Joining me is Alex Paugam, Senior Director Business Development and Global Strategic Accounts at Harmonic, who’s been at the forefront of this transformation. Welcome to the show. Hello, Naytram. Thanks for having me. Really happy to be to be here on this podcast. Awesome. Let’s start with the big picture. The broadcast industry has seen major shifts in how content is monetized. From your perspective, what are the most exciting changes happening right now in monetization strategies? Thanks. Great question. So I think from a high level perspective, we’re seeing really more and more content owners and broadcasters bringing ad supported models to their OTT platform. The first reason is to generate more revenue, but also customer acquisition. And I think now the trends are more not really about how do I, personalize or insert targeted ads, but how do I make it more reliable? How do I cope with huge audiences as more and more people are moving to streaming? And so you will need sometimes to to insert ads at the same time for tens of millions of concurrent viewers. So how do I cope with huge audiences? And and and really, how do I use tech to enable new monetization, better targeting, and and higher engagement? Yeah. You mentioned tech. What are the dynamic ad insertion monetization current technologies being used and having a positive impact right now for customers? So if we did if we now deep dive a little bit into the tech, and and from our perspective at Harmonic, the way we we address this is using targeted advertising. So manifest manipulation technology is the ability to replace ad and replace content and make it targeted to the end user. So since we are pulling some feeds and it’s a unicast relationship, you can really personalize, that unique session and replace the ad and make it targeted per end user. You could also use that same technology manifest manipulation to replace the content. And so it’s it’s standard, server-side ad insertion (SSAI). But now it’s moving to be able to use that standard technology but on more advanced workflow. So coping with, low latency delivery, UHD, and be able to insert those ads potentially to tens of millions of concurrent viewers and have targeted ads, different one for you, different one for me, for for that huge audiences. And and and now it’s moving to how do I target it more per end user, per geography, per demographics, and also per on what’s in the content. So it’s leveraging that manifest manipulation for really advanced workflow. That’s that’s currently the the the the technology that is being used from a video pipeline perspective. Okay. And what are the innovations and new capabilities that you see coming? So we’ve been I’ll talk about what we have been doing, and that also relates to the the the broader market. We’ve been, investing a lot into those targeted ads technology to enable advertisers to, insert the right their ads at the right moment on the right, game or content and to enable publisher to have better targeting, so higher revenue, but also create new inventory that could be a little bit less disruptive than moving to complete ad breaks. So one of the innovation we’ve been investing in is called in-stream advertising, is having the ability downstream at the SSAI level to insert ads that are not a full ad break, but, new ad format. So it could be a squeeze back, so I have the ad around the video. It could be a double-box, so I I squeeze back a little bit the video and I have an ad playing next to it, or it could be an L shape. And, that in-stream ad technology is really helping the advertiser put their ads exactly at the moment they want and also the publisher to create new inventory that is premium because you and I are still watching the soccer game when we watch that ad. Yep. And it’s targeted. So it’s not on the production level. I create that graphic and ad for everyone. It’s done at the SSAI level. So it’s targeted. So it’s higher CPM. It’s less disruptive, and you’re not increasing the ad fatigue, but providing more monetization capabilities for the broadcasters, the publishers. That’s really one of the strong innovation we’re working on. And then there are some some some other, workflows where we are embedding AI to provide better targeting. And so that’s mostly around, contextual metadata creation so that we are leveraging our own AI tech and also, cloud provider AI tech to analyze the video, analyze the content, and be able to create metadata, so both on live and VOD content, create metadata that describes what the content is about, what the sentiment of the content is, and also give some information about what’s what’s displayed. Is it brand safe? So so we have the ability to better target at the right moment because we understand with and enrichment the data what is happening, Better target at the right content because we understand better the content. And also, show the right ad for the right content. Meaning, if it’s a news channel and you’re talking about a plane crash, you don’t want to display an ad about an an airline just afterwards. So it’s brand safe as well. So using AI to create contextual metadata that enable triggering of the right ad at the right moment for the right brand and also, enrich with markers where there aren’t. Because we do see a lot of customers that receive feeds without markers. So we can understand when there is a switch to an ad, create that marker, and then later on replace those ads and make them targeted. Are your customers seeing a better better value for these types of, services where you’re no longer interrupting the stream, you’re doing a squeeze back? You know, I personally when I watch a football game, I don’t feel like, you know, I’m watching an ad anymore because I’m still seeing some behind the scenes. So are customers, you know, inserting or doing using leveraging this for more, placement of ads in soccer games instead of waiting for halftime? Yes. That’s that’s exactly the idea. There are some sports where you will spend forty five minutes on a soccer game without any ad, but so many moments where the guys are rolling on the floor and not much is happening. Yeah. So, that creates opportunity to insert to create new inventory, to insert new ads, but the viewer is still engaged. So you’re still listening to the audio. You’re still watching the content, and you are seeing an ad, but at a moment that could be less interesting. So, it’s not we we’ve we’ve run and we’ve seen, reports and analysis. It’s not increasing the ad fatigue from a viewer perspective, but it’s creating new inventory, so new ad revenue. And it’s premium because your viewer is engaged, so it’s it’s higher CPM. So it’s it’s really premium in a high value inventory. And for the advertisers, it’s the ability to target at the right moment and have their ads displayed to a lot of people that are really engaged and didn’t go to the restrooms. Yeah. Nice. What so yeah. What, what are the new business models that are enabled by these new technologies? So, if if we focus on this, we really see, an increased CPM value for those in-stream ads and new formats. In some areas, it’s three to five times. But it’s it sticks to the standard SSAI and and and ad ecosystem and digital ad ecosystem. So it creates new inventory, new revenue, but it’s still the same business model. It’s CPM based, and it’s kind of a rev share overall. So it’s standard, but it creates new, monetization capability. So it’s really, a win win from an advertiser, publisher, and and technology enabler perspective. Oh, cool. And, you know, what are the considerations to take into account, when it comes to pitfalls to avoid, you know, in in order to maximize the return on investment for these types of, new technologies? I I see mostly three. The first one is is the ability to cope with growing audiences. As people are we we do work with five thousand customers at Harmonic, mostly broadcasters, operator, more and more, leagues and federations. And we see a lot of the projects that we were doing on prem for broadcast workflows. We see those project turning into streaming because there is a lot, and 2022 was a tipping point in terms of number of subscribers on the broadcast and the streaming side. Streaming is growing, so more and more viewers on the streaming side. So it means that you also need to cope with growing audiences. And so when you do targeted ads, you need to insert ads at the same time and to replace from millions, if not tens of millions of people. So one of the thing to really, take close care of is having a resilient architecture that can cope with those audiences. Because that piece, the ad stitching that is inserting the ads, this is in the critical path of the video. Yes. If that fails, everything fails. Yeah. And so if you are inserting tens of what what we are doing currently, tens of millions of concurrent ads for NBA in the US, if that fails, ten millions of people get the black screen. And at the moment where you are inserting the ads, so exactly at the moment where you don’t want to have a black screen for your customer who is the publisher. So having an architecture that is resilient is is is paramount. And we have built that, ten years ago to be cloud native, to be geo redundant. So we are always deploying our SSAI into two availability zone in the cloud. If one fails, the other take over, and that’s how we are able to provide some high uptime SLA of 99.995%, which is the the highest in the market. And, when you have failure, we’re able to switch very, very quickly to the other side and avoid having a black screen on on the end user. So that’s that’s one of the big pitfall. With people moving to streaming, you need robust architecture. And, and another one is having also and that’s more kind of an advice is when you choose your partner, also look at their integrations with other bricks of of the whole video puzzle. And here we’re talking about monetization, so the whole monetization puzzle. So having the ability to be integrated with different ad decision server, different DSPs, different SSPs is is key because you might change one or the other at some point, and you don’t want to to have to reinvent the wheel and do a lot of integration. So having a a solution that is integrated with the vast majority of big ADS is really, really important for flexibility, for resiliency. Okay. And then, you know, what what advice would you give for, you know, new customers starting out on this journey? Like, you know, clearly, Harmonic has been doing this for a number of years, and you’re helping these customers to migrate to these new workflows. What kind of advices would you give a new customer coming into this world? So it’s it’s kind of summarizing what you’ve just said is, look for pre integration, look for robustness because you don’t want it to fail. And, and look for innovations that that are able to generate you new revenues, provide new capabilities to better target, to better understand when to place the ad and what which ad to insert. Because in in the end, it’s a better viewing experience. It’s better and and and a better viewing experience is longer viewing sessions, and better targeting is higher CPM, so higher monetization. Okay. And you mentioned AI. Do you personally use any of these AI services? Have you ever tried ChatGPT or any of those? So I I’ll answer for Harmonic, and then I’ll answer answer for me. At Harmonic, we really have three ways to work on the AI side. We use it in our R&D to optimize and and and, and be more productive. And we’ve seen, since we started using it in 2022, an increase in 20% more user story from our R&D. So we are just doing more. We are developing more. We are more productive. The other part is we are embedding AI into our code to provide optimization, automation. So for example, we can, optimize the video so that it it takes less footprint, less bandwidth from a storage or CDN perspective. So it’s, in the end, cost reduction for our customers. And and then we have a lot of other ways to use AI into our code to create better targeting. And the last part is really bringing the best of breed of partners, AI partners, that are very close to the video pipeline and close to our tech. We build integration and we bring them part of our solutions of kind of a resell. Yeah. We act as the front and they are under the hood so that we provide to our customers development, but also key partners that are enriching and providing new capabilities. And then on my side Yeah. I use ChatGPT forty times a day. Forty times a day? Yeah. Always. Okay. So what’s, what was your last query to ChatGPT? That’s a good question. Let let let me check. Oh, it’s a it’s a live one. I’ll I’ll tell you. Oh, yeah. Simple. This morning, I I made a post about yesterday, went to the Devoncroft Summit, and we also had our, usual partner reseller, event. And I took a few pictures. And so this morning, I said, please write me a post as if I could have written it in a less formal way and, for LinkedIn, of course. And, the two sentences were just perfect. So it helps me build, keep on top of my, LinkedIn influencer. Oh, I missed it. I I’ll have to go check that post out. Oh, awesome. Well, thank you thank you for joining us on the podcast today, Alex. We’ll see you later. We we look forward to hearing more from you soon. Awesome. Thanks. Thanks.