[00:00:00] Luke: From privacy concerns to limitless potential, AI is rapidly impacting our evolving society. In this new season of the Brave Technologist podcast, we’re demystifying artificial intelligence, challenging the status quo, and empowering everyday people to embrace the digital revolution. I’m your host, Luke Malks, VP of Business Operations at Brave Software, makers of the privacy respecting Brave browser and search engine, now powering AI with the Brave Search API.

[00:00:29] You’re listening to a new episode of The Brave Technologist, and this one features J. P. Schmitz, who’s the SVP of ads at Brave and a Brave board observer. He’s been building tech companies since 1995, including Klix, which became Brave Search. In this episode, we discussed the evolving relationship between browser, search engine, and the internet, and how advertising has shaped their development.

[00:00:49] Balancing privacy and efficacy when leveraging LLMs and ways we can empower users to make simple but impactful changes to protect their privacy.

[00:00:57] Owning their data and shaping their own future [00:01:00] on the internet. now for this week’s episode of the brave technologist, JP, welcome to the brave technologist. How are you doing today?

[00:01:11] Jean Paul: I’m very good. Very well.

[00:01:12] Luke: Let’s kind of set the table here a little bit. Why don’t you tell the audience a little bit about what work you’re doing at brave and kind of why it’s really important at this time how you see it affecting the company in general and the broader world.

[00:01:25] Jean Paul: Well, at the moment I’m taking care of ads, technology, product and sales. And I think the ultimate goal is, very simple, right? Is to make Brave sustainable as a company, which basically means profitable in all terms to do so, you know, providing good value for advertisers and respecting, the values of the company, towards the users.

[00:01:44] So a simple problem, difficult solutions, I guess.

[00:01:47] Luke: Yeah. And let’s dig into that a little bit too. I mean, cause it’s one of those things where. Everybody uses search, but not very many people really know much about it under the hood or, kind of the evolution of it, walk us through a little bit of the evolution [00:02:00] of search engines and how important advertising has been with search and the evolution of that.

[00:02:05] Jean Paul: So why do we need search engines, right? We need search engine because in the mid nineties, the web emerged as an open system, right? Before that, there was online, right? You had compuServe, you had AOL. So there were closed systems, which did not seem to need much of a search. Because if you wanted to go to the weather app, you would type weather and you would go to the app.

[00:02:24] And if you wanted to go sports code, there was only one. It’s only when the web became open that suddenly you had like 1000 of everything. And there was a need to first find them because you didn’t know they existed. You weren’t sure they were there, and then more and more people started to build in expectations as to what was out there, etcetera.

[00:02:43] And then it was a matter of ranking. So the first search engines, AltaVista, Excite, You know, there were a bunch of, them at a time we’re not doing that much ranking. It was really about finding, and it was just like keyword matching. Let’s just try to find out what is out there. What transpired after a [00:03:00] while is that, and this is kind of a segue to advertising is that people found out that when they were on top of that page, traffic was coming in.

[00:03:08] Right. And so they started an early version of search engine optimization, which at that time was about, you know, writing a white keywords on a white background. So the user wouldn’t see it. And somehow the search engine would be fooled and put you on top the result page. It’s only really around the time where Google came out and another company called ink told me that people figured out that one of the best ways to rank was to rank via the.

[00:03:35] Text that one page would use to describe another page, right? So there was a real difference between you saying on your page, I’m the best restaurant in new york city and the new york times dot com saying this link is the best restaurant in new york city. And that was really the secret behind google, right?

[00:03:56] To create this index that also included the text that [00:04:00] other people were using to describe your page. So Google and Inktomi, Inktomi became Yahoo search, became, you know, extremely popular because of the highest quality that they were delivering. SEO was still a thing. And Google figured out very quickly, well, not only Google, because there was a company called goto.

[00:04:16] com, which became Overture, which was bought by Yahoo, ultimately invented the concept of keyword advertising, which is basically, you know, you just buy the keywords that you think are relevant for your page. It turns out that about 18 percent of the queries are commercial. So people are looking to buy something.

[00:04:35] And for these queries, actually, the paid answers actually are better than the organic one. I can give you a bunch of examples if you’re interested in why that is. But essentially people started to buy keywords and the price became so high that it became. Uneconomical for bad actors to like, I mean, there’s exceptions to that, of course, but in general, the valuable keywords are too expensive [00:05:00] for scammers to outbid the real deal.

[00:05:03] Well, this is why search advertising is such a killer business model because people tell the search engine what they want. And in 18 percent of the cases, so the answer is commercial and the search engine gets money for that. And the other 82 percent they can give for free.

[00:05:18] Luke: And that’s really going for what the user’s intending, right?

[00:05:21] Jean Paul: Yeah. And if a user is not interested in buying something, but indicate clearly in the query that they just want to get information, then it may become uneconomical for, you know, retailers to buy that keyword, right? So there’s a natural balance that comes in. And because we are brave, we care about privacy and the search engine business is remarkably private because there’s no need to know who the person is.

[00:05:43] It’s just, you need to know What they just typed, right? They tell you what they want. So, there’s no need to say, Oh, you know, I’m not gonna give you an app for that lipstick because you know, the person looking for it is a man that doesn’t make sense in search advertising is someone who’s looking for lipstick.

[00:05:58] You give them lipstick advertising, [00:06:00] right? So a lot of the problems with privacy actually better in search because it’s not targeted at who you are. It’s targeted about what you just said. You want it.

[00:06:10] Luke: Right, right. No, that makes, it makes a lot of sense. When we look at this world of search and kind of how brave comes in to play, I mean, you’ve got such a few players in this indexing space where they’re, they actually own the index, right?

[00:06:22] Like how is brave able to kind of come in with an independent index this late in the game? Like maybe we can give a little background around how brave search came to brave, just to kind of get people familiar, because it seems like almost one of those. near impossible industries. Do you like kind of break into at this point?

[00:06:40] Jean Paul: So just to give a little bit of timeline, you know, Brave was founded in 2008, 15, but seven years prior to that, I had the, I wouldn’t call it a crazy idea. I had this crazy goal. And I can tell you more about why I had this crazy goal of trying to rebuild the search engine from scratch. I mean, there were a number of [00:07:00] reasons.

[00:07:00] one was boredom. I had pretty much done everything that that could be done in terms of e commerce, content, publishing tools. I had built a bunch of websites in Germany at that time, and I simply became interested, and I had done a project with Google a few years before. So I kind of knew this world of browsers and search engine, and it struck me as being infinitely more interesting than just building websites.

[00:07:27] And second. I was also relatively clear in my mind that I did not want to go work for Bing or Google or anything like that. Right, so, or Firefox at that time. I was asked once by the French government to create a French search engine for the French people. I’m not French, I just happen to speak French, so they’re probably mistaking me for I gave them really solid argument at that time.

[00:07:49] Why that wasn’t possible. I said, look, it’s going to be too expensive in infrastructure. It’s going to be too difficult to find data to build the first version of the search [00:08:00] engine. And it’s going to be very difficult to find people because Google basically will overpay them and you will not find them.

[00:08:07] And I remember flying back to Munich from Paris thinking, yeah, all good arguments, but it’s kind of bullshit because I kind of know people that would not work for Google, but would actually want to build a search engine. Then. All of these are still working at Brave today, infrastructure, you know, I knew that Amazon was working on something called AWS, that you could rent infrastructure instead of buying it.

[00:08:30] And so the problem really got reduced to data, like how can you crawl the web, I use a word, It’s not very precise, but let’s say crawl the web without getting fooled by all the SEO stuff. And so that was the, the first problem we kind of worked on at clicks. And the second problem was distribution.

[00:08:46] And that was a lot of, that was a difficult thing. And another problem that was interesting is how do you get access to users when you only have Half or a third of a search engine. Imagine the situation we had in 2014. I [00:09:00] want to say where we could answer, let’s say, 30 to 40 percent of the questions perfectly, but we didn’t know the answer of the other ones.

[00:09:08] Luke: Right.

[00:09:08] Jean Paul: How do you create a search engine that has user interaction? Because you’re going to need user interaction at one point to get better. And. And that’s a problem we solved by having a drop down at that time. Firefox extension that would give you the answer right away if we knew it. And if we didn’t know it, they would just go to the default, which was mostly Google and that extension turned out to be insanely successful.

[00:09:31] So we, we start to get this flywheel running and in fact, Firefox invested in clicks at that time and wanted to roll us out in Europe as a default search experience. Unfortunately, in 2017, yeah, Google came back to them and told them to stop that nonsense. But you know, silver lining, we ended up at Brave, which I think is a, is a much better home for us.

[00:09:54] Luke: So kind of almost like a hack in, in being able to work with the browser and the users to kind of backfill [00:10:00] the index in a way, right? Like, and to kind of work off of what actual people are browsing. I mean, I think people are unaware of just how much of that adversarial kind of gameplay and just like well poisoning there is with search engines and indexing, right?

[00:10:14] Jean Paul: Well, if you just go out on the web and crawl like a spider, right? Like you take one site and then you follow all the things, but you find out is that the biggest problem you’re going to have is, is exactly to filter out all the bad stuff. So you do have to have another way of prioritizing what you want to index.

[00:10:34] And for that, you do need the help of users. You didn’t need to like you cannot experiment too much. Like it’s easy to take 10 percent noise away, but it’s hard to take 90 percent like when it’s 90 percent of what you see is noise. It’s very, very deep that highlights again the value of search, right?

[00:10:49] There wouldn’t be so much crap out there. Because these bad websites are not done because people just write bad contents. They are done because people want the traffic from [00:11:00] search engines. Right. So it’s very difficult for a machine to detect what is made by another machine. remember once talking to Prabhakar Raghavan, who now leads Google search, And he was showing us the thousands of websites about lupus, and they were not understandable.

[00:11:18] These were early versions of LLMs, like these were language models that were not, they were easy to detect for humans, but computers had a terribly hard time because To computers, it looked like perfect English. It had the right statistical properties of English without making any sense. And all of these websites would link to one website that was created by a human.

[00:11:39] Because they knew that Google would have a tendency to just follow the links and land in, in that page. So very difficult when you start from scratch to deal with that. Right. But luckily we managed to buy data or always in a super private way, right? Like we don’t have F data that is user data. We don’t know anyone and we don’t know what they’re doing and we don’t even know [00:12:00] connection between pages, but, but we do know.

[00:12:02] Which pages are important

[00:12:04] Luke: just for folks that aren’t familiar, right? Like clicks was also kind of going at this from a privacy perspective.

[00:12:10] Jean Paul: Of course,

[00:12:10] Luke: Yeah, no, no, it was a privacy first company, although the way we got there is maybe a bit interesting because you know, when I started and with team, we started building the search not. Our prime objective. Our prime objective was how the hell do we build a search engine? It’s only at one point we realized that having a browser was a very important step in distributing a search engine so that we could not necessarily rely on Firefox and we needed A browser we controlled, but because we were essentially buying data left and right in the markets, we realized how much this tracking was a crazy problem.

[00:12:48] Jean Paul: And we can talk a lot about tracking as well, but we have better people are brave to talk about tracking, but we realized how big it was. And so when we build that browser, we build it for ourselves and we say. Well, you know, we might as [00:13:00] well have a lot of tracking protection in there because this is the browser.

[00:13:05] We’re going to use going forward. Most of us had extensions on top of our browsers, right? We would use, I don’t know, you block if it was a run at that time or ghostry or privacy badgers and all of these wonderful extension. So when we created our own browser, we figured we must have protection.

[00:13:21] Luke: Yeah, it’s one of those things I love too, is that, I mean, like having been at Brave from the 2016 days and one of those things that was striking to me as somebody working at a browser company was just how often people confuse a search engine with a browser.

[00:13:34] And so like, here you guys are building out this like wonderful search engine, you know, that’s kind of privacy minded too. And then we’re able to kind of bring those two things together. It really like, you kind of get the best of both worlds, right?

[00:13:47] Jean Paul: You know, at the end of the day, The search engine is like a steering wheel to a car.

[00:13:52] It’s understandable that people do not make the difference because it’s hard to imagine a search engine without the web. It’s hard to [00:14:00] imagine a search engine that is not connected to a browser because, you know, what are you going to find in the search engine? You’re going to find a place that you want to go to.

[00:14:07] Right. We can talk about AI and whether that changes or not, but people want to go to the open web. They want to see the menu of that restaurant where they’re going to go for dinner tonight. Right. So a search engine without the browser is kind of crazy. It’s like having a steering wheel and then you can turn it, but it doesn’t lead you anywhere.

[00:14:25] And having a browser without a search engine is basically unusable.

[00:14:29] Luke: Yeah. Or, you know, you get into these problems where you’re kind of become captured by the deals you’re getting, right? Like where if you see, you know, Firefox, for example, or Mozilla, where they’ve got this Google deal and, you know, they’re kind of, Have this hamstrung by this in a lot of ways, but the ability to have both and be independent like super crucial, I would imagine to growing the revenue

[00:14:50] Jean Paul: to this point, right?

[00:14:51] If you have a browser, you need a search engine. If you have a search engine, you need a brother. So the way this problem has been solved in the past, and this has [00:15:00] caused a lot of problems for which Google is now. In court, I mean, they’ve been declared guilty already. So it’s just a question of remedy and penalties now.

[00:15:06] But when they had the search engine, they realized that they had a really good search engine. But people do not understand how difficult it was for them to find users. And luckily, they had a lot of capital. So they could go out and Essentially pay a well at that time, a massive amount of money to become the default search engine, and they were looking for distribution because they were competing against Microsoft, which had the distributions were Explorer and had a tendency at the time and again to be kind of a monopolist as well.

[00:15:37] know, Google had trouble and they had to spray money around. To get distributed. And if you have a browser, as we know, at Brave and as others have discovered in the past, like Firefox, is you really have, like, no one pays for their browser. So what are you going to monetize it with? And it turns out that search is just the perfect monetization.

[00:15:55] So that’s all in. Well, the problem that Of course, this is when it becomes a [00:16:00] monopoly, right? Where Google buys all the browsers like Samsung, Apple, Iowa, you know, Safari and Firefox and tells them to make it super difficult to switch browser because every time they switch, they lose 40 or whatever the number is in the US.

[00:16:14] But these two are joined at the hip and it’s, I think for us engineers, it’s, easy to make the distinction, but for normal users, I don’t believe there is that much distinction anyway.

[00:16:23] Luke: No, I mean, that’s, that’s one of those things we saw early was just how often people would mix it up. I mean, we did a deal with DuckDuckGo for a couple of years before we had Brave Search and people were referred to one as a browser, one as a search engine.

[00:16:35] It was surprising. But then you also think about like, And this is the other thing with Google, where you’ve got Android, it’s kind of this ecosystem at the device level with that search bar, too. And everything just kind of works together.

[00:16:46] Jean Paul: I mean, if you want to extend that problem is people do not really make a difference between the operating system and the browser.

[00:16:53] Most of the time. I mean, apps making that a little bit more difficult, but at one point, the only application that you would install on your Windows [00:17:00] computer was a browser and then you were done. Right. Right. And then the next step is. You know, what is the real difference between the browser, the internet and the search engine, right?

[00:17:09] The search engine requires an open internet. If the internet does not exist, or if the open internet doesn’t exist, you will not need a search engine. App stores do not need search engines. App do not really need search engines either. So it’s the connection between what we call the internet, which is, is the web really, right?

[00:17:25] But the internet. The browser and the search engine is kind of one in the same thing and for users. And it turns out that Google has adopted this lesson as well for themselves. They say, well, we might as well own everything, right, but that’s where it gets a little too far.

[00:17:43] Luke: There’s so much noise around this right now.

[00:17:44] You’ve got, like, antitrust action in the EU, right, and then also in the U. S., and obviously both have very different ways of approaching this, like, from your point of view, as somebody who’s been in this space for a long time and seen these things kind of play out in different [00:18:00] areas before, like, where do you see this going?

[00:18:02] Jean Paul: We have to see, right? It’s going to be, it’s a tough job that this has to do. I don’t want to be, you know, I’m not the kind of person who would say, Oh, I would know exactly what to do. But it is clear that there has been a ruling that Google was ruled a monopoly. Of course they’re going to appeal, et cetera, et cetera.

[00:18:19] But I assume that this will hold, this is pretty solid argument. And there’s going to be a bunch of private lawsuits. Of people seeking damages, that’s going to be okay, but that’s not going to change the world. It’s just going to transfer some money from Google to someone else. And then there’s going to be the remedy part.

[00:18:36] And this is going to be interesting because Google has a lot more experience with this than the DOJ in the US because Google has gone through this exercise in Russia. They’ve gone through this exercise in Europe. They know what works, what doesn’t work in the sense of what works for not for the remedy.

[00:18:56] So obviously we’ve been asked, or I’ve been asked, we talked to people at [00:19:00] the DOJ to figure out, I mean, bottom line is, and I don’t mean this in a, in a sort of self serving way, but if the remedy do not help brave, then they didn’t work because you know, we are the only small search engine out there.

[00:19:18] Right.

[00:19:19] Luke: I mean, let’s dig into that a little bit too, because I think that’s another area people, don’t really understand very well either, because someone might say, here’s something like that and be like, well, what do you mean? We’ve got, there’s DuckDuckGo, there’s, there’s Yahoo, there’s all these different search engines available.

[00:19:32] What do you mean? There’s only like two or three? Like when people think about search engines, where are they missing the mark on search engine versus index?

[00:19:39] Jean Paul: Yeah, so so basically there’s two ways you can build a search engine is you can do everything by yourself like we did at clicks and not brave or you can just take an existing search engine that can be Google or Bing and package it, you know, put a different name on it.

[00:19:54] And both of these exist. So, for many, many years, there’s a company called start page that had a Google [00:20:00] deal where they could put Google results and their own name on top of it. A bunch of telcos in the world have packaged Google before, but Bing is the bigger player there. So, so people like Dr. Go or Ecosia in Germany, all very nice products and very nice companies have basically a Bing as a backend.

[00:20:17] So obviously the remedies are going to help them as well through. Bing mostly because they don’t only provide the results by a Bing, they also get the money from Bing, right? So, so the advertising is, is not sold by them, but by Bing and they get a revenue share. Brave is unique in a way that we, create our own index.

[00:20:38] We are 0 percent dependent on Bing and we have taken the difficult challenge. Of monetizing directly without any intermediaries, with our customers.

[00:20:48] Luke: You’re almost kind of building that advertising business on the search side up from scratch, basically. Right. Absolutely. Yeah. We touched on this a little bit earlier, but I’m really curious for your take on this too, especially because Brave [00:21:00] has this AI answer, you know, engine or summarizer kind of built in, like AI is kind of getting more and more integrated with search, people are using these prompts for more query type of engagement.

[00:21:11] Right. Yeah. What’s your take on the current state of AI and search? And do you think it’s gonna one’s gonna take over? Because people, some people are saying I’m just gonna use propensity for everything, right? Like forever. All my searches, where do you think the rubber is gonna meet the road there?

[00:21:24] Jean Paul: I’ll take it step by step from the point where the user is asking the question.

[00:21:28] One unresolved questions in AI is. Where does it live? Right? And some people will answer, well, it will live everywhere, but that’s not quite an answer. I think that it is clear that people will use Search fields wherever they are in the operating system or in the browsers to type in this request that they have, right?

[00:21:50] And then the request that people have will not change so much. I mean, the expectation about the quality of result probably will change, but the type [00:22:00] of Things that people look for is not going to change that much, right? And if you look at three general categories of things, which cover most everything in on the web today is, is people can ask to go somewhere with their browser.

[00:22:15] Right. And that’s not going to be solved by a I like if you type Facebook because you want to go to Facebook, the AI is not going to be super helpful. What is helpful is the brother to bring you to Facebook, right? Facebook is an obvious example. But if you if you want to know, you know, the opening, you want to check out the restaurant that you’re going to go to tonight or you want to watch a porn video or something.

[00:22:36] You want to go there. You don’t want the I to tell you about it, right? Thank you. Right, right. And so the second type of, of queries that people have is the commercial ones, the 18%, right? I’m looking for this particular headphone set. I’m looking for the shoe. I want to buy skis or something. And it is unclear or AI will be helpful in that, but I don’t think it’s going to hurt the [00:23:00] advertisers,

[00:23:00] Luke: right?

[00:23:01] Jean Paul: I have good evidence that the advertisers will embrace whatever the user wants to have and will. essentially try to be present in that using the same method as they do today, which is to basically buy their way into it. So I don’t see that many changes there. The biggest change will be in everything in between, which is, I’m interested to cook this recipe, or I want to know how to do this or how to do that.

[00:23:28] Up until recently, people would simply just type that they would look at the results and they will go to BBC Good Food to cook something, or they will go to, You know, some website that explains them how to fix that device or something. And these sites would almost exclusively finance themselves via advertising.

[00:23:47] Right?

[00:23:48] Luke: Oh, yeah. Look at those recipe sites, right? Like, yeah,

[00:23:51] Jean Paul: they rely on your attention to sell that advertising. Now, the problem is if the AI goes out to these websites, reads the content, summarize it back to [00:24:00] you in the search engine, you will get a better answer, but they will not get traffic. And that is concerning in a second degree, because in the first degree, it doesn’t really matter.

[00:24:09] You get, you know, the user gets what they want, et cetera. But obviously if these sites do not find a way to finance themselves, then they will stop doing them. And then, I don’t know, we all going to be stuck in recipes from 2024 and not, no content will be written forward because it’s not monetizable.

[00:24:28] Right. So, so I think that a new contract has to be sort of. It’s not a real contract, right? But an implicit contract has to be navigated to between people who write content and how they monetize themselves and search engines and AIs, et cetera. Just like, I guess, in the nineties, we had the same problem, right?

[00:24:48] Like it’s, you know, there was no financing. It was unclear how you get traffic. Right. We did, we did it anyway. Right. And then eventually it turns out that That search was the highest [00:25:00] monetization possible. And then other things, use, you know, ended up in programmatic, which is a big problem, but this is where we are.

[00:25:07] That’s an equilibrium that is going to be disrupted for sure by AI. But I think it’s wrong to worry about, you know, it’s a little bit like worrying about overpopulation on Mars before you build a rocket, right? We still trying to get AI that are actually immensely useful and not just. amazing because they are amazing, right?

[00:25:29] We just didn’t expect that. But, but it’s not like it replaced really anything in daily life, especially regarding search at the moment. I don’t think it’s right to fear it. I think it’s right to just navigate it.

[00:25:43] Luke: We had Joseph from search on earlier, right? Like when we did the AI answer engine first released that, and there’s a really interesting kind of parallel here between what I’m hearing around the search advertising piece and knowing which queries to monetize with, with [00:26:00] ads.

[00:26:00] And then also like with what Joseph was saying too, about like. Knowing which queries are the best to show an AI answer for. And it kind of seems like, you know, using that knowledge, we’re probably going to find really interesting ways to know, okay, well, which AI answers are the best to throw a product in, right.

[00:26:17] Or, or to help the user, right. And they

[00:26:19] Jean Paul: can overlap. In fact, we’ve been surprised that the advertisers response in. The times where we overlap overlap means that for one query, we show both the AI and an ad has been overwhelmingly positive. Because one thing you have to understand is advertisers are not interested in buying clicks that will not convert, right?

[00:26:40] We are not interested in selling clicks that don’t convert. Because if we do, then the CPC goes down, et cetera, et cetera. So having an AI followed by an ad or an ad followed by an AI, it doesn’t really matter, really filters the people between those who still want information and those who say, I know one transaction.[00:27:00]

[00:27:01] So I would argue that this is good for both, you know, the platform brave, the advertiser and the user because the user gets what they want. They can actually choose, right?

[00:27:10] Luke: Right.

[00:27:10] Jean Paul: The platform gets the lowest amount of impression. I mean, they get a better click quality and the advertiser also gets a better quality.

[00:27:17] So I really don’t see any, any problems with that.

[00:27:20] Luke: No, I mean, I think one of the cool things about the way Brave’s been approaching it, too, is this, we’re really good about citing sources, and then now we’re starting to see, like, other questions or ways it could be helpful, I mean, a lot of these wins you get on the transparency side are also kind of opportunities to put another follow up item there, like, oh, you’re, Looking for this, here’s a pot pan from whatever vendor, right?

[00:27:41] Like lots of opportunities there to make something useful. It seems like for the user, that’s also kind of a commercial win. It’s super interesting. It’s cool to hear too, that the advertisers have been noticing that one, because you know, who knows how often they pay attention and then, and then to, that they’re not turned off by it.

[00:27:58] Jean Paul: Absolutely not.

[00:27:59] Luke: [00:28:00] What are some of the biggest challenges that you’re facing? I remember when we first launched ads, you know, like coming into this from a privacy perspective, especially with the types of ads we were launching with, where you weren’t going off of the user’s intent with the search queries.

[00:28:15] A lot of pushback from traditional advertising. Are you facing a lot of those same challenges as you build up the search business? Or what are kind of the biggest challenges you’re seeing now, now that you got your arms around this whole advertising product side of Brave?

[00:28:29] Jean Paul: So the short answer is yes. And it’s an evolving sort of landscape.

[00:28:34] The core problem, right, is that The world has become a third party world. And by that, I mean that in a transaction between a publisher and an advertiser, there is very often a third party that observes the the transaction in order to essentially provide two services. One is targeting and the other one is attribution tracking.

[00:28:58] So that’s the core problem because [00:29:00] browser like brave basically remove all these third parties. So so that layer is just not available right now. Is that a real problem or not? This is a very interesting question. I guess that that is going to occupy the world for the next few years in search. First of all, there is no targeting, so we can already forget all of this profiling of people, et cetera, which is really bad.

[00:29:22] There’s no way that you can make an argument that this type of profiling is actually good. It’s not good. So in search, we don’t have that problem because the profiling is done by the query itself. There’s no need to know about who you are, what you did before. It’s just about what you want now, and you just said it.

[00:29:39] So there’s no privacy issue there. So the only question comes down to attribution. Attribution is I got that click from Brave and I want to know if that click actually leads to a transaction in the next five minutes or the next two days or whatever people measure. Most of the super big advertisers do this by themselves.

[00:29:59] So they are [00:30:00] first party, meaning that the information Like if you go to Wayfair or to Amazon or to Booking, which are all big customers of ours, they need no help and they are able to do this by themselves. And not only it works very well, but they are super satisfied with the traffic they get from Brave.

[00:30:16] It’s an absolute love fest. The problem comes from companies that for one reason or another, mostly because they’re not technical enough or big enough to afford this, have Essentially took the more convenient way of using, let’s say Google or Adobe or another company to do this measurement for themselves.

[00:30:38] And this is where the big mistake happened, allowed these companies to use their domain name to record all the information. Why is this a big mistake? It’s a big mistake because. These retailers are leaking information to Google. We know we’ll end up knowing more about the electronics industry because they have three or four or five or 10 or 20 customers in that industry.

[00:30:58] And they follow all of the [00:31:00] conversions and they end up building profiles of people that they can use for retargeting itself. So again, super, super evil stuff then and not acceptable. Now, believe it or not, Google is arguing to their customers that they should move to first party. Because it turns out that there are so many tools like brave and you block origin and grocery and what not that using third party do not provide a clear attribution picture and Google knows since 20 plus years that customers will measure well also spend well.

[00:31:35] So it does reach a point where companies using third party misattribute a lot of things which contribute to them spending less even on Google.

[00:31:45] Luke: Oh, yeah.

[00:31:46] Jean Paul: And so Google is, is promoting a first party attribution, but you know, it’s easier said than done because these companies are not technical because if they were technical, they would have done first party attribution all the way.

[00:31:59] So now [00:32:00] there’s a tension between the marketing department who buys our ads and the IT department was supposed to do the move to Facebook. first party and it takes two years instead of taking five minutes, right? Google had announced that they would deprecate third party cookies in Chrome. That would have helped because that put a lot of pressure on everyone.

[00:32:19] Unfortunately, they decided, or they were forced, I think in that case, to back away from that, which released a bit of the pressure. So I

[00:32:26] Luke: think the

[00:32:27] Jean Paul: problem of attribution in search Is painful at the moment, but it is a problem that will result itself over time. So obviously I would prefer if it was resolved today because of revenue will be, would be infinitely higher, but you know, it’s, it’s one thing.

[00:32:44] The other problem is targeting, right? The other problem, and that’s more for display type of ads is to say, I want to buy a person who likes video game and basketball, right? So how does that work today? You have a bunch of companies that [00:33:00] most people don’t know the name of. will put their trackers on every web page, including basketballs and gaming sites.

[00:33:06] And then they will create online profiles of these people. And when you hit a third website, completely unrelated to all the rest, that website will know, or I have a basketball person interested in basketball and gaming. And there was this company willing to pay Whatever, that is something that absolutely doesn’t work in brave.

[00:33:26] That is something we don’t want to have working brave. And this is basically called programmatic advertising. If I could call it a problem is that the buyers at the moment are in a phase where they prefer that type of buying. I’ve been around long enough to remember, you know, eight years ago where advertisers would tell me that they would Absolutely refuse to buy any of this thing.

[00:33:49] and now they all want to buy it. And in three years they will not want to buy it again either.

[00:33:55] Luke: There was a real pushback against this. I mean, that’s when I was doing ad product stuff and it was, you [00:34:00] know, as this stuff started proliferate, right? Like with programmatic where everybody was really up in arms about Google and everybody else getting all the data right now, it seems like it’s just got so easy for them to just say yes to the machine kind of thing.

[00:34:15] Jean Paul: Absolutely. And again, there will be a phase in in advertising where that goes away. I think as a general rule, I’ve never really met an advertiser that wants to breach people’s privacy,

[00:34:27] Luke: right?

[00:34:28] Jean Paul: I’ve never met a user who likes this. And so, you know, we should be able to solve that problem.

[00:34:35] Luke: Yeah,

[00:34:35] Jean Paul: it’s just annoying that The consequence for us at Brave obviously is less revenue.

[00:34:40] And if we have less revenue, then it means that we may need more capital to get to where we want to be. But, but it is a problem that is solvable and maybe we can find interesting ways to solve it because it is clear that advertisers do need to target. In search, you target by the query. So that, that’s a no brainer, but in other advertising, you know, if you want to [00:35:00] advertise for crypto products, you want to show it to people who are, you know, at least able to understand your ad and you want to avoid showing it to people who would just get puzzled by it.

[00:35:09] But you know, at Brave, we have ways to do this client side. We have ways to do this in a, in a totally privacy preserving manner. The real problem is, is these companies that somehow owns everything you have done in the past five years and monetize it between publishers, which are again, good people and advertisers, which are again, good people.

[00:35:28] So the principle of advertising and publishing is that. Publishers have too much attention or they have a lot of attention and advertisers are seeking attention. It’s like a bank and a customer, right? Like some people need money and so they can get a loan and people have money so they can get interest.

[00:35:42] There’s nothing evil in that. Where it gets a little bit problematic is when there’s a third party just Breaching the entire world’s privacy in order to solve that problem.

[00:35:52] Luke: Right, right, right.

[00:35:53] Jean Paul: The problem itself is perfectly fine. It’s just the solution is a little bit perverted.

[00:35:57] Luke: One of the reasons why this is all so hard is because nobody’s [00:36:00] really ever tried it the way that we’re trying it either, which is kind of cool.

[00:36:03] Like if we can really crack this nut, I think it’ll be something truly different from market.

[00:36:09] Jean Paul: As always with privacy products, right? Is it’s important to build something that actually people prefer and make it private at the same time, right? Because it turns out that neither users nor advertisers really, really say, first I want privacy, then I want performance, right?

[00:36:30] People basically want the thing to do something good that they want. And then, you know, it should be private.

[00:36:38] Luke: Yeah. Yeah.

[00:36:39] Jean Paul: So that’s why, you know, there’s a lot of companies at the moment saying, oh, you know, LLMs should be more private, et cetera. This is good. But you know, the problem of LLMs at the moment is that they, like, you wanna use the best LLM possible.

[00:36:51] You don’t want to use the most private LLM possible. That is not good. Right,

[00:36:55] Luke: right, right, right. You gotta be able to compete with the convenience factor. That’s one of the things [00:37:00] that’s so great about what we’ve been able to do with search is just that once we’re out there and you see Google integrating AI talking about putting, you know, screws on as a pizza topping, and then you see Brave actually telling you about mushrooms or pepperonis or whatever, you know, like, you’re like, cool, we’re like differentiating and the quality is really good.

[00:37:18] Jean Paul: We definitely need to be better and For advertising, there’s one kind of easy way to get to be better is to earn more money, right? And it turns out that if you don’t have a third party in the middle, you will have more money available. That’s just mathematics, right? So the ability that if we are able to provide advertisers with good value and we don’t have a third party in the middle, either they’re going to spend less or we’re going to earn more.

[00:37:46] And that in and out of itself in business to business, this is a value for users. This is different. Like you have to be faster, for example, it turns out if you block trackers, you get faster, right? You use less battery, [00:38:00] you use less network. So, so there’s advantage for the user as well with privacy. So I think we have a good base to solve that problem.

[00:38:08] The problem, of course, is if the whole world is working one way and you’re going a different way, it takes time.

[00:38:13] Luke: Yeah, fighting against the tide a little bit or going upstream. But hey, I mean, this is something we’ve never really shied from in this place, and we’re super excited to have you at the helm on the ad product side.

[00:38:24] I think it’s going to be really cool to see, you know, how this goes over the next coming months and you’re as part of like, you know, the key driver for Brave. We covered so much today, and I really appreciate you making the time to come on, because I think it’s the stuff that our users, our audience are super excited to hear about, and things that they really didn’t know about either.

[00:38:43] Is there anything we didn’t cover that you might want our users, our audience to know about?

[00:38:48] Jean Paul: I mean, I could talk for hours about this, as you may have noticed, basically, as, as a user, I think it’s, you know, the choices that you may make are important and true. It’s [00:39:00] convenient to use the default, but, you know, the default browser, the default search, et cetera, but, you know, just look a little bit further than that, that helps everybody, everybody, including you.

[00:39:12] and if you are an advertiser, just give us a call.

[00:39:14] Luke: Yeah, yeah. Fantastic. Fantastic.

[00:39:17] Jean Paul: I mean, you get good value, but it’s also kind of important. You know, the business proposition that we have is quite interesting, right? We basically go to advertiser and we say, at the best, you could make your business 1 percent more.

[00:39:31] Because that’s what we are, right? We are basically 1 percent of the market. Let’s say,

[00:39:34] Luke: so

[00:39:36] Jean Paul: it takes a little bit of courage for them to do that. And it turns out that a lot of them have been doing it and are quite happy about it because sure, they get 1 percent more, which, you know, if you make a billion a year, 1 percent is still.

[00:39:49] Sizable.

[00:39:50] Luke: Right. Right.

[00:39:51] Jean Paul: On the other hand, I’ve noticed that they are very proud of doing something different is definitely less boring than just buying Google and it’s quite [00:40:00] enjoyable for everyone. And it’s profitable, right? But our pitch is not, are you going to double your business? Right. Just let’s be, that’s not possible.

[00:40:08] Exactly. Yeah. Where can people find you? I am out there, but don’t expect much social commenting from me. I’m an introvert. So I guess a podcast, I haven’t used my Twitter account in a long time.

[00:40:29] Luke: In that event, I will definitely have to have you back on, to check back in on, on how things are going, as we go forward.

[00:40:35] And again, JP, I really appreciate the time. I thought it was a great conversation and I think our audience is going to really enjoy it. So thanks again for coming on today.

[00:40:43] Jean Paul: I hope so. Thanks a lot, Luke.

[00:40:46] Luke: Thanks for listening to the Brave Technologist podcast. To never miss an episode, make sure you hit follow in your podcast app.

[00:40:53] If you haven’t already made the switch to the Brave browser, you can download it for free today at brave. com. for listening. And start using Brave Search, which [00:41:00] enables you to search the web privately. Brave also shields you from the ads, trackers, and other creepy stuff following you across the web.