If Google sent you zero traffic tomorrow morning, would your P&L still be breathing by lunchtime?
Agility requires re-engineering your revenue model at the speed of AI.
Today we’re going to talk about how AI chatbots are siphoning off audience attention—and what forward-thinking publishers can do about it instead of panic-refreshing their analytics dashboards.
To help me discuss this topic, I’d like to welcome Sean King, Chief Revenue Officer and GM of Media & Entertainment at Veritone.
About Sean King
Sean King is the Chief Revenue Officer & GM, Commercial at Veritone, leading all sales and marketing activities for the organization in addition to overseeing Veritone’s Commercial division, including SaaS technologies and managed services. With a keen ability to drive growth and operational excellence, Sean has played a pivotal role in scaling Veritone’s AI-driven solutions. His leadership was instrumental in establishing Veritone Licensing as North America’s premier AI-powered content licensing firm and in guiding Veritone One to become a top AI-driven audio and influencer advertising agency before its successful divestiture in 2024. With over 20 years of experience in sales management, technology operations and strategic partnerships, Sean is passionate about optimizing business operations, unlocking new opportunities and applying AI and synthetic media to drive innovation.
Sean King on LinkedIn: https://www.linkedin.com/in/ryansteelberg/
Resources
Veritone: https://www.veritone.com
The Agile Brand podcast is brought to you by TEKsystems. Learn more here: https://www.teksystems.com/versionnextnow
Don’t Miss MAICON 2025, October 14-16 in Cleveland – the event bringing together the brights minds and leading voices in AI. Use Code AGILE150 for $150 off registration. Go here to register: https://bit.ly/agile150
Connect with Greg on LinkedIn: https://www.linkedin.com/in/gregkihlstrom
Don’t miss a thing: get the latest episodes, sign up for our newsletter and more: https://www.theagilebrand.show
Check out The Agile Brand Guide website with articles, insights, and Martechipedia, the wiki for marketing technology: https://www.agilebrandguide.com
The Agile Brand is produced by Missing Link—a Latina-owned strategy-driven, creatively fueled production co-op. From ideation to creation, they craft human connections through intelligent, engaging and informative content. https://www.missinglink.company
Transcript
Greg Kihlstrom (00:00)
If Google sent you zero traffic tomorrow morning, would your P &L still be breathing by lunchtime? Agility requires re-engineering your revenue model at the speed of AI. Today we’re going to talk about how AI chatbots are siphoning off audience attention and what forward-thinking publishers can do about it instead of panic-refreshing their analytics dashboards. To help me discuss this topic, I’d like to welcome Sean King, Chief Revenue Officer and GM of Media and Entertainment at Veritone. Sean, welcome to the show.
Sean King (00:29)
Thanks so much for having me, Greg Pleasure to be here.
Greg Kihlstrom (00:32)
Yeah, looking forward to talking about this with you. Definitely a timely topic here. Lots of lots of people thinking about this before we dive in though. Why don’t you start by giving a little background on yourself and your role at Veritone?
Sean King (00:43)
Excellent. ⁓ Well, about myself, I’ve been in the media and entertainment space as well as the advertising side for about the better part of the last 20 years and have added various roles across operations, sales, go-to-market and strategic partnerships. But at the end of the day, I looked and love looking at a lot of managed service-based businesses, which are a lot in media and entertainment and advertising, and looking at ways in which you can better operationalize those businesses bringing new tools, writing new workflows that can help take a mundane people process and make it much more streamlined and much more effective. So when I had the chance to come work at Veritone for the better part of coming up on nine years now, wow, mean, talk about a space where AI can provide, can really be a force multiplier for this entire industry at large.
Greg Kihlstrom (01:34)
Yeah, love it. Well, yeah, let’s let’s dive in here then. And I want to start with I think definitely something that’s top of mind for anyone who, you know, search traffic is a significant, you know, whether it’s revenue driver or whatever, whatever the case may be. Gardner is saying that AI search experiences could cut traditional search traffic by 25 percent by 2026. So significant. Tolbert just reported a 49 percent surge in retrieval bot visits this year alone. How are those numbers reshaping the urgency for publishers to rethink? What is their core business model?
Sean King (02:14)
I mean, those are really dramatic numbers. When you take a step back and look at those ones. And frankly, I mean, it’s a wake up call to a lot of these groups. And what you’re really seeing is you’re seeing almost the rebuild of the internet taking place with a lot of these large language models and other search. I mean, more and more individuals are going to LLMs and whether they’re open AI, Gemini, cloud, whatever it may be, they’re looking at those as the way that
They’re becoming their trusted search engine. It’s there. But the reality of the situation is organic search has already been kind of declining for major outlets, major publishers, which means they have to kind of pivot from the traffic they depended on coming from the search into new ways. So really this means that those AI platforms, their content archives, you know, that they’re using, you know, they have to be turned into usable, valuable training data sets.
So that way they can start to be found by these different groups and can continue to allow them to do that because everyone’s consumption for content, everyone’s consumption, whether it be print or film or video, it’s still there. It’s just the ways they’re going about finding that content is shifting. And so it puts an added pressure on these groups to make sure they’re that more agile to make sure that they can meet the demands of where their consumers are going to.
Greg Kihlstrom (03:36)
Yeah, yeah. And I mean, in a sense, this is we’ve been designing websites for bots, you know, for I mean, there’s been bots ever since there’s been an Internet almost, you know. So like in a sense, this is we call it something different, but it’s it’s it’s a little bit the same. But but in other ways, it’s it’s certainly different. And so, you know, from from a consumer perspective, you know, people are still consuming content. I mean, they’re they’re
Certainly according to the stats, they’re consuming it through LLMs more than they certainly were before, but they’re still consuming content directly from publishers as well. How does a publisher make sure that they’re reaching real people in addition to those bots?
Sean King (04:19)
Well, I mean, like you said before, mean, people are consuming more content than ever. But these are making sure that you’re making your assets more accessible to being able to be found by, we’ll call these bot like AI intermediaries. Rather than visiting these ⁓ publishers kind of directly, they have to find new ways in which in gang, not just with their standard audience, I’ll consider the loyal, the humans of it but making sure that those assets are data ready so that these AI intermediaries, these AI tools, can make sure that that same content is discoverable in these new ways.
Greg Kihlstrom (04:56)
Yeah, yeah. So Veritone put out an ebook recently that talks about viewing content as data as part of a strategic mindset shift. Can you talk a little bit more about this? know, what does it mean and how does this connect to the growing role of AI bots as we’ve discussed so far?
Sean King (05:15)
This is one of my more favorite topics to think about because it is a mind shift. If you’re an engineer, you’re in product, every day you’re up there and you’re writing code, at the end of the day, what do you do with that code? You put it into GitHub, you put it there, you put it into a safe repository that’s there, so it’s accessible, so it’s safe. In the media entertainment, you don’t think of that as data. You think of it as film, ⁓ as videos, as entertainment. You have to kind of think about that the same way in each thing.
I’m creating every day, whether that be this podcast, whether that be a short, a reel, a video, a sporting event. All of that film that you’re creating, all of that, all of that is data. So what do you need as an organization? You need to make sure that you have a partner that can ingest all of that content, that can append the necessary data to that on kind of a frame by frame rate. Who’s on the screen? What is it talking about? What objects may be there?
You know, what are those ways in which that I can make sure I understand the most about this content and where is it in a really effective data lake, so to speak, that’s kind of time correlated that I can make sure that all of that content is accessible. It’s discoverable. And you got to make that a data centric approach. So, you know, you always think about it as a creative approach, a conversational approach. You have to keep those same you know, creative ways of going about in which you’re creating content. But through that workflow, you need to think about, okay, at what process am I turning this into kind of AI ready data assets? So that way later I can make sure I’m getting the most out of my content and getting, making sure I have the most understanding about it. And that’s kind of one of the things that Veritone does a lot with our partners is we kind of orchestrate that transformation of content, of film, of
audio, a video, a photos, a paper, anything that you want to, and kind of orchestrate that transformation into AI ready assets that can be accessible through applications or to create custom workflows to help solve for what you’re looking for or find the operational efficiencies you’re seeking.
Greg Kihlstrom (07:26)
Yeah, yeah. So then so that so then that’s what Veritone maybe talk a little bit more about the role of Veritone’s platform then and also just your experience so far with that mindset shift, you know how you know how customers are adapting to that.
Sean King (07:44)
Great, so look at Veritone at large is we’re a software company and we have a platform that’s called, we call AIware. And every single day across our customers in both the public sector and the media and entertainment or commercial sector, we’re ingesting about 150,000 hours of content a day. So think in 2024, we did something short of a little more than 58 million hours of content. Wow. I mean, it’s a lot. And as that content is coming in, we’re appending
all the necessary metadata to that content, depending on the use case for that customer. And once that’s there, you can access AIware through either our applications. So when you think about it, think of AIware as your operating system, your applications, as how you interact it. I always like to give the example of like Microsoft Office. You you have to have your IBM operating system or Microsoft operating system, but you may only use Excel or Powerpoint. So we have in the media entertainment, a lot of our groups use Digital Media Hub, which allows you to interact with your content. It’s an AI-enriched archive management tool that’s there. So as that content is coming in, we’re appending all the data, we’re making it searchable, making it discoverable. They may want to power workflows. They may want to find all their content, make it accessible for secondary and tertiary use cases like content licensing or data monetization of their assets which is a lot more in media entertainment groups are doing more and more these days. But you have to have a common data standard for how these things are coming in, how things are getting labeled, how things are getting appended. And it may look a little different from our media entertainment customers or someone in the public sector like a local police department. For us though, a camera is a camera. So it doesn’t matter if it’s a body worn camera, a dash camera, or a camera looking at Augusta National golf course or the US Open doesn’t matter to us. Camera’s a camera, but the workflows are very different and how you’re interacting and what you’re wanting to do with that content is quite different. And we just kind of help all of our customers, both in public and commercial sectors, get the maximum yield that they can get out of their assets.
Greg Kihlstrom (09:56)
So we’ve talked about making the content searchable, discoverable, all of that. Now let’s talk a little bit about protecting as well as monetizing that content. So, you know, we’re seeing a flurry of licensing deals between AI companies and some major outlets, smaller publishers though, you know, maybe, maybe worrying they’re going to be left out or sidelined. You know, what, what negotiating leverage to mid tier or niche publishers still have and how can technology level that playing field?
Sean King (10:27)
Great question in those. Look, mid-tier publishers hold significant value and leverage because more of their content tends to be specialized and is something that is gonna be way more domain specific than a lot of the major publications. And there is a unique vertical training aspect where these could have significant value for not just commercial, but for a potential AI use cases that are there a lot of those ones come into it’s really what are the needs of the data scientists or what are the needs just like they are if you’re licensing for a commercial. You know, I want to get this great footage for a documentary I’m doing or something that’s there. I mean, it’s the same type of kind of domain specific needs that will come in. So what I always suggest is rather than focusing on kind of individual pieces or individual specific types of content, having the publishers kind of position themselves kind of with
some more ongoing strategic data partners. I may be one with 5,000 hours of very specific content. Well, great, if I can work with a partner that I know has many other different partners in that space, there’s opportunities in what I’ll consider more of the tonnage. Well, I may be a mid-tier publisher, but I have very domain-specific information. Well, there’s 15 others that are like me.
Well, when you put the 15 others, like you actually have something that’s quite meaningful and valuable and it helps bring more opportunity to all those groups.
Greg Kihlstrom (12:01)
Yeah, yeah. So then, you know, is part of the strategy strength and numbers then, you know, is that that as you mentioned, alliances of publishers banding together, is that, you know, is there any downside to that? Or, you know, is that kind of the path?
Sean King (12:17)
I mean, it’s not necessarily a downsize. It’s just, it’s making sure that there’s the right type of opportunity. I mean, most of the groups in these sides, everything is still being dominated by the hyperscalers today, which obviously are looking for lots and lots, but you know, we haven’t necessarily seen the next wave of all coal consider like these very specific, LLMs domain specific LLMs.
So it’s not saying that there wouldn’t be, and frankly, we’ve seen this movie before when YouTube came out. And then how did that turn into these smaller groups, what’s happening to it? And you started to see all the different creations of these NCNs and everything else that started to take place. Just like what took place on the web before and where all the traffic was started to go to the Netscapes and the AOLs and the YAHUs. Well then, we see, I mean, we’re gonna see a similar evolution as this place it’s course at.
Greg Kihlstrom (13:10)
Yeah, yeah, makes sense. So last topic I want to talk about is something that know Veritone has written quite a bit about as well, is responsible AI. let’s talk, follow that down to some future web talk as well. Veritone’s AI for good principles, transparency, trusts, and security, and compliance, and empowerment aim to keep AI on the straight and narrow.
which, you know, good, definitely on the right, in the right way. How do those guardrails translate into, you know, day to day decisions about using and licensing publisher content inside these generative models, LLMs, so on and so forth?
Sean King (13:51)
Well, I always like to say, mean, it’s, you know, I think we’ve all heard this, just because you can doesn’t mean you should. Right. Right. And I think that’s just a general purpose anyone should think about and how they’re using AI. You know, with Baratone specifically, and I got to give credit to our founders, you know, Ryan Steeleberg and Chad Steeleberg that kind of went, you know, back in 2014 with this, you know, the vision and hypothesis that in order for this to be truly safe and scalable, you almost have to democratize AI.
And that’s kind of what the vision was for AIware, which was, we have over 1100 AI models of which, that’s almost 58 million or so hours, I quoted, think we used 864 throughout last year. And it’s not because one’s better than the others or it’s to in the use cases is that in order to make sure you’re solving for the right type of outcomes, you cannot be beholden to singular models or singular usage. You have to make sure that you have the accessibility to as many as possible that really becomes the backbone for kind of mapping out and creating the right data points that serves as the basis for kind of, you know, everything that goes forward. But really when it comes into like ethical usage of those, I mean, there are so many different things that we see. And unfortunately in the media, we see all the nefarious use cases of it, someone took someone’s deep fake or whether it be their voice or their image, but what we’re not seeing a lot of it and the things that we are most usually proud of are how AI can be used to protect IP in those ones and advanced safeguards, client controlled usage of their rights, how that goes into licensing optimization tools and using against unauthorized defense and different things across that. We’ve done some things over the last couple of years, like we did the vault with the CAA, one of the largest agencies and they created the CAA vault, which kind of as I similar to think of content as data know, CAA had the foresight to think of, well, we got to think of people as data. When we’re creating these digital twins of individuals, you know, it’s not the outcome is it, it’s the training data that’s the value. It’s the, it’s the scans of their voice, of their faces, of their bodies, everything that’s being used to do that. You’re taking these, you know, human individuals in the, in the, in the real world and you’re putting them in these digitally immersive environments they felt that the same principles that we expect in real life needed to actually go into this digital universe, which was really awesome and something we’re proud of working on it with that. Because it’s again, it comes back to those, your name, your image, your likeness of individuals and all the assets that you own, those same principles of ownership need to make sure that they’re protected in these digital use cases. And that’s why you kind of have to have that radical transparency over all of your data over all your assets, which kind of goes back to me referring to, you have to start thinking about your media, your content as data. Because if you’re not thinking about this way, how are you going to be able to track, govern, and make sure anything is being safely deployed in these environments? I mean, it’s a core principle to being AI first. Is that data?
Greg Kihlstrom (17:08)
Yeah, yeah. Well, and I would also imagine, you know, another factor here is that regulators around the world are hinting at, you know, fair compensation mandates for content owners. Again, you have to be able to quantify that in a way before you can compensate fairly. Right. You know, if you, you know, if you were talking with lawmakers right now, you know, what would your advice or what would you propose to balance kind of innovation with publisher sustainability.
Sean King (17:38)
I mean, I think it’s a combination of both legal and ethical considerations that has to be paramount as these AI models are developed. We just need to make sure that the days of scraping the internet are over. And frankly, it’s already been done and that’s been done. But as we’re going into these next frontier, as we’re going behind these gated walls of where this other date and other goes, that everyone needs to look at it from the legal and ethical consideration standpoint.
Greg Kihlstrom (18:07)
Yeah. And so, you know, from the publishers standpoint, then, you know, given that there is uncertainty, regulations are, you know, I’m sure there’s some in the works. There’s always there’s always some uncertainty there. What advice do you have for publishers that need to make some potentially long term decisions that may have implications for whatever those regulations ⁓ end up being down the road?
Sean King (18:33)
Well, let’s face it, the regulation will evolve as things happen to it. So I think right now it’s critical to establish some flexible systems and frameworks for publishers. And I think most importantly is clear visibility on how, where, and by whom their content is being used. Is it just for R &D purposes? Are they looking for display rights? You have to have these flexible frameworks.
because I think with those, it’s going to be easier to adapt and kind of as both the laws evolve, but also as these different modes of monetization and usage and other things may evolve. It’s just really important to be proactive and, you know, just thinking about really the how, the why and the what is being used and where it’s being used. So you can make that, you know, you can make the judgment called the business decision.
And more importantly, how can you minimize your future risk?
Greg Kihlstrom (19:31)
Yeah, yeah, well, Sean, thanks so much for joining today and sharing your insights. One last question for you. I like to ask everybody, what do you do to stay agile in your role and how do you find a way to do it consistently?
Sean King (19:44)
It sounds corny and cliche, but I like to play. And what I mean by that is I love trying all the different models. love testing new and different things. I love playing around with new tools. You know, I always joke around that I’m technical enough to be dangerous, but not enough to do any real work. Right. When I count this one and I’m just, I’m getting more and more amazed as the weeks and months and years go on how closer that just curiosity and these different tools are making me that much more technology first. And so it’s just exciting just, and I encourage everyone just if you’re unsure, play around.
