring their products to market as a commercial entity?

We don't see that trend. We don't see that trend. I do expect some of the traditional kind of legacy broadcast infrastructure to structure players, maybe one of them to break through and eventually launch a compelling offering, but that's not the case today. But I think in the high end of the market, it's going to be Harmonic and AWS for some time.

There are no further questions in the queue at this time.

Okay. Well, good. We're doing good just at the top of the hour. We'll wrap it up here. We appreciate your participation today. Our hope is that we gave you extra level of insight into our business, what's driving it, our technological differentiation. And our view of the opportunities going forward. But this is an evolving space. It's a dynamic space. I think key to remember is that we have a fantastic brand, we have fantastic technology, and we're going to continue to invest and to take advantage of the opportunity and look for additional incremental growth opportunities. As a baseline here, as we presented today, though, we think that our video business is in a great position for a new phase of growth and transformation. I think the last thing I would say is please also plan to join us a week from today, where we will do a similar deeper dive on our Cable Access and broadband businesses and give you similar kinds of insights, both technological market evolution as well as multiyear financial outlook. Until then, thank you all. Stay well and look forward to speaking with you next time. Have a good day.

Thank you.