Latest / Signal & Noise: Bringing Clarity to Media, Data, Tech and AI / Signal & Noise Special Edition: Advertising Week 2025
Transcript
- 0:00Hello, this is Rio Longacre here with my Co host Brett House for
- 0:03signal and Noise podcasts. And this week we have a special
- 0:06edition that come at you live from Adweek 2025 in New York
- 0:10City. Lots of people here.
- 0:12I think 20,000. I read that earlier today.
- 0:14It's a packed house. We're having a having a good
- 0:17time with a lot of great conversations, including with
- 0:19our special guest here, Jonathan Barnes.
- 0:21Jonathan, really excited to have you talk to our viewers here.
- 0:23Maybe we can spend a couple seconds, introduce yourself and
- 0:26tell our viewers about who you are.
- 0:28Yeah, thanks for having me. It's great to be here.
- 0:30My name is Jonathan Barnes. I'm a founder and CEO of a
- 0:33company called Population Science.
- 0:36In short, we help organizations reduce ad fraud so they can put
- 0:41more money towards working media earlier.
- 0:42Today we were talking a little bit about couple topics. 1 was
- 0:45curation and the other was outcomes.
- 0:48I know that's been a big buzzword this year, kind of
- 0:49won't go away. Love to hear what you're
- 0:52thinking, what you're hearing. Yeah, I mean, there's a lot
- 0:54going on in the space. And I think both of these
- 0:58buzzwords, if I'm even curious why you care about those things
- 1:03now because they've been around for so long.
- 1:07And I think what's different now is we actually have the tools at
- 1:10our fingertips to take control and and curate quality
- 1:14inventory. There's a wide array of
- 1:16publisher analytics tools. DSPSSSPS have become more
- 1:21advanced with the solutions that they're able to to provide AI,
- 1:27you know, buzzword, all buzzwords is, is, is definitely
- 1:32do a lot of interesting things in the space.
- 1:34I think you kind of put all those things together.
- 1:36And while we've talked about these things for a long time, I
- 1:40mean the time is now to be able to really curate and build
- 1:45really an end to end stack. Whether you're media buying in
- 1:48house at an agency or however you're doing it, you can, you
- 1:52can, you can finally do these things.
- 1:55And are we talking about outcomes a lot because of ad
- 1:57fraud and some of the concerns around transparency in the media
- 2:00ecosystem and whether or not there's, you know, players in
- 2:03the space that are doing things that are stealing impressions
- 2:06that aren't really driving outcomes.
- 2:07Is that, do you think that's a big driver of, of why outcomes
- 2:10based marketing is, is the, the big meme of the year?
- 2:13Well, I, I think it's all of the above.
- 2:14I mean, I go, obviously I'm very passionate about ad fraud and
- 2:17that's definitely a big part of it.
- 2:18But I think it's just signal in general.
- 2:21I mean, there's billions of impressions going through every
- 2:25day. I won't even get into the
- 2:28transaction ID that's that's going on.
- 2:31It's a whole other topic. Go to pre bid next week if you
- 2:34want to learn more about that. But I, I think it's just people
- 2:37are, are tired of being sold in this industry and they want to
- 2:42now have the ability to take control of themselves and, and
- 2:46and learn from themselves. What's quality signal, whether
- 2:49that's rooting out ad fraud, whether it's data efficacy, data
- 2:55activation strategies, I mean, you name it.
- 2:58Again, the tools are there that you have the ability to take
- 3:01control now. So I urge everyone in the space
- 3:04to just do it. Just get out there and do it.
- 3:07Thanks, Jonathan. Love it.
- 3:08Thanks for coming. And by the way, love the hat.
- 3:10Thanks. Hello Alyssa, great to see you
- 3:14here. Thanks for joining.
- 3:15Why don't we start by you telling our audience a little
- 3:17bit about yourself and your background?
- 3:19Yeah, thanks for having me. My name is Alyssa Firth.
- 3:21I lead ad Tech at Cordero, which is Omnicom's digital
- 3:25transformation consultancy. So focusing on helping brands
- 3:30bring together martech and AD tech and really think through
- 3:33how they can leverage their tech stack to drive more
- 3:36connectivity, transparency and performance with their media.
- 3:39Yeah. We were talking offline about
- 3:41data collaboration and data activation and how first party
- 3:44data, while it's not all that you need because it's oftentimes
- 3:47not differentiated from brand to brand, brands are not
- 3:51necessarily doing that the activation of their first party
- 3:54data across the the digital ecosystem effectively, right.
- 3:58Where do you think the blockers are the friction is and at least
- 4:03taking that first baby step to to sort of data activation
- 4:06before you start to bring in other data sources for sort of
- 4:09increased signal? Yeah.
- 4:11I mean, I think at first it's do you have all your data together
- 4:14and in the right place? Do you really understand what
- 4:16first party data you have and the value of it and how you can
- 4:20use it? And I think, you know,
- 4:22secondarily, it's how do I actually get it from my internal
- 4:25ecosystem into a paid media environment?
- 4:29And then when I do that, how do I actually prove that, that it's
- 4:32driving any outcomes? And so that closed loop is
- 4:35really what's missing. I think CDP, CDPS play an
- 4:39important role, but data clean rooms also play a really
- 4:42important role in that because they provide that ability, that
- 4:45intelligence layer. So it's like a portal, right,
- 4:48for you to send your first party data out.
- 4:51Well, not really even send it out, right, but activate your
- 4:53first party data and then get that measurement back in.
- 4:56So it's less of a black box and you can create more transparency
- 4:59with your paid media. You mentioned data cleaning
- 5:02rooms. I remember a few years ago there
- 5:04like the hottest topic you went to can.
- 5:06It was everywhere. It was like data cleaning rooms.
- 5:08Everyone had their own data cleaning rooms.
- 5:09And then I think adoption maybe didn't live up to expectations.
- 5:13I don't know how much of his hype and I know you've worked in
- 5:15the space for several years. Where do you see the state of
- 5:19the industry in terms of data clean rooms?
- 5:21Are they being adopted? Are they not being adopted?
- 5:22Where are we and where do you think we're going to?
- 5:24Go yeah, it's really tough. I mean, I think that brands have
- 5:27been experimenting with data clean rooms over the last couple
- 5:29years. They've been treating it almost
- 5:31like science projects, right? And I'm going to test out media
- 5:34clean rooms, maybe use a a neutral clean room.
- 5:38I think adoption's been really hard because it's not just the
- 5:40tech that's complicated, but actually the people and the
- 5:43workflows around it. How do I get my IT legal and
- 5:47marketing teams all to elaborate together and really understand
- 5:50the true value and create business cases around it?
- 5:53So that's been a really tough challenge for adoption, as is
- 5:56agencies and figuring out, you know, how are they going to use
- 5:59data clean rooms and then what is the role of agency versus
- 6:02marketer in data clean rooms? But as we see brands place more
- 6:06value on 1st party data, on measurement and transparency and
- 6:09also their AI agendas, you know, we need to start to move from a
- 6:13place of science projects and testing into making data clean
- 6:17rooms actually a strategic infrastructure investment so
- 6:19that you can build a really robust data ecosystem that
- 6:23includes both 1st and 2nd party signals.
- 6:26Also, thanks Melissa. Appreciate your time.
- 6:28We've got Tim Rowe, the editor from State of Streaming.
- 6:32Thrilled to have you. Thanks for stopping by.
- 6:34This was sort of a serendipitous encounter in advertising, which
- 6:39happens. It sure does.
- 6:40Tell, tell our audience a little bit about yourself and what
- 6:42you're doing at state of streaming.
- 6:43Absolutely. So about 5-6 years ago started a
- 6:47podcast all about billboards actually that has morphed and
- 6:50migrated into all things streaming TV, understanding that
- 6:53the challenges facing out of home is really what is facing
- 6:57streaming at large. So that focus has shifted to all
- 7:00streaming news on state of streaming.com.
- 7:03Nice, nice. So media Radar, we're focusing
- 7:05certainly a lot on on the rapidly growing sort of video
- 7:09web everywhere advertising ecosystem, the CTV advertising
- 7:12ecosystem. It's a big topic of advertising
- 7:14week. It's a big topic across the
- 7:16publisher and advertiser ecosystem in general.
- 7:19What do you think's really driving a lot of that shift?
- 7:22And we've seen a lot of move towards user generated content,
- 7:25the creator economy versus the Hollywood economy from
- 7:29originally produced content and a huge shift over the last five
- 7:33to seven years. That's that's that's really
- 7:35showing a domination of the the content ecosystem aligned to to
- 7:39UGC. How do you think that's driving
- 7:42some of the changes in the in the streaming and video
- 7:45everywhere advertising ecosystem and and and advertisers in
- 7:49general? How do they actually understand
- 7:52that and take advantage of that of that sort of change in
- 7:55consumer behavior towards UGCI think it's evidence of what we
- 7:58always knew to be true that the consumer has control and the
- 8:02unreasonable consumer isn't going to go back.
- 8:04We got used to free one day shipping and having our content
- 8:08the way we like it on demand and the platforms that we like to
- 8:10watch it on. A few years ago, as I was
- 8:13watching my son kind of as a, as a, you know, young student
- 8:16navigating and lockdowns and things like that, I noticed he's
- 8:20watching YouTube on the living room TV all day long.
- 8:24I started to ask him, like, are you watching YouTube?
- 8:26What are you doing? And he said, I'm watching TV and
- 8:30kind of clicked then that was 5 or 6 years ago.
- 8:33This is just the way that his generation is going to watch TV.
- 8:36And I think the same can be said for for us and how we want to
- 8:40consume content. So UGC, sports on demand, it's
- 8:44really the way that we want to consume media.
- 8:46And now the market having to respond to that.
- 8:50I think out of home over the years has kind of been
- 8:52overlooked or neglected. How is the shift to digital out
- 8:56of home kind of I think bring it, integrate it with the rest
- 9:00of advertising, really create closed loop in terms of
- 9:03experiences, data reporting. Can you maybe comment on that a
- 9:05little bit? I think what's interesting is I
- 9:07think that 25 years of of display has built a tolerance
- 9:12for maybe things that aren't perfectly measured or perfectly
- 9:16targeted and understanding now the fraud risk, how much of my
- 9:20ad budget is actually being absorbed by things that aren't
- 9:23humans. I can reallocate that budget to
- 9:26digital out of home. I can suddenly increase my
- 9:29incremental reach with real humans, people that can buy my
- 9:32products or services. So integrating digital out of
- 9:35home into the media plan is that next natural transition I think
- 9:39from display to something that's a larger format offering similar
- 9:44unit economics but with bigger impact and reach?
- 9:48That's awesome. Thanks for coming, Tim.
- 9:49Really appreciate it and good seeing you for having me.
- 9:51We are privileged to have Scott Messer, the founder and
- 9:54principal of Messer Media, who's very well known in the ad space.
- 9:59And Scott, I've, I've looked a lot at your at your LinkedIn
- 10:05post. I saw your presentation at the
- 10:06Admonster sell side summit. You really established the fact
- 10:11that publishers have a problem, there's a monetization problem,
- 10:14there's a content problem, there's an LLM problem, but you
- 10:17really laid this out in sort of three core principles.
- 10:20Sort of admit that you've got a problem first in order to solve
- 10:24against this, this challenge that the ecosystem is facing.
- 10:26Can you talk a little bit about that for foot folks?
- 10:29Our audience that hasn't seen this presentation, hasn't seen
- 10:31your post on LinkedIn, what that looks like.
- 10:34There's three kind of core principles and what your
- 10:35solutions are for publishers going.
- 10:37Forward, absolutely. So set the stage a little bit
- 10:40when I was on stage AD Monsters in Asheville as the Co host of
- 10:44the event, I practice something that I call Adtech Therapy.
- 10:48It's a service that we provide inside Messer Media.
- 10:51It is the rehabilitating your revenue and fixing your problems
- 10:54that you can get back to crushing your goals.
- 10:57And part of therapy is sort of going through the, the processes
- 11:01of grief. And one of the first ones is
- 11:03denial. And as I was thinking about what
- 11:06I'm going to do for my, my kicking off speech for this
- 11:08conference is thinking about getting publishers to the right
- 11:11place mentally so that we can have productive conversations.
- 11:15And part of that is like letting go of the past a little bit.
- 11:18So I laid out three basic premises for them.
- 11:21The first is that programmatic is broken and it simply cannot
- 11:24be unbroken. The second part is that the open
- 11:29Internet has stalled out from a growth standpoint.
- 11:32Where there was $70 billion going into the market in like
- 11:352015, 2018-2019, there's $80 billion going into the open
- 11:39auction now. And that is a tiny fraction of
- 11:42growth compared to the rest of digital media.
- 11:45Expenditures went from like $100 billion up to 400 or $600
- 11:50billion depending how you do the math.
- 11:52So while everything else is 4X, the digital web has grown by
- 11:56less than 10% or 12%. That's bad and I'll talk about
- 12:02why in a little bit. And then the third part is that
- 12:07LLMS have devoured the long tail of content and that is gone.
- 12:12So the content that of yesterday has been consumed and publishers
- 12:17need to think about where we go next from a content perspective.
- 12:23So as you think about these three things, what I told
- 12:26publishers to do is you must push for transparency in
- 12:30programmatic. We need to get a better system
- 12:33for advertising, otherwise the money is going to move
- 12:37elsewhere. And you're seeing things like
- 12:39this with the trade desks move to open ads recently, the fiasco
- 12:42with transaction IDs, and so on and so forth.
- 12:46With the open web, we don't compete against each other as
- 12:49publishers, we compete against the walled gardens, and we have
- 12:53to get our collective shit together to fix that.
- 12:56Similarly, when you think about like the performance of the open
- 13:02web, my greatest fear is that advertisers simply take their
- 13:05money and leave. And if performance is broken,
- 13:11why would an advertiser want to spend money in the open auction?
- 13:14If it continues to get smaller and smaller, it's going to be an
- 13:17unviable place for them to spend money.
- 13:20And that's bad for journalism and creators overall.
- 13:23And then when you combine that with the fact that LLNS are
- 13:26starting to shrink the amount of traffic that's going on, you're
- 13:29getting to a smaller and smaller world to the point where there's
- 13:35going to be nothing left. And I fear that the budgets that
- 13:37the LLMS are going to take as they move into advertising,
- 13:40because open AI and perplexity will move into advertising very
- 13:45soon, the money and where those budgets are coming from isn't
- 13:48like a new search budget. It's going to come from open web
- 13:52display. It's a comparable experience,
- 13:54maybe a little bit coming out of search, but predominantly from
- 13:57open web display. And that is going to be the
- 13:59final nail in the coffin, the open web monetization ecosystem,
- 14:04that's bad. So I need everyone to get up to
- 14:07the starting point of like here we are programmatics in a bad
- 14:10place. The open web is having
- 14:11performance issues and we're having a serious traffic issue.
- 14:15OK, now that we agree upon that, can we start finding solutions
- 14:18and move forward? Scott, this is real.
- 14:21I love that. You know, it's been a tough
- 14:24couple decades for for publishers.
- 14:26You know, they've they've they've really struggled.
- 14:28Monetization is falling quite a bit and I think it continues to
- 14:31collapse, especially some publishers.
- 14:33What would be your kind of advice for them to besides
- 14:37calling the ad tech therapist? What be your advice to them in
- 14:39order to? Well, when you do call the ad
- 14:40tech therapist, this will be your advice that you get and
- 14:43it's free for you guys today. So firstly, even though
- 14:48yesterday's content was stolen, we have tomorrow's content to
- 14:51fight for. So think about that.
- 14:54You have a future and it's worth protecting.
- 14:56Which means establishing baselines with LLMS and how you
- 14:59want your content to be handled. Are you blocking bots?
- 15:02Are you setting up frameworks and marketplaces that you can
- 15:04actually license your content? You have to be business ready.
- 15:09It's like publishers are in the business of giving information
- 15:11and giving answers, and whether you're giving that directly to
- 15:13an LLM or directly to a user, you need to have a business
- 15:16model and the frameworks for monetization.
- 15:20So get those things set up and then work on content format
- 15:24innovation. A media company's job is 3
- 15:27things, produce, distribute and monetize IP.
- 15:31And then you keep running this thing around like a flywheel
- 15:33where the monetization keeps funding content.
- 15:36Your website is not your IP. That is a distribution channel
- 15:39and a monetization place. You must have more places to put
- 15:43your content. So start innovating content
- 15:45formats and move beyond your web and start building your brand
- 15:50and going to the meeting the consumers where they are.
- 15:52And that might mean going back into the walled gardens and
- 15:55going into other locations. Like podcasts?
- 15:59Thanks, Scott. Great having you here.
- 16:01Always a pleasure, Rio. Thank you very much.
- 16:04Krish Raja, who is coming to us all the way from Australia, I
- 16:09believe, right, Melbourne, from Sydney, from Sydney, Yeah,
- 16:13Sydney, Australia. And you've got some, we've
- 16:14talked, you know, before. You've got some really good
- 16:16thoughts on publisher monetization and on the
- 16:19importance of owning data and how critical that is to publish
- 16:25your sort of content strategies and ability to better monetize
- 16:28their data. And we'd love to hear what you
- 16:30have to think. Yeah, it's the topic we love.
- 16:31Yeah. I've been in the publishing
- 16:34monetization space for a long time, more than 15 years,
- 16:37publishers directly at tech providers, martech, and now I
- 16:42run advisory in Gen. AI and data for content
- 16:45businesses. And yeah, look, I think the
- 16:48thoughts have roughly followed a similar thread throughout those
- 16:51years, which is that publishers want people to come to their
- 16:54properties. And in a in AI world, there are
- 16:57more threats of disintermediation and being
- 17:00removed from your customer. And so I believe that at the
- 17:04core, first party data is important.
- 17:06First party data is your signal, your understanding that you, you
- 17:10know, your customer and in, you know, in the years gone by, all
- 17:15we had to be able to do that is an e-mail address.
- 17:18Someone types their e-mail in, you capture their e-mail and
- 17:21that's your main 1 to 1 mechanism.
- 17:24That's, you know, not necessarily the future of how we
- 17:28monetize in, in open web publishing.
- 17:30Certainly we've always found it a little bit difficult to, to
- 17:33turn e-mail addresses unless we're monetizing it in, you
- 17:36know, actually in Facebook. We've always had to find a bit
- 17:39of a difficulty to get the scale we want to get the the level of
- 17:43sophistication we want when we're running e-mail address
- 17:46matching. And so yeah, look, I think first
- 17:49party data is the critical juncture to embedding a good AI
- 17:54LED monetization strategy. I think AI minus data equals
- 17:59risk getting equals hallucination.
- 18:02It includes potential disintermediation from your
- 18:05customers. And so I you know my big
- 18:07proponent of first party data. What about third party data?
- 18:09What does that fit in Chris? Well, I think it's very
- 18:12relevant. I think that you need to have
- 18:14information from other businesses and their customers
- 18:17or other data sources in order to enrich and scale your data
- 18:21set. Your data set is unlikely to
- 18:24contain all your customers and if it does, it's unlikely to
- 18:27know everything there is to know about that customer.
- 18:30And so absolutely, I think third party data is really critical.
- 18:34I just don't think it will be the only strategy in the future,
- 18:38which it kind of has been for a little while.
- 18:40It's been, it's been the main event for for a lot of
- 18:44monetization. And should pubs be afraid of AI
- 18:48or embracing AI? Or maybe both?
- 18:49What do you think? Both.
- 18:51I think a bit of fear is healthy.
- 18:53I think you have to, as a non AI native business, it's important
- 18:56to take this seriously. This is AI is not, you know, a
- 19:01fad and it's not going away. I think we all agree on that.
- 19:05And so I think a little bit of healthy kind of sit up and and
- 19:08go, OK, well, this is, this is a big deal.
- 19:10We've got to, we've got to take this seriously.
- 19:12But optimism for sure. Who has more access to loyal
- 19:16customers who, who has access to content and proprietary content
- 19:21that is able to be amplified with AI, the big content
- 19:26businesses. So if they if they nail their
- 19:27USP and and apply AI in strategically correct areas for
- 19:32their for their business, it's going to be a game changer for
- 19:34them. Thank you.
- 19:37Exciting to have Alexis Hot Lightner on our podcast.
- 19:40Alexis, Welcome Amy telling our viewers a little bit about
- 19:43yourself. Sure.
- 19:45So I run my own consultancy now, have worked at the walled
- 19:48gardens, Google, Facebook, Amazon, working to help out,
- 19:53figure out how we can get advertisers to better understand
- 19:57how to optimize their media dollars with with tech
- 20:00capabilities. And now I'm off on my own
- 20:03running Arc Digital and currently running a supply path
- 20:07optimization study on behalf of Google, Open X and Asus run by
- 20:12the 614 group at the Brand Safety Center.
- 20:15It's just I love the fact that came to this industry financial
- 20:19services initially, which I thought was so interesting.
- 20:22Maybe you could tell our viewers a little bit about how when you
- 20:25came to this industry, financial services, which is definitely
- 20:27rigged, and when you came into advertising an attack, you were
- 20:31a little surprised and you found out there were so few
- 20:35regulation. So when you think about it, the
- 20:38stock market and the advertising markets are similar in that
- 20:42they're both auction based. But really it stops there.
- 20:46If you think about the regulations when you buy a stock
- 20:50or sell a stock, you know who's on the other side of the trade
- 20:53Regulations like know your customer and DTC like clearing
- 20:58houses. There's a clearing agent who
- 20:59tells you what stock you bought at what price and when.
- 21:03That's absolutely not the way that the programmatic the system
- 21:07works, the way there's multiple intermediaries, tons of
- 21:12reselling of inventory, hidden costs, hidden intermediaries.
- 21:16So really you don't know where your media dollars are going and
- 21:20sometimes those media dollars are going to fraudulent or
- 21:25publishers, some who are the buzzword of last year was.
- 21:29Yeah, lots of standards, but not enough let enforcement
- 21:32regulation. What we talked about before that
- 21:34before this episode, we tell us a little bit about the study
- 21:37that you're putting together around supply chain optimization
- 21:39and transparency. Because I know you have a real a
- 21:41lot of interesting thoughts on how that can be improved to kind
- 21:44of match some of the more regulated industries that just
- 21:48have less friction, right and more transparency built in.
- 21:50Exactly. So one of the big problems and
- 21:56basically we're going to study in 2024 and for dramatic
- 21:59advertising and they realized that they basically interviewed
- 22:03advertisers who all said they didn't have enough information
- 22:06to take interest when it comes to understanding and making
- 22:10decisions about their supply path.
- 22:12And that's kind of what this study is aiming to do is how can
- 22:17you know these sponsors, Google open exit basis, they want to
- 22:20learn what do advertisers work out when it comes to their
- 22:24supply paths and how can they surface that data in a different
- 22:27way. If you think about it, everybody
- 22:29wanted access to log level data and then it was really hard to
- 22:34like honest that data and do anything interesting or
- 22:37insightful with it. So.
- 22:39Be careful what you ask for. So really that's that's the goal
- 22:44of the study is to understand what it is that and then help to
- 22:50kind of understand what we have to our tech platforms and what
- 22:56do they need to build and help them to validate their road map.
- 22:59That's great. Thank you for joining us.
- 23:01We've got one of my favorite guests, Shiv Gupta from U of
- 23:04Digital. Shiv, welcome.
- 23:05Great to see again, as always. Can you tell our audience a
- 23:07little bit about yourself? Sure.
- 23:09I'm Shiv. Thanks for having me on.
- 23:12I have been in AD tech for a long time.
- 23:15I started my career at AOL. I was there for about 10 years.
- 23:17Was that Criteo mostly in sales and sales leadership type of
- 23:20roles. Now I run a company called U of
- 23:22Digital. We provide training for the
- 23:24marketing and advertising industry.
- 23:25We train on tech, we train on the zeitgeist.
- 23:28We train buyers and sellers and obviously we train on AI because
- 23:32that's what everybody needs to learn about right now.
- 23:34Awesome. So we, you mentioned AI
- 23:37enablement and you digital's at the forefront of this.
- 23:40I know that's a lot of what you're doing.
- 23:42I'd love if you could maybe talk to our audience about the state
- 23:44of AI enablement, big companies versus small companies impacting
- 23:47workers, What's going on in the spaceship?
- 23:49Yeah, So I mean, right now, obviously AI from a product
- 23:52perspective is moving 1,000,000 miles an hour.
- 23:55Companies are, you know, ChatGPT and open AI is like shipping
- 23:58speed is insane. Like every week they're putting
- 24:00out product. And naturally that means people,
- 24:03some people, a lot of people are falling behind, right?
- 24:05They can't keep up with everything.
- 24:06They don't understand all the innovation, they don't
- 24:08understand the use cases, etcetera.
- 24:09So we're seeing that everywhere. It's rampant.
- 24:12People are struggling to keep up and a lot of companies are not
- 24:16providing structured ways for their people to keep up.
- 24:19And it's actually interesting, like some of the most foremost
- 24:22AI companies, the big tech companies are not providing like
- 24:25the tools that their people need to keep up, which I find to be
- 24:28fascinating. And I think the companies that
- 24:31will do a better job of that either internally or through
- 24:34providers like us will, will leap ahead, right, because
- 24:38everyone's shipping product. But if you can bring your people
- 24:40along for the ride, like the people need to be the ones using
- 24:42it. Fair to say people are the
- 24:44bottleneck, is that it? I think that's accurate.
- 24:46Yeah, I think, I mean, listen, there's probably other
- 24:47bottlenecks. My personal, what I see and what
- 24:51I do is try to undo the bottleneck around people.
- 24:55And people are a massive bottleneck right now to AI
- 24:57adoption to AI kind of like speed to market for companies in
- 25:01all sorts of ways. When you're looking at AI
- 25:05adoption, is it, is it impacting jobs?
- 25:08Is is eliminated? There's a lot of fear about
- 25:10that, right? Is eliminating jobs is creating
- 25:11jobs or do we not know yet? What do you think?
- 25:15I think it's eliminating jobs, right?
- 25:17Most companies, market caps have grown in the last year, right?
- 25:22The market is doing really well and big tech companies are
- 25:24growing and open. AI is worth $500 billion, but
- 25:28there are not as many jobs out there and employment,
- 25:32unemployment is high and companies are laying people off,
- 25:34right? So it's it's not good.
- 25:38I'm sure that is just companies right sizing like after kind of
- 25:40the COVID like like insane hiring spree versus like actual
- 25:45structural because AI is starting to eliminate roles.
- 25:49Do things that you know what one person can do, maybe the job of
- 25:54three or four people, right? Which you're right if they're AI
- 25:56enabled. What you think?
- 25:57I think it's a mix. I mean, I think there's
- 25:59definitely like still bloat that people are cutting from over
- 26:02hiring during the pandemic, but I think it's a mix, right?
- 26:04Because I think you'll see companies attribute layoffs to
- 26:09AI specifically, right? And so you got to, you got to
- 26:11take them at their word. I think some of them are doing
- 26:14it proactively because they actually haven't integrated AI
- 26:17into their work streams fully yet, but they're kind of trying
- 26:20to get ahead of it. Maybe.
- 26:21I also think they're using it as a forcing function.
- 26:23So companies saying like, hey, we're going to let people go or
- 26:26we're not going to hire because you have to figure out how to
- 26:28use AI to ship product or do whatever you do, commercialize,
- 26:32etcetera, right? So I think it's a mix of both.
- 26:35Any advice to our viewers on AI adoption?
- 26:38Like personally or company widescale?
- 26:42Any words of wisdom here? Well, so one thing I would like
- 26:45to say is the people that are running the companies, the
- 26:48visionaries, the product, people that are shipping product, they
- 26:52need to think about enablement. A lot of companies don't think
- 26:54about it. They expect people to learn
- 26:56through osmosis and that's not scalable and it's not working,
- 26:59right? So I think people need to think
- 27:01about structured enablement more, especially in the age of
- 27:04AI. Otherwise people are going to
- 27:06fall behind. So that would be 1 like piece of
- 27:08advice. And that could be, again, you
- 27:09could hire somebody in house to do it.
- 27:11You can use a company like us to do it.
- 27:14You have to do in a structured way.
- 27:15And then the other thing I'd say is a cultural shift, right?
- 27:18Like, first of all, leaders have to lead by example.
- 27:21They have to they have to use the tools.
- 27:22They have to figure out how to like show people that they are
- 27:25using it and that they have to like, you know, one thing I
- 27:27posted recently on social media was like they have to make sure
- 27:30IT and legal teams are not in the way and compliance teams of
- 27:34using the tools. That's a big bottleneck too
- 27:36right now. So couple of couple pieces of
- 27:39advice. As also you have thanks so much
- 27:40for coming by. As always, appreciate your
- 27:42insights. Enjoy the rest of that week.
- 27:44Thank you. We've got Karsten Vida from West
- 27:48Media Research with us. Karsten, it's great to see you
- 27:51see each other at all these events, becoming a nice habit.
- 27:54Maybe you could start by telling our audience a little bit about
- 27:56yourself. Yeah, I'm running a little
- 27:58research boutique W Media Research and we cover
- 28:02advertising technology, the business of advertising
- 28:04technology. And most of the stuff we do is
- 28:08custom research, custom go to market advisory.
- 28:11Awesome. Well, one topic that I know
- 28:13you've written about a lot recently, I saw you had a great
- 28:16write up in Market Texture the other day on it is about AI
- 28:21search of like AI search and how that's changing the way people
- 28:26look for things. It's changing the way that
- 28:28people look at search results. It's certainly impacting the
- 28:31amount of traffic going back to publishers and it's impacting
- 28:33their monetization in a meaningful way.
- 28:35Love, you could talk a little bit about that.
- 28:37Yeah. So what I've been looking at is
- 28:40the impact of AI search in chat bots on user behavior.
- 28:44So what we see is that a lot of users switch from traditional
- 28:49search to AI overviews in search or chat bots all together.
- 28:55And the problem with that is that the AI based search results
- 29:00or chat bot, you know, messages are they give you the complete
- 29:06research. You don't have to go out anymore
- 29:08and click like 5 or the 10 blue links to find what you're
- 29:11looking for. Instead, you already have the
- 29:13result right there in front of you, which means you don't have
- 29:16to go anywhere else anymore. And and the result of that is,
- 29:19is that traffic for independent, independent publishers primarily
- 29:25is going down quite dramatically.
- 29:27How dramatically so? So there's a ton of numbers all
- 29:31over the place anecdotally. And what I did recently was
- 29:35pulled it all together and just to see how it falls out across
- 29:40the percentages. And since for most publishers,
- 29:43they lose like 30 or 40% of their of their traffic, which
- 29:47then means they lose 30 or 40% of their advertising revenue.
- 29:51And for many publishers, that's very serious revenue loss.
- 29:56Are CPMS staying consistent or are they starting to drop?
- 29:59Are they going up? There are fewer auctions.
- 30:01I haven't heard anything about that, but I would expect them to
- 30:03go down just because there's less demand.
- 30:07And so it's not only a problem for the publishers, it's also a
- 30:13problem for the vendors downstream because they
- 30:16facilitate the advertising that the publishers are running on
- 30:18their sites or apps. And if there's less advertising,
- 30:22then they make less money too. So for both of these publishers
- 30:27as well as as vendors, they have to look for diversification.
- 30:31They have to branch out into other stuff.
- 30:33Is the open web in trouble? Possibly yes, possibly yes.
- 30:39What worries me the most is the fact that so traditionals,
- 30:44despite all The Who ha ha about AI, you know, it's so fantastic
- 30:49and everybody's using AI. Traditional searches stood 30
- 30:54times the traffic of more than 30 times the traffic of the chat
- 30:57bots. And yet even though chat bot
- 31:00traffic is like it's tiny compared to search already, we
- 31:04see these losses in revenue or traffic on the publisher
- 31:07websites. What's going to happen if even
- 31:11more users adopt chat bots and AI based search?
- 31:16Then you're going to see even more dramatic traffic loss on
- 31:18the publisher. 30 to 40% drop is that's a lot.
- 31:22That's a lot, but there's cases where there's smaller sites
- 31:27because there's one travel site out of Australia, 90% traffic
- 31:31loss, they just have to shut down, right?
- 31:33It's impossible to make up for it.
- 31:35What are you telling them Publishers Like what advice are
- 31:38you giving them? You mentioned diversification.
- 31:40They have to diversify, so subscription is definitely at
- 31:44the top of list. I've talked to a big publisher
- 31:49today and they say, you know, by Christmas 90% of our users will
- 31:53be subscribed and they just kept making meats and meats anymore
- 31:59ends meet anymore, you know, So it's a pretty dramatic
- 32:04situation. Event business is another one,
- 32:09optimizing, producing data that I can use and then license them
- 32:14out to AI. That's another path to make
- 32:17money. But yeah, it's gonna be tough.
- 32:21By the way, it's also the reason why you see this whole stuff
- 32:24around transactional IDs, right? So used to be your publishers
- 32:31offer the same piece of inventory on like 10 different
- 32:34platforms, right? 10 different exchanges or even
- 32:37more. Well TID you get some clarity
- 32:39and supply path. Right.
- 32:40Well, it used to be the TID. No matter where this inventory
- 32:43showed up, it was the same TID, which meant the buyers could
- 32:46compare. So they could figure out where's
- 32:49the best price for the same piece of inventory across those
- 32:5310 offers, right. And what publishers did was via
- 32:56their pre bid organization, which is a header bidding
- 32:59organization for publishers. They said, hey, we're going to
- 33:03change that. So now every time the same
- 33:06inventory gets offered somewhere else, it's got a separate
- 33:09individual ID. Doesn't sound like that incense.
- 33:12It almost sounds like that's incenting bad behavior if you
- 33:14ask me, right? I mean.
- 33:15Well, you can't really blame the publishers.
- 33:17I mean, they're in such a desperate situation and this
- 33:20makes up for like, this makes like 1015% more revenue.
- 33:24That's pretty good. So they're already like, you
- 33:28know, half out of the bottom, but these big losses you.
- 33:31Have but isn't it incenting like the publishers are getting an
- 33:34outside share of let's say monetization because they're
- 33:37gaming system a little bit by you don't think so?
- 33:40They're struggling so much already, which is not in their
- 33:43business interest. And it's not in our interest
- 33:45either, because if there's no independent publishers, then we
- 33:48don't get independent news, we don't get independent
- 33:52entertainment, we don't get connectivity with other people.
- 33:56And we don't get democracy either, right at some point,
- 33:58right? That's political implications,
- 34:00correct? The Republic needs open debate,
- 34:04open, you know, discourse. And if there's no independent
- 34:07publishers anymore, it's not a good thing.
- 34:11This is super interesting and there's a lot to unpack.
- 34:13Thank you for your time. This is great.
- 34:16Thank you for having me, I appreciate it.
- 34:18So my name is Ethan Steininger. I'm the founder and CEO of Mix
- 34:21Peak. We are a multimodal AI
- 34:24infrastructure platform, and So what that means is we give
- 34:27developers the ability to combine unstructured data in
- 34:31various ways that are newly possible.
- 34:34So developers use us for contextual targeting, brand
- 34:37safety and creative attribution. Fantastic subscript taxonomy or
- 34:42taxonomy. So maybe overview is a little
- 34:44bit about it. What is it?
- 34:45Why is it important? Yeah.
- 34:47So taxonomies at its core are kind of hierarchical mappings of
- 34:52relationships. And so we think of taxonomies in
- 34:55the context of content taxonomies, which is really,
- 34:59let's say you have a video and that video contains certain
- 35:03subjects. If we're able to map the
- 35:06different subjects in the video down to the time stamp, then we
- 35:09can do better targeting. We can ensure that those
- 35:12concepts don't appear next to other concepts.
- 35:14So for example, if tobacco is a flag concept, it's really hard
- 35:19to identify, hey, those leaves in the corner of that video is
- 35:22tobacco. And so this is what granular
- 35:24taxonomy is really enable in the AD ecosystem.
- 35:27And you guys are doing some work with the ID on on this content
- 35:31taxonomization. Can you tell the the audience a
- 35:33little a little bit about what you're doing, what that looks
- 35:36like and what the kind of end results will be as you roll
- 35:39something some of these these frameworks and approaches out to
- 35:42the to the load market? Yeah.
- 35:43So IAB actually goes through a number of versions of taxonomies
- 35:47over time. So there is a whole swath of
- 35:50customers that are using IAB taxonomy 2 point O.
- 35:55They have their own taxonomies. And so there's a concerted
- 35:58effort to consolidate these taxonomies into what is now 3.1.
- 36:04And so we're working actually on a taxonomy mapper that takes
- 36:08previous versions and combines them together in order to adhere
- 36:12to the new 3.1 standard. And what 3.1 enables is just
- 36:16more granularity on it. So if you think about like
- 36:19contextual targeting, brand safety, all of these fields, the
- 36:22value is in the granularity of the targeting.
- 36:25And so if you can, if everybody speaks the same language, then
- 36:28you can ensure that everything is placed next to what you want
- 36:31it to be placed next to. Eras Levin.
- 36:34He's a good friend and well known commentator on many
- 36:37subjects. Eras love if you could tell the
- 36:39viewers a little. Bit about yourself.
- 36:41Yeah, I think what's sort of relevant here is I've been in
- 36:44space a long time. I was at Google for a very long
- 36:47time, sitting sort of in between the really the intersection of
- 36:51supply and demand, both on the supply side and then the demand
- 36:53side. And I really saw a lot of
- 36:55changes over many years in the industry, things that I thought
- 36:58were not really healthy for the ecosystem.
- 37:01I tried to fix many. I was successful fixing some and
- 37:04some I fell short. And so now I'm trying to fix
- 37:07those things from the outside as a consultant helping different
- 37:09companies focused on those solutions.
- 37:13Eris, you've been a very vocal critic of the term outcomes,
- 37:17outcomes era. A lot of you could tell us a
- 37:19little bit about that. Yeah.
- 37:21I mean, ultimately, I've seen the sort of gaming, the
- 37:24gamification of this industry. Generally speaking, I think it's
- 37:30ultimately because buyers took their eyes off the ball of how
- 37:34marketing effectiveness should be measured.
- 37:37And ultimately I think what sort of pave the way for that is
- 37:41buyers reliance on vague terms that sound good on paper, but in
- 37:46reality translate into something that is not really tied to
- 37:50effectiveness. And so outcomes more often than
- 37:53not, while we think they're business objectives and KPI's,
- 37:57but more often than not they end up meaning something that is
- 38:00really a vanity metric. So something like attribution
- 38:03that isn't actually incremental or maybe even worse than that,
- 38:07vanity metrics, viewability and completed views that are
- 38:10completely detached from actual effectiveness.
- 38:13And so I sort of, I think that we've been in this outcomes era
- 38:17for a very long time pushing towards those vanity metrics.
- 38:20And what I'm very adamant about is that we have to now pivot and
- 38:23transition away from outcomes, or at least away from terms that
- 38:26are so sort of dual meaning. And that even though some people
- 38:30up high really do mean the good and nice and noble
- 38:35interpretation generally on the ground, that gets translated
- 38:38into something that is just easy to do and achieve and doesn't
- 38:42actually tie to effectiveness. And that, that leads us to our
- 38:44next question is that you've talked a lot about quality era
- 38:47of advertising, like we're entering this quality era and
- 38:50that seems to be thematically related to what you just
- 38:53discussed around outcomes and what this really means.
- 38:55And, and how would you define that?
- 38:57What are your talking puts around around quality
- 38:59advertising or the quality era? Yeah.
- 39:01So I think ultimately most marketing effectiveness is
- 39:06actually determined. You know, historically and even
- 39:09through the digital era, it's based on how often you reach a
- 39:13customer and the impact that can be measured over the short and
- 39:18the long term. That reach was usually actually
- 39:22quality adjusted in the past, but in the digital era that got
- 39:24a little bit disconnected and we only looked at quantity.
- 39:26We looked at these vanity metrics and then how many you
- 39:28got and then tying it to some sort of outcome, which often
- 39:31wasn't incremental. And so the idea of quality, it
- 39:36actually it's a subjective term, right?
- 39:38It's really big. But ultimately it's thinking
- 39:41about the relative value of media.
- 39:42Not all impressions, not all of each, not all media is equal.
- 39:45And so how do we define the relative value of media
- 39:49according to different criteria, whether it's the time of day or
- 39:52the Geo or how prominent the ad is on a page, Is it large or
- 39:55small? Pay different amounts for that.
- 39:57And I think ultimately that's the ingredient that helps drive
- 40:01the real effectiveness and the real business outcomes that we
- 40:03care about. Anne Thiel.
- 40:05Anne, if you could maybe tell our viewers a little bit about
- 40:07yourself and your background, that'd be amazing.
- 40:09Welcome. You got it.
- 40:10Thank you for having me. My name's Anne Thiel.
- 40:12I lead advertising and revenue operations for Cox Automotive.
- 40:17What that means is that I have agency and publishers on my
- 40:23team. So I lead the publisher side for
- 40:26add operations for Auto Trader in Kelley Blue Book and I lead
- 40:30the agency side for dealer.com. I just finished a panel and we
- 40:35were talking about how all of the tech stacks work together.
- 40:42Agentic AI is is. I know it's the buzzword this
- 40:45week. Probably all.
- 40:46Around. I've heard a lot of it.
- 40:49Agentic AII am rebranding in my own department as the Great
- 40:54Connector. So in my department, we have two
- 40:57separate tech stacks. We want to be able to utilize
- 41:01the teams across both publisher and agency.
- 41:06We want to make sure that we're using all of the tech in the
- 41:10right way. We want to make sure that we're
- 41:12using our audiences in the right way.
- 41:15And with Agentic AI at the enterprise level, we will be
- 41:19able to connect everything from the tech stacks and that's both,
- 41:23right? So there's two of them.
- 41:24We will be able to do that and we will be able to utilize like
- 41:29our media buying team will be able to utilize them across all
- 41:33of the different tiers. So the OEM agencies all the way
- 41:37down to the brick and mortar dealers, we'll be able to
- 41:40connect all of that impression inventory, Auto Trader, Kelley
- 41:43Blue Book, our Cox DSP through from fluencydealer.com, we'll be
- 41:49able to do all of that and we'll have media planning Tier 1 to
- 41:53Tier 3. Sounds like a really amazing
- 41:56bold vision at an enterprise. It must be challenging.
- 42:00What are some things you've been running into?
- 42:01Like some maybe potential pitfalls or just things that
- 42:05have been tough that you think you'd like to tell other people
- 42:08about if they're going to go through similar journeys.
- 42:10Thank you. Great question.
- 42:12I it is very important number one to pick out the pain points
- 42:20in the process. So when you're getting all the
- 42:22way down to the very tactical level of the folks who are doing
- 42:26the media buying or doing the trafficking on the teams, you
- 42:29need to understand what are those repetitive processes that
- 42:32they're doing that are taking the time automate those first.
- 42:36Maybe your agentic solution can also do some automation.
- 42:40We have already gone through the painstaking process of
- 42:43automating like a lot of the work flows.
- 42:46So it's easier for us to implement the enterprise agentic
- 42:49solution. So that's where I would start.
- 42:52If people like, if you haven't done that yet, you need to look
- 42:55at ways to automate processes first before sort of
- 43:00implementing that that agent that will be actually performing
- 43:04the processes for you. Do you think it's just let's say
- 43:07Gentek workflows or Gentek AI? I mean how is it going to change
- 43:11experience for like business, business users as they use
- 43:15software? Do you think it's going to be
- 43:16completely different in a couple?
- 43:17Years I do, actually. I think that the teams, I think
- 43:19operations teams are going to look a lot different.
- 43:22I think salespeople are going to be able to have campaign data
- 43:26and selling points at their fingertips when they're sitting
- 43:29in front of a client, they're going to be connected, you know,
- 43:32through the agent to all of our systems.
- 43:35They're going to be able to say, oh, we do have a million
- 43:38impressions for you available on the homepage of Auto Trader next
- 43:41month. If you would like that.
- 43:42That's a very like non complex, right like answer and an example
- 43:47there. But they're so for sellers
- 43:51having all of that information at their fingertips, they won't
- 43:54have to go back and put in a sales force request and get the
- 43:58impression inventory. And that takes weeks.
- 44:01I think for sellers to be able to have that information readily
- 44:06available when they're sitting in front of a client is going to
- 44:08be game changing. Sounds amazing and I'm sure we
- 44:11could talk about this for hours. This is really interesting.
- 44:14I don't want to take any more of your time, but thank you so much
- 44:16for stopping by. Thank you.
- 44:18Bob Wozak, who's the CEO of Med Connect.
- 44:21Bob, great to have you here. Can you tell us, tell our
- 44:23viewers a little bit about yourself?
- 44:24Good to be here. Your name is Bob Wozak.
- 44:26I am the founder of mad connect. We are a intelligent
- 44:31connectivity layer for the mar tech and AD tech industry.
- 44:35I've been in this space for north of 25 years at this point
- 44:39as ACEO now three times, AGM twice and ahead of product
- 44:43twice. This is a really cool topic and
- 44:45it's you just said it a second ago.
- 44:47ICL or intelligent connectivity layer, love if you could tell us
- 44:50what this is, why it's important.
- 44:52Yeah, we built. An API connectivity hub.
- 44:55Originally, because the market was moving from third party to
- 44:581st party data, we thought there was a need for greater
- 45:01connectivity between the platforms you license to build
- 45:03your. Martech and ad tech stack.
- 45:05Ultimately, with AI, it has evolved into an intelligent
- 45:08connectivity layer. As we see, that's a foundational
- 45:10layer to connect all the platforms that companies and
- 45:13their users are using. Today to be able to perform
- 45:16their various tasks and functions and now with AI you
- 45:20can build that and automate that through API connectivity.
- 45:23And we talked about how this connects to basically every ad
- 45:25tech and Mart tech platform that a company is, whether it's a
- 45:28publisher and agency, a brand is currently managing in their sort
- 45:31of total stack, right, right. Could be warehouses, could be
- 45:34DSPS, etcetera. How do you ensure that it seems
- 45:37like it's a critical component to making all this stuff work
- 45:40together, sort of this underlying data logic and, and
- 45:43connectivity between all of this stuff, because that's often
- 45:46where the fragmentation is within, within a brand and they
- 45:48have problems kind of getting their data house in order.
- 45:51How do you guys, you know, how are you seeing this in like come
- 45:55to life and with clients? Are you seeing some of these
- 45:57prototypes? Get out and and test it in
- 46:00market. Yeah.
- 46:00Look, we think that, yeah, the average enterprise right now is
- 46:03licensing somewhere between 80 to 120 different platforms to
- 46:06build out their martech and ad tech stack.
- 46:08Each one of those platforms has a private pool of data in there
- 46:11for the enterprise for its various function, whether it's
- 46:14for activation or measurement or reporting or enrichment, you
- 46:19know, profile generation. So, you know, those platforms
- 46:22have to be connected and what the ICL capability is.
- 46:25Essentially, it's a large switchboard to connect, plug all
- 46:27your systems into and all their APIs and then, you know, allow
- 46:31that to be able to run the various use cases that a human
- 46:34would run systems. And instead of that, you'd be
- 46:37able to have an LLM or an agent that you develop, you know, run
- 46:40your automation for, you know, audience enrichment, audience
- 46:44activation, porting, communization, all the various
- 46:46functions that an agency, a brand, a publisher or platform
- 46:49are doing today from a human standpoint.
- 46:52Now it can be done with an LLM by plugging into the ICL.
- 46:56That's awesome Bob, love it. Thanks for the explanation and
- 46:59enjoy the rest of EDWY. Absolutely, thanks for having
- 47:01me. Hi, this is Rio Loniaker.
- 47:03You're my key takeaways from what I've seen today as I've
- 47:06been walking around and spending time with people.
- 47:08Number one, really busy event. I think I read earlier today,
- 47:1120,000 attendees overall. It's a big number and you can
- 47:15tell I got here this morning and the line wrapped quite literally
- 47:18around the building and it took me about half hour to get in.
- 47:22And luckily I saw Kenny Paul from Chalice.
- 47:25I was walking down and then we ran to Kendra Barnett.
- 47:27And so it's good to run into two friendly faces there.
- 47:30And we ended up getting in two other key themes.
- 47:33Number one was a lot of talk today about like AI and a data
- 47:38governance and responsibility, especially copier.
- 47:41I think it's important as we look at as AI evolves, there's
- 47:43more agencies and marketers adopt it.
- 47:46What AI frame, what frameworks are we using to understand is
- 47:48data consented? Should we be using it?
- 47:50Is it if we're if we're using models as a data copyrighted?
- 47:55A lot of talk about that today. And the other talk, I thought
- 47:57this is actually the biggest one of the day was about creators
- 48:00and a creator economy. Believe it or not, there was 150
- 48:03different creators presenting today during Advent.
- 48:07So pretty cool. Just show us how far this has
- 48:09come, the value, the premium value organizations are putting
- 48:13on the creator economy and the fact that creators are creating
- 48:17amazing content. So I think it really speaks
- 48:19volumes that there was such an emphasis on it.
- 48:21And I think we're going to see more and more of that.
- 48:23Those are my key themes for today.
- 48:24Thanks very much for listening in.