← Always Be Testing
Always Be Testing
#72 Optimizing Publisher Earnings with AI | Cornelius Frey, CEO, NucleusLinks
00:34:58
Loading summary
Transcript115 lines
- [00:07]
A
So, yeah, we're just gonna just keep it super chill and hang out on the couch. We've got a little video here. Mike's teed it up. It's already running, which is freaking awesome.
- [00:16]
B
The whole time.
- [00:18]
A
Smart.
- [00:18]
B
Nice to see you again.
- [00:19]
A
Yeah, nice to see you. Everybody out there, different drinks on the. Oh, good call. Yeah, please.
- [00:26]
B
I can give you a top up anyway.
- [00:28]
A
Juice.
- [00:28]
B
Cheers. One last night.
- [00:30]
A
That's okay.
- [00:31]
B
It's a nice little contrast to last night.
- [00:34]
A
Yes. You gotta mix in the health with Vegas. Where, where are you coming in from?
- [00:40]
B
Berlin, Germany.
- [00:41]
A
Okay.
- [00:42]
B
Beautiful Berlin.
- [00:43]
A
Yeah.
- [00:43]
B
We live with a family and have a growing team base. Amazing European. Yeah. Conquering ambitions all out of Berlin.
- [00:52]
A
Okay, that's great. Yeah, that's great. How many folks are there?
- [00:55]
B
There are now, I think closing in on 50. There were 30. So I found a little company called Opinionary, maybe you know, the little needle, the poll on Times, NBC and so on. And Affinity out of Mumbai bought it one and a half years ago.
- [01:10]
A
Okay.
- [01:11]
B
And since then everybody stayed on board and we're growing.
- [01:14]
A
Congratulations. Congratulations. Tell folks what, what you guys do with the Paneri and with Affinity. What's the, what's the product, what's the service for those that don't know.
- [01:23]
B
Yeah. So Affinity is a holding company, 500 people closing on 100 billion revenue out of a bootstrapped little team of five, 17 years ago in Mumbai that's now running, you know, in total, seven business units, number of ad networks all the way from branding and impact like McAnvas VV, working with OEMs, browser players and a suite of SaaS products which I happen to run. One of them is opinary helping publishers drive first party data conversions by engaging their users more meaningfully on site. And the other one is Nucleus Links, which is why we're here at asw, which is basically the AI driven operating system in the wild world of affiliate operations. Because that is publishers don't even know how much money they're bleeding in manual operations, how much painful work is going on unnecessarily every day, and how much opportunity there is once you bring in the power of AI. So that's what we're trying to bring.
- [02:26]
A
That's amazing. And so a publisher is a situation where. Describe what's a good use case for a publisher. That's like, okay, I kind of need to assess where I'm at in terms of addressing the links on my site, leveraging AI routing to the right destination. Can you share an example of that? Or how does that Work.
- [02:46]
B
Yeah. So let's imagine you're a publisher and publisher can be, you can be in the coupon space, you can be buy now, pay later, you can be a content publisher and you have a healthy and growing affiliate revenue operation. And it's, you know, there's money there and it makes sense and everybody's happy. Kind of the situation that these publishers might not know that they're in is a losing money because there's so many more budgets available. Likely their setup now is they've got a merchant and either got a direct partnership or they've got a primary network. Set it, forget it, sail off ever happily into the sunset. What most players don't know is there's so many, not just CPA and CPS networks, but even CPC networks and all kind of other budgets available. Let's even take someone like Walmart. There's 25 different networks available that you should be plugging into, but no one in their right mind can, for every single click or every minute switch. Oh, where's there a deal available now for which region, for which user? So there's something, we've built a tool that allows you on a per click basis to get the highest yield, the highest earning per click by automatically for that individual click, picking the best highest earning network.
- [03:59]
A
That's awesome.
- [04:00]
B
Second part is, yeah, there's broken links, stock outs, deals, and you find out maybe after an hour, maybe after a day. And that's kind of the way it's been for 15 years. The fact that this can lose you up to 15 or 20% of your revenue is lost on, you know, most because they don't know how much you can get by identifying those broken links in real time. Real. Another pain point is I have to log into every network and Amazon separately to get my numbers together to get my reporting and somehow cobbled it together on my end as opposed to having unified reporting across all earnings streams in one place. So those kind of things. Basically the metaphor we like to use is an affiliate. Someone in affiliate operations and a publisher is like a pilot and they're doing a great job. They're amazing pilots. But somehow we've gotten used to them having to build their own plane and being ground controlled and running the airport at the same time.
- [04:58]
A
Yeah, it doesn't make much sense. No.
- [05:01]
B
Better behind it.
- [05:02]
A
Yeah. And their instrumentation is not as accurate and efficient as it could be from what you're saying as well. When someone's running that role as a publisher and they're having to do all those things, do you have some estimates of how long that takes. Is that a full time job? Is that two full time jobs? And then do you have a sense of like, hey, this is what we were able to save you in terms of just labor and time?
- [05:28]
B
So it is. If you talk to people in those roles, they will happily tell you they are smart, hungry and hardworking people who will happily tell you that up to two thirds of their day are spent manually creating links, fixing links, finding new deals and doing things that frankly, they could be spending so much better by optimizing relationships with a few direct partnerships, doing meaningful business work. So easily, two thirds of time is spent on menial, annoying work that these incredibly capable operators can really use to do better things.
- [06:06]
A
Yeah, that makes sense. I love that. And there's so many, there's so many monetization options and platforms and advertisers. It's just, it's immense. When you mentioned the AI piece and the routing and conversion piece is interesting to me because obviously you come from some great A B testing background. We've actually had an AB testing conversion optimization service within RBL. Right now we're obviously more focused on affiliate Influencer, but it's kind of near and dear to our heart to think about how do you route people to the most ideal destination? How do you think about the user experience? How do you a B test that? And then how do you think about AI and enabling that? Can you maybe double click a little bit on the AI element? Because obviously that's a hot, just a little bit of a hot topic for everybody.
- [06:57]
B
Yeah, me too. So a B testing is where all of this starts out. A B testing would be say, hey, we used to send traffic for Adidas to Awin and now let's try out B. Maybe Sovereign has a better deal. That'd be a B testing. What we do is different. So we do constantly because we've got the scale every minute send basically test clicks to all those networks to inform our learning algorithm with constant data input for publisher A merchant B region C type of user D. At this time, where might we be getting the highest earnings? So we send, basically we invest up to half a percent of traffic in gathering data in real time. By sending test traffic everywhere, it might not be winning money at that time, but that informs the AI both in real time. Okay, no, here's a new deal that didn't exist when the link was set two weeks ago. But now there's an amazing deal to be had. But also it informs a learning algorithm that's able to predict where, at what time, in what region, what merchant higher VPCs are available. And they can validate at the end of having sent the traffic there or invalidate. So it's based on a B testing. It's constant A to Z testing that informs a learning algorithm.
- [08:15]
A
That's amazing. And if you're a publisher and you're going through that learning phase and obviously you have a certain amount of volume which dictates that learning curve and that effectiveness of the conversion, how long do you think it takes to kind of get through that learning phase? Or if you're a publisher, what are some of the things that you observe from their wins there with getting through that algorithm learning phase?
- [08:38]
B
So when we say win for us, it's a win to lift publisher earnings on those accounts that we feed into the Nucleus links engine, not by like 2%, although that's nice. But a win for us is 25% plus, wow, lift in earnings from those accounts. So there's so much money being left on the table. And to get to that, get to those 25 and sometimes have a big search shopping player who just came on board, they're getting 102% uplift. Forbes seeing similar numbers, a huge buy now, pay later. So this goes across the board. You ask, what does it take to get there? We are fortunate that we build on a big foundation of learnings. There's couple of things. One is onboarding. We create the links automatically because we got APIs into the network. So that's by now having built this amazer, you know, takes a day or so. You need a certain scale which, like, let's say you have a thousand merchants. We'll need at least 50 clicks per day per merchant to make it meaningful, which, I mean, isn't a crazy number. That's already sufficient to get there. And then there's one more part which every affiliate operator will know. Signing up to new networks can be a pain because it's all legal, you know, razzmatazz included. And we did two steps on that one to ra that one up. Because that itself can take somebody two months to say, hey, here's, you know, we have. We say, send us your merchants. We tell you which networks you should be working with. Either say, cool, we go ahead, we quickly sign up, or we've built what we call network plus an engine. We're not a sub network ourselves that builds the links to the networks on the publisher's behalf directly. Like a sandbox for us for them to tap into, let's say the not 50, but 10 networks that given their merchant mix we know are going to drive those 25% plus earnings and all that included means we can be up and running producing those kind of lifts within two to three days. Wow. And do all of that in a 30 day completely risk free money free trial. Because we're happy to say we put our money where, where our claims are.
- [10:46]
A
Yeah, that's amazing. And then are they are. When you do get to monetize and you do start the charge and to commercialize it for yourself and your business growth, how do you typically structure that?
- [10:58]
B
So we are a software as a service company, we're not a network ourselves. How we make sure our software slab based basically cost per click in the end how we make sure that works for everybody. You say we've got a iron ceiling in there which means we only ever take money from the uplift we create and never more than 30% of the new revenue that we've created.
- [11:25]
A
That's amazing.
- [11:25]
B
I mean beware of anybody who walks into any room and says I've got guaranteed free profit for you. Yeah, that sounds like a sketchy proposition. But I think with this model we're, we're close and on pretty firm ground.
- [11:38]
A
And the pricing is backing up your bold claim. Which is, which is nice.
- [11:43]
B
And I think it's also something you see. I mean you asked about AI. It's interesting how pricing really in the whole tech industry is changing, right. There used to be just flat SaaS fees and we're really seeing that. The first step used to be usage based fees and now more and more. I think as agentic AI is coming to the place, outcome based fees are really going to grow. So there's a whole lot of businesses and VCs out there that you say flat SaaS pricing forever. I think a lot of that is changing very rapidly.
- [12:13]
A
We've talked a lot about usage based in our experience and our just usage as a company and in guests on the pod. And yeah the outcome based thing hasn't been talked about as much. So I like that you're bringing that and thinking about agents and the effect of agentic AI. What are some AI and agent based things that you have on the horizon that you are able to share around.
- [12:38]
B
The business around Nucleus links One big part that's already there which existed before agentic AI was a big thing. But broken links are an insane issue. 404 stockouts we briefly touched on that do two things. First of all, immediately upon recognizing hey we got an earnings issue here on Sending traffic to that network, even though technically this should be paying great rewards, it's just not happening. We route the traffic another way automatically. And the agentic layer also alerts the publisher because this is something that needs to be fixed immediately. It doesn't help if somebody is asleep at 3am and finds out the next morning. But it's also transparency that what we're finding publishers really going for is, yes, they're super happy to hand off all that manual, menial, annoying work, but very rightly, they want to know what's going on. They want to know which networks are winning out. They want to know where their earnings are coming from. They want to know how merchants compete. They want to know where broken links are taking place. So I think the magic that this kind of layer of AI is allowing is both independent action, but not independent action under this huge cloak. And you say, I don't know, it's a magic machine, I better not touch it. No, people want to know and should know what's happening in order to trust those systems. And I think that's how you build adoption and trust.
- [14:04]
A
I love that. And when the publisher is kind of thinking through these pain points, deciding what they can do, who within the organization or what teams are you finding are kind of most responsive to the conversation and kind of getting things implemented, getting the test started.
- [14:21]
B
Basically, it's two kinds of roles across the coupon, the content and the BNPL space. They sometimes have different names, but there's somebody in charge of e commerce revenue CROs, they're just in charge of bringing the money in and going there. Hey, without risk, you know, we can lift your affiliate earnings. That's something, that's a pretty attractive value proposition there. But then also no less important are the folks at the forefront of the operational business. So somebody in charge of affiliate operations, the people building those links, fixing those links, tracking those earnings, logging into every single network every day to find out how things are going. Those are the people who use this technology every day. So yes, the people in charge of every get excited and are. But we also try to make the life of the operator easier. So it's those two parts. The roles have different names.
- [15:15]
A
I love that. It's been awesome being here in Vegas for Affiliate Summit. I'm sure you're having all kinds of good conversations and we're kind of like trend spotting and seeing what's out there. And some of the stuff is kind of same old, but some of the things are new and interesting. Is, is there anything that has kind of Been an aha moment for you during this, this conference in Vegas or anything that to note or share.
- [15:38]
B
Yeah. So one true part about our time at ASW and these places, I haven't been seeing any big talks, you know, on stage or anything. It's for us, it's our team for, you know, my first aha moment is ever. Our team is spread across 10 hours worth of time zones. So incredible. That's always just an aha moment. It's incredible.
- [15:58]
A
We're close, but you guys are a lot more of that.
- [16:01]
B
So just seeing that team and maybe quick double click on that is what we pride ourselves on is a team that's incredibly hungry and loyal. So we actually. To join the team, you don't need to know what a CPA to Z is or whatever. We hire for aptitude, brain and hunger, often straight out of college or some other institution and then people stay with us. That makes for incredible energy. Anyway, that's internal. Yeah. Spending time with clients here and seeing the openness towards trying things. I think you always have a fine pendulum swinging, especially in this industry, between, yeah, I should be doing all those things, but it's just there's so much on my plate and our product team is busy and yeah, you know, maybe we look at new stuff like AI, but it won't be before, you know, H2 or something.
- [16:51]
A
Yeah.
- [16:51]
B
And there's times when people are ready to go out, try stuff, I feel. And you tell me what might be behind that at the moment. People are very open and ready to try and work and find those next solutions and the things that help them. So just this seems to be a very open moment in the industry right now.
- [17:10]
A
That's amazing. Well, I think you've got a great offer, a great story, you've got horsepower, for lack of a better term, and the technology and it's a, you know, pricing model that's pretty darn compelling. You're saying, hey, there's not, not a lot of risk here. So I could see why. And you're. It's also at a moment when these publishers are not exactly. No one's job is easy if you're an advertiser, if you're an agency, if you're a publisher. Publishers are. Some of them are getting squeezed by Google, some of them are getting hit with AI from a different way. So it's. I think it's an important reminder to like, okay, they can start leveraging AI to their benefit to make their monetization more effective. So for the traffic that does come into Them they have the highest confidence they've ever had hopefully in saying hey, we're routing it to the right retailers, we've got links working and functioning, we're not missing out on the opportunities as a publisher. And then everybody benefits from that ecosystem because the publishers are really the heart of this ecosystem I've often said and talk about and you see it firsthand, they're your customers.
- [18:11]
B
And maybe you touched on something interesting which obviously moving the industry. The Google changes to search results, the rankings for publishers and their coupons businesses specifically. One thing that's really stuck out to me as we very tiny at the morning moment because those changes came in continental Europe with a delay just two weeks ago there was this massive bullwhip changing half the industry and there's lots of European publishers here in Las Vegas this week as well. And I think a couple of things are interesting. It's the pragmatism with which that's being dealt with. So you have the publishers whose traffic got massively hurt but they're already adapting and they have been preparing and building content offers that are not to be disadvantaged from brain engines because genuine related content. You've got interesting industry players like C4 Rocks and Web Gear is helping them with it. We saw tag 24 out of Europe growing in the US on a very content commerce driven model unencumbered by those Google changes. And you've got the sort of the pure play non content commerce players and the coupon sites who are seeing a real revival in their business and capitalizing on it. So I think sometimes you're at inflection points in the NC's where everybody's just sort of hand wringing and saying we're so helpless. Wish of Powell is what are we to do? I'm not thinking this is a moment like that. I think people are finding solutions. People are growing less dependent, you know, on the big walled gardens and that's for the better.
- [19:42]
A
Yeah, building up first party where they can, you know.
- [19:45]
B
Absolutely.
- [19:45]
A
Email. Building up kind of a genuine authentic social following and something that's really exciting for consumers. I mean that's obviously a thing we're starting to see across the board. I'm starting to kind of get on that soapbox if you will to talk about trust and how important is because if you have that you're kind of getting to the core of the Google Eat framework which is driving a lot of those algorithmic changes and you have something that's more defensible which is exciting and you don't have to I think there is a lot of opportunity we did see over the last 12 months and I've talked about this a bit. Some of the content sites took a hit.
- [20:24]
B
Yeah.
- [20:24]
A
But in aggregate, like clients, all publisher types, like, we're seeing growth going up into the right portfolio growth, revenue growth. And I think you're right. There's winners in the space that are creating that genuine connection and doing the things that are needed to kind of win in a time when, you know, there's. Yeah, there's some headwinds, there's some challenges, but it's not insurmountable.
- [20:45]
B
Yeah, absolutely.
- [20:47]
A
That's amazing. That's amazing. What's next on the horizon for you guys? I mean, you got some exciting things going. I will say having our team here, I think I picked up on what you were saying. Like, it almost feels like a mini, off site, mini reunion to be here in Vegas with a lot of good people, you know, and those relationships, as cliche as it sounds, are so important. Not just, you know, with your team, but partners like you guys.
- [21:08]
B
Yeah.
- [21:09]
A
Advertisers and networks of people that you want to work with. How have you found it? Just being able to connect with people, you know, coming, coming. Quite a ways to come.
- [21:18]
B
Quite a way. Yeah. Amazing. And if you ask what's next on the horizon after these three days, I think first off, very good, long sleep. What's next on the. There's one more hat I'm wearing. So we from the opinion side, you know, joined the leadership team of Affinity Global. There's one more hat I'm wearing which is on the M and A and capital market side.
- [21:37]
A
Amazing.
- [21:38]
B
And that's brought with it a number of really interesting discussions. You know, as a, mostly as a global company, but with a base in India, we are eyeing the public markets there as well. India has a roaring stock and IPO market which is something not every, every avtech business around the world has access to all the time. So that's a long term but very interesting prospect for us. And we are, you know, still growing at 30% plus organically, but also looking at some acquisitions. So something that here at ASW is always. I mean, there's amazing businesses growing, profitable, tech driven in our industry that are looking, that have been added for maybe 15, 20 years. The next step isn't obvious. Maybe the founders are also, you know, finding joy in other things in life or want that next S curve of growth which is even greater. So that's conversations that we're having that I'm really enjoying because it allows you to connect with some of the real, you know, indomitable builders in this industry. And it's just awesome.
- [22:34]
A
That's amazing. I mean, you've built, you've been involved with tech, you've built in Europe, you've taken it globally, you know, to kind of interact with such an exciting market that is India, I think is really, I mean, you step back and look at it and there's so many predictors and positive trends coming out of that region. And collaboration with entrepreneurs in the US and Europe, I mean, it's, it's pretty fantastic. Is there anything that you've observed from it or just you've learned from that market over the years that you found interesting or that you think is worth sharing with the audience?
- [23:07]
B
Yeah, there's a number of things. I think just an economy growing at close to 6% GDP. It's obviously like real. It's not a type that lifts all the boats, but lifts a lot of boats. There's just so much energy and hunger and things happening every corner. You feel it. Especially my contrast is Germany, which has been contracting for third quarter in a row. It's a different vibe out there. So part one that helps everybody on a more micro level, the inventiveness and hunger to build globally meaningful businesses, often in a very resourceful way. So a lot of the giant tech firms out of India been bootstrapped. That makes for very nimble businesses, very entrepreneurially driven. You know, you don't just drop $40 million in VC cash from a, you know, seed round on some idea. You build what matters.
- [23:59]
A
That's great.
- [24:00]
B
Maybe a third one also contrasting like product and engineering. There's a Silicon Valley school of product management which has created a whole industry, you know, out of product management. There's a lot of merit to it. I have found it very interesting sort of to witness the much more pragmatic way. There's no long product development, you know, feature roadmap plans or something. You work on client feedback. You work build it and they will come. You build random stuff, but you get a deep understanding of a market of a client and then you go and build it and bring it to market rather than, you know, creating a million documents. I found that very liberating as a contrast to how also we at Opinionary have been building product for a while. And yeah, I think the fact that what's changing is in many economies, but India used to be an incredible place for analytical horsepower, for great huge consulting and outsourcing houses. And I don't Think many people have realized that India is going to be another one of those places going to be building consumer and enterprise products for the world market. And that's already happening. So I think watch this space.
- [25:12]
A
I love it. The M and A thing is exciting, especially as someone who's built company and been acquired. And it's awesome to see. I think affiliate is what we call it kind of mature and get better and more players involved. And are there things that if you're excited about to consider as an M and A opportunity, obviously builders, obviously tech, I'm assuming, but I would love to know more about what are some of those criteria? What are some of those things that kind of how does something get up on the radar for you all to go, oh, this is worth exploring, this is worth talking to?
- [25:47]
B
Yeah. So I think the way I look at it in our corporate setup, it's three layers. One, the numbers need to be right. You want in our case a profitable, meaningful and growing business. That's table stakes. Second one is people. The culture shines through and it shines through from the people who started the business and who run the business. It needs to be a cultural fit of people. I keep saying hungry, but that's really our spirit, right? Being hungry to build that next iteration of our industry rather than just randomly taking care of some ARP opportunity here. @ the core, we want builders. And the third one is solving a meaningful problem problem in a meaningful market. So there needs to be a tech component for us particularly two things are important at the moment, which is exposure to the US market where we're growing the fastest already our largest market. But yeah, you know, getting down even more here and work in the search space. Access to search budgets, work on that side because we see still huge potential down there as well.
- [26:52]
A
Very cool. When you say work in the search space, can you double click on what that means for this context?
- [26:58]
B
Yeah, basically making search budgets and search traffic available and dealing with it outside just, you know, the results page, but using it in a wider ecosystem and finding intent intelligently in spaces that might not historically been label search, but in terms of intent, purpose and performance are very much search adjacent. We think there's, you know, we see ourselves a bit. The whole affinity philosophy is truffle pigs go into spaces that are underexplored. One of those is working with incredible handsome manufacturers and browser players who haven't been monetizing the enormous traffic as well as Google has been and building the unique tech to help them do it. Another one is finding what's Maybe now industry niches and turning them into sizable parts of the industry rather than going into RTB where everybody is, you know, batting the head in.
- [27:51]
A
Yeah.
- [27:52]
B
Go find those new spaces.
- [27:53]
A
Arbitraged out pretty well.
- [27:54]
B
Pretty much, yeah.
- [27:56]
A
And paid. And organic. Or is it primarily organic? Search. When you. When you say search. I'm just trying to.
- [28:01]
B
There could be it. Yeah.
- [28:03]
A
It feels like there's. There's an opportunity for both. From what I'm here.
- [28:06]
B
Absolutely.
- [28:07]
A
That's. That's fantastic. Maybe coming down the home stretch here, maybe some. Some quick hits. Anything you want to tell the audience a little bit about Cornelius that maybe they don't know about you or hobbies, life, things that you want to share?
- [28:21]
B
Yeah, sure. So there's one thing that involuntarily few people walk out of a conversation with me without finding out. I'm a terrible but very excitable amateur musician.
- [28:33]
A
Oh, nice.
- [28:34]
B
Sadly, you'll probably never hear the name of my band.
- [28:39]
A
Now you have to tell us.
- [28:41]
B
You don't have to linger. Good. No. Watch this.
- [28:44]
A
Bass.
- [28:46]
B
No. Full truth is I. I'm. I'm a amateur bass guitarist playing with incredible musicians and they. They endure me. We even have an affinity company band. So we had a company offside in Goa and. And got together. So. Yeah, if there's, you know, in 20 years time, a second layer to my career.
- [29:05]
A
Okay.
- [29:05]
B
We like to tell ourselves we're sort of in psychedelic rock. Probably more dad rock. Psychedelic fun is what we say. Dad rock is what it is. That's one thing. Raising three beautiful daughters.
- [29:19]
A
Amazing.
- [29:20]
B
Speaking of India, took the whole family to India. So traveling with kids. My one part is kids will do a lot of things. Typically just the parents are scared. Kids are fine. Yeah.
- [29:31]
A
Exactly. How old are your daughters?
- [29:33]
B
They are five, seven and ten.
- [29:34]
A
Oh my gosh. That's amazing. Amazing. I got four year old daughter and two year old boy.
- [29:38]
B
Oh, wow.
- [29:39]
A
That's. That's. Those are exciting ages. Congratulations. I love it.
- [29:43]
B
Thank you.
- [29:43]
A
I love it. Anything resource, book, guide, video that's like just kind of caught your attention that you want to share. That's a must read or informing kind of some of the ways you're thinking about things.
- [29:55]
B
Given I'm 20% in. This is a bit preposterous to tell everybody to read it, but I've been listening to the audio version of Yuval, Noah Harari's new book, Nexus on information System. Basically the nature of information and what it means for the AI age. And one part he says basically information doesn't equal Truth automatically. Information's purpose isn't to convey truth. Information's purpose is to connect us. And that's fine. No need to pass judgment on that. Let's be very mindful of the fact that more information doesn't automatically mean more truth and what that means for how we design the information systems of the future. And social. The barrage of content, you know that did all things with our societies was just the first layer. The next part is coming and I don't think you need to be a, you know, an AI doomer to say there's real potential to shape this and it's going to be interesting times.
- [30:50]
A
Love it.
- [30:50]
B
I don't have the conclusion again. I'm just couple of hours in. Seems like an exciting.
- [30:55]
A
I'm all in on that book now that you got me hooked. Hooked on it. So hopefully other people will be finding interesting for people that are kind of. Okay, I gotta dive into this AI thing. I need to learn more. Is there anything that you kind of guide people to to say hey this is the basics. Get started with like get your up your AI game a little bit. Is there anything that you kind of guide people on?
- [31:15]
B
I found two parts been interesting to me. One is, is in the back issues but the Economist ran like an eight week primer serious you know on LLMs foundation models, application agenda, AI and a lot of those things you just find out with the incomparable skill of drilling down things to the basic. You think you know all of it. We know all the buzzwords.
- [31:37]
A
Yeah.
- [31:38]
B
But putting it together in a way that my test is always can I explain it to my 10 year old daughter? And that's always when you fall flat and you know, yeah, I can have a one hour discussion at ASW here in Vegas on AI. Can I explain that meaningfully to my daughter? No. Means I probably don't actually get it as much as I think that is a great resource to get into and on a weekly basis. Somebody who reads very widely and condenses that reading is Benedict Evans, the old Andreess Norowitz partner. I just find his reading recommendations on the topic give me more than I have time for in a typical week to keep the ball rolling.
- [32:15]
A
That's awesome. Maybe kind of something tactical as we kind of come down the home stretch. You obviously are seeing so many publishers look at okay, how do we guide people to the right destination? How do we. And that's going to be so different across so many different things. So probably hard to distill down. But is there maybe a couple of Tips that you can share from. Because so much of the POD is all about learnings. Anything you can share to say, okay, you should route it this way in this case or you should be thinking about this at least to make sure the user gets a better experience. It might be counter to what you guys do because it's so test experiment data, AI driven. But is there any, like, tips for the audience to think about if you're a publisher? Like, yeah, what should be sending to and why and when mobile, you know, like love to know what that might be as a guide.
- [33:05]
B
So I think we have this age old drive to call for the next big thing in publishing. At first it's social, then it's video, then it's audio, whatever, and then it's web three. And sometimes these are mirages. So I'm always a little careful. This is the one way you do it. Now what we see in our publisher base with players is more you pick a way to do it and stick to it and build a user understanding for what you do. This can be super different. This can be. On the one hand, we just mentioned the Economist. The Economist personalizes nothing. The Economist says, here is what you need to know in a given week. It's our point of view and we bring it to you and you know what you've got. Let's take Microsoft, on the other hand. Microsoft is on a journey also on the consumer products to a huge individualization. I don't think either of those approaches are right or wrong. Both of those are doing an amazing job in picking one, in making very clear what their path is and then executing on it.
- [34:04]
A
I love that.
- [34:05]
B
And I think that's what every publisher need to do. Don't be stuck in the middle with an undifferentiated, unclear thing. Be clear what you do. Stand for it, tell the story and do it.
- [34:14]
A
I love it. What an amazing way to wrap it up. And it's. It's all about that. If you're doing that, there's an authenticity and a truth there that you referenced earlier, which I think comes up a lot for us in these conversations. So amazing. Here we are in Vegas. It's nice and warm. We're going to head back to the show, but thanks everybody for joining us. And Cornelius, you nailed it. It's been a pleasure to get to know you more and learn more about what you're doing.
- [34:37]
B
Thanks for joining.
- [34:39]
A
Thanks, everybody.
- [34:40]
B
Thanks. Hi.
- [34:41]
A
Awesome, man.