How a CMO prepares for a world beyond search

[00:00:00] Llibert Argerich: I think one of the things that I’ve realized in the last year and a half is that the advent of AEO and GEO has really given content marketing and content in general a much bigger role to play, right? So content’s always been important, but in the world of AI, it’s fundamental, right? Because in SEO, you need to have the right content in your platform, right?

[00:00:27] So you need to be able to appeal to the Google bot, and it will scroll and say, “Okay, yeah, this is a relevant authority in this particular topic, so we’re gonna put them on top of the search results.” So AI has exploded that, right? So the trust, the authority, the expertise really comes across the content, but not just the content in our website, but the content distributed everywhere in the internet, right?

[00:00:54] How people talk about us on Reddit, how people talk about us in different social media channels, um, the presence we have on YouTube, right? And all of that is what builds expertise, authority, and ultimately conveys trust. Because that’s how LLMs decide whether to quote you on a particular question that a customer asks on their interface and whether they are gonna show results from either content on your website or pick up that pro, that supplier from the feed that we’ve integrated and show the result there.

[00:01:33] Jason Hemingway: Welcome to In Other Words, the podcast from Phrase, where we speak with business leaders shaping how organizations grow, adapt, and connect around the world. I’m Jason Hemingway, CMO of Phrase, and today’s guest is Llibert Argerich, the first-ever Chief Marketing Officer at Thumbtack, a home services platform that generated nearly $500 million in revenue in ’25. Before Thumbtack, he scaled Udemy’s customer business from over $250 million to nearly $700 million across 65 languages. And before that, he spent 8 years at eBay running marketing across 8 different countries. He grew up in Andorra, nestled in the Pyrenees between France and Spain. He speaks 4 languages, and he’s lived in 5 countries. Llibert, great to have you on the show.

[00:02:16] Llibert Argerich: Yeah, thank you for having me, Jason. I’m excited for this conversation.

[00:02:19] Jason Hemingway: Let’s get underway, and I think you grew up in Andorra, which is a relatively small country of, you know, 80,000 people, but you’ve since built your career across, I don’t know, 5 different countries. So, when you look back, where did that kind of curiosity, that thought about the wider world, particularly in marketing, I guess, particularly begin?

[00:02:43] Llibert Argerich: Andorra is, I believe, the 13th smallest country in the world, right, and 80,000 people; it’s a valley in the middle of the Pyrenees. So, by nature, it’s very sheltered, right? And physically, geographically, it also culturally. And from a very early age, I had that curiosity, that itchiness to say, I need to see the world. I need to see what’s happening outside the boundaries of this country. And there’s amazing things about growing in a small place. You’re protected culturally. You know everybody. It’s great. And at the same time, you’re missing out on a big world wide out there, right? And yeah, from an early age, I was like, yeah, I just need to know what’s happening outside the boundaries of this country, right? And started with, like, traveling to Spain and France and ultimately just, like, traveling around the world and really just soaking in different cultures and different ways of thinking and different, just like, ways of living, right? And that’s been something that’s been driving me ever since.

[00:04:00] Jason Hemingway: That idea of moving between countries, I think if you look at your career, it’s kind of central to that. And if we start with something like eBay, and when you were leading, you know, you’re leading marketing across 8 different markets. What did that – I don’t know, maybe that multilingual background or that kind of cultural travel you did, what did it help you see that maybe something like data or, you know, that kind of thing couldn’t?

[00:04:26] Llibert Argerich: One of the benefits of growing up in Andorra was that I grew up speaking three languages from a very early age, right? Catalan, Spanish and French. And I was in between different cultures, particularly Catalan culture, Spanish culture, and French culture. So, from earlier on, I was very aware that there was different ways of thinking and doing things, right? So, it’s not the same thing to live in Barcelona, live in southern Spain, or live in Paris, right? So, that shaped earlier on the idea that in order to communicate and live and appeal to different people, you’ve got to adapt to their culture, their way of thinking, their way of doing things, right? And those things are not the same in different countries, in different geographies. So, at eBay managing 8 different countries—that was like the kind of default mode, right? It’s like, you cannot go to market in Germany the same way you go to market in the UK, or the same way you go to market in Italy. So, constantly thinking about, okay, so there’s a global platform enabling commerce globally. But when you talk to a buyer in Italy, it’s very different than talking to a buyer in the UK because their motivations and intrinsic cultural norms are different. So, ever so slightly adapting messages and communications to the local markets and cultures is key.

[00:05:59] Jason Hemingway: Yeah, I think that’s a really interesting thing, isn’t it? It’s all about adapting. You know, you might have a central thesis, a central message, but it’s how does that actually adapt and land in a market? You know, you don’t want to dilute the brand message, but you do have to adapt it so it’s culturally resonant with people. I think that’s a really, really interesting sort of thing. And given your experience, I guess moving around has kind of helped you think about that.

[00:06:23] Llibert Argerich: From eBay, then you took that experience.

[00:06:26] Jason Hemingway: Well, you tell me, but it seems like a very different environment. Udemy, which is, for those who don’t know, it’s a huge online learning platform, something like 40 million learners accessing content in 65 different languages, if I read that correctly.

[00:06:40] Llibert Argerich: That’s right.

[00:06:40] Jason Hemingway: Did that change your view of what it means to make an experience feel genuinely resonant or designed for the customer?

[00:06:46] Llibert Argerich: It was very similar to eBay. eBay is a global online marketplace, right? So, it was global. It’s my remit that went from the 8 markets I’d managed at eBay to be responsible for the entire global platform and marketing. So, it just took the challenge to another level, right? And from that perspective, yes, absolutely. It’s like now you need to market in India and Japan and Italy and the U.S and Mexico. So, it’s very different. And I think the common denominator is like you still have a brand, right? And that brand needs to be appealing globally, right? There’s a value prop. There’s a positioning. There’s a value that you deliver to instructors, which is the supply side at Udemy, and customers, which are the learners, which is the customer side, right? And that technology, that enablement, that connection between customers and suppliers, between the learners and the instructors, is global. But how you communicate to these users is different; how you position the courses, how you price the courses has to change country by country, because the power of acquisition, the disposable income, changes country by country. And therefore, you need to think not just about the message, but the user experience. What courses are more appealing in different countries, and how you price them, right? So, you need to really think about the holistic experience when you’re going to market to make sure that you’re adapting to the realities of that market.

[00:08:28] Jason Hemingway: That’s an interesting point, isn’t it? Because from a marketing comms point of view, you always think about how the message resonates. But actually, it’s the experience. The experience isn’t just the message. The experience is how the people interact with the brand, how they actually receive pricing; it’s all part of the marketing mix, isn’t it? Or how you position the product. So, I find it really interesting that you concentrate just as much on pricing as a kind of cultural aspect to that, isn’t there? Because it’s about price points that are relative.

[00:08:55] Llibert Argerich: It’s a reality of each market, right? So, at Udemy, different courses appealed to different audiences, and there was a different demand. Therefore, you cannot think about and again, disposable income is a big thing, right? So, a $50 course in the U.S would be prohibitive for most people in India, right? So, adapting that… Design courses were very popular in Italy, but coding courses were very popular in India, right? So, the market, the offer-demand dynamics really mandated the price elasticity and what kind of pricing you needed to apply to courses in different countries, right, to really respond to that offer-demand. Plus, some courses were created by Indian instructors, some courses were created by British instructors, and therefore that changed the dynamics as well.

[00:09:51] Jason Hemingway: Yeah. Interesting. And I guess that, just on a note, does that become very complex? How do you manage the complexity of that? Because you’ve got lots of different things to manage, lots of different levers to pull according to different markets.

[00:10:03] Llibert Argerich: That’s why you hire very smart people, PhDs, who can set up algorithms that look at price elasticity, that dynamically price courses based on the offer and demand in different countries, right? So, there was like a very complex operation behind it, where, as a CMO, I set the strategy, the direction, the guardrails for the communications and the positioning, and then working with the experience team, with the pricing team, to make sure that our offering, from a course, from a merchandising perspective and from a pricing perspective, adapted to that strategy.

[00:10:45] Jason Hemingway: Yeah, I love it. And it’s kind of that holistic strategy setting, looking at all those different levers, is interesting as it plays out in your career now. And we’ll talk about that in a moment, I think, because you’re responsible for a lot more actually than just plain vanilla marketing, let’s say. So, let’s talk about Thumbtack now, where you’re at today. And for those people that don’t know, Thumbtack connects homeowners with local service professionals across the U.S, the United States. So, really, after years of adapting brands across different countries and cultures, I guess the question I have for you is: how is operating a more local, let’s say, more singular market? How has that changed maybe the way you think about building experience or even trust with the customer?

[00:11:31] Llibert Argerich: It’s been a fascinating transition from global marketplaces to a hyper-local marketplace. So, the simplicity of marketing to the United States is, like, it’s one, two languages, right? So, it’s English and Spanish for the most part of it. But then the complexity arises from the hyper-local nature of the marketplace. So, Thumbtack has 300,000 pros nationwide. But if you’re a homeowner, you don’t really care that we have 300,000 pros nationwide. You care that we have the one pro who can come to your home at the time you need and do a good job, right? And for a pro, you care that we can provide enough jobs for you in your serviceable area, right? So, that’s where the hyperlocal marketplace creates complexity because Thumbtack is a technology platform that connects homeowners and pros nationwide, creating value for both. But in reality, it’s hundreds of thousands, if not millions, of hyperlocal marketplaces because the value exchange happens at the zip code, at the neighborhood code, right? So, we need to really make sure that that value exchange is created at the hyperlocal level, right? So, when you think about our areas of influence and when I think about marketing, it went from a global scope and thinking, oh, how do people in India think, to, hey, how people in Rincon Hill, San Francisco, the neighborhood connect and look for pros, and what’s our balance of, like, how’s our demand? How many homeowners do we have in Rincon Hill, San Francisco? And how many pros in different categories? And what’s that balance? So, our challenge is really to make sure that we have the right demand, the right supply to meet that demand, and the right demand to really give pros enough jobs and find that balance across.

[00:13:57] Jason Hemingway: You can sort of feel the common thread in what you’ve said is, you know, when you think about cultural adaptation, which is where you came from, now you’re adapting on a local scale, which is really interesting, but it’s a common thread of how do you adapt things to certain audiences that you’ve segmented, really, I guess.

[00:14:12] Llibert Argerich: Adapting to global audiences at a country level to, like, hyper-local, we’ll talk about an example on Seattle, but this hyper-local relevance is like, we just launched what we call a Seattle Market Takeover. And the premise of that is like, okay, so, what do Seattleites care about in terms of their homes, right? And the key insights that came through are very specific. Moss is a problem in Seattle, but it’s not in Phoenix, right? The humidity, the weather is very different, so you cannot go to market in Seattle the same way you would go in Phoenix, right? So, creating a clear understanding of what that local home ownership, what do homeowners care about, and what do local pros care about, and then, the types of jobs they do is very important. So, that hyper-local relevance and understanding the local audiences applies whether it’s nationwide or super-local.

[00:15:17] Jason Hemingway: Yeah, I guess it’s about relevance, isn’t it? You’re adapting to make it relevant. You used an example of ‘moss’, which I hadn’t even thought about. Moss is a thing. Of course, it’s not going to be Arizona. You probably don’t get much in Arizona. It’s quite dry. But let’s switch gears a touch and just talk about your marketing and your channel mix, you know, and I think, you know, that local relevance has to carry across lots of different, you know, broad mix of channels. And one of the things you’ve talked about is how you’ve expanded Thumbtack from kind of this heavy reliance on kind of search into sort of an 11-different-channel growth engine. So, if you could just explain what are those channels, again, it’s quite complicated with that amount of channels and the kind of local relevance that you’re trying to do or local adaptation. What are those channels? So, how do you keep that experience consistent across them?

[00:16:09] Llibert Argerich: As you know, one of the key premises for a CMO is just make sure that you don’t put all of your eggs in one basket, right? So, coming into Thumbtack, one of the key premises is like, hey, let’s really have proper diversification of where our customers are, right? And search continues to be a massively important channel because it has high intent. People are looking for plumbers, and then capturing that intent on the money is important, but that’s not nearly the total market, right? So, we want to go and appeal to people across the spectrum and people who may not yet be ready to hire someone, but are clearly homeowners and have these problems. And these either repairs, or maintenance, or improvements of the home are top of mind, right? So, that’s where the diversification expansion of the channels, and again, expanding beyond Google searching to Bing, expanding into AI chats and discussions, right? So being present on ChatGPT, on Claude, having multiple streaming platforms with video advertising and so on, being on YouTube, which was something that we scaled last year very significantly, all different social media platforms, right? So, you get the breadth, right? So, there’s a lot of different surfaces, a lot of different platforms that really allow us to talk to customers where they are, right? And the consistency comes from having a very clear brand position, a very clear set of value propositions, a very clear set of reasons to believe, right? And showing up in these different channels in a relevant way that are native to that channel, but that carry the value props for Thumbtack, right? So, at Thumbtack, we really care about connecting pros and homeowners with that value exchange. Our mission is to help. And to be the partner that helps homeowners and pros accomplish more. For a homeowner, that accomplish more is they have a home and they want to make that home either the dream home that they’ve always wanted to have, the safe place to raise their family, the place they can put their feet up, whatever home means to a homeowner. For a pro, it’s about building their business, to provide for the family, to be successful, to expand whatever success means to that pro. So, that’s the platform. That’s the messages we go out across different channels in very different ways that feel native to the platform, but ultimately connect to that value proposition and brand messaging.

[00:18:50] Jason Hemingway: Yeah, I guess it’s connecting that value prop with one of

[00:18:52] the authentic places that people will be, and making it not so interruptive or non-relevant, I guess, is the kind of way to think about it. So, we alluded to this question earlier, but I’m intrigued that your remit now spans marketing, product, data, engineering and sales, which is great for a CMO. But has being accountable for the entire growth engine changed how you think of the role of the CMO?

[00:19:17] Llibert Argerich: It’s been fascinating, challenging and empowering, right? Because ultimately, the role of the CMO is not just to put creative out in the world, right? It’s to connect the message, the value prop that you put out there and the actual experience that people have, right? So, how we go to market, the scalability, the consistency, the different channels. You can imagine that managing 11 channels requires quite a bit of infrastructure, right? From just connecting with Google and Meta platforms’ data exchange to be able to manage performance, to having a lifecycle infrastructure that helps orchestrating emails, push notifications, SMSs, to having landing pages that are actually personalizing to the person that comes in to make sure that we’re as relevant and we provide as good an experience starting from day one. In order to do that, you need cross-functional teams. Marketing is not enough. You need product managers that can work on the infrastructure and create requirements. You need engineers that build that infrastructure. You need applied scientists that can think about algorithms, LLMs, and how to implement AI into this infrastructure, right? So, as a CMO, I’m really straddling between the brand, the message, the creative, and the technology that makes all of this possible, that delivers ultimately the experience that will lead homeowners and pros to engage with Thumbtack, to convert, and ultimately to come back because we provide the best experience they possibly can get.

[00:21:07] Jason Hemingway: Yeah, look, I think it’s great. I mean, you’re also responsible for the strategies to start things off, aren’t you? Where should we be? What should we be doing? What shouldn’t we be doing? I think people always miss the strategic thinking at the beginning is key. And then everything else comes underneath, from the tactics and all of those things. But it’s really nice to hear a CMO that’s got a remit that spans all of that. I mean, it’s great because you are really in charge of the company’s growth agenda and how to connect with all of those different audiences.

[00:21:34] Llibert Argerich: I really believe that the modern CMO is that breadth, right? It’s really the growth officer, right? And that goes from positioning and how we talk to customers, which is fundamental. It’s never going to go away, but also, down to the nitty-gritty: how do we measure, how do we prove that we drive value to the P&L?

[00:21:57] Jason Hemingway: I think you’ve anticipated what my next question might be. I was listening to you on, I think it was the Marketing Vanguard podcast, which people should listen to. And you argued on that that marketing will always be treated as a cost center unless it can, you know, unless it can prove its contribution to the P&L. And you just talked about, you know, down to the nitty-gritty of metrics and those kinds of things. So, how do you think about, or how do you prove the contribution of marketing? What’s your thinking on that?

[00:22:25] This is my

[00:22:28] Llibert Argerich: most strongly held belief, and this is as a CMO, right? So, in order to really scale from an early-stage startup to a public company, marketing needs to be a profit center, not a cost center, right? And from day one, anytime I take a job, and particularly here at Thumbtack, it has always been about: how do we grow our audience? How do we appeal to more people across the United States? And how do we keep driving value to the P&L and the bottom line, right? And that comes from a very clear measurement infrastructure, very clear attribution, very clear incrementality, very clear ways of experimenting and measuring, right? And ultimately, how I prove that is through a handshake with a CFO, right? Our finance partners, our CFO, they need to understand how things are being measured. Any marketer who’s wrestling with attribution knows that there’s no perfect attribution, that there is no real source of truth. At the end of the day, you need to understand the limitations of what you’re measuring and how you’re measuring. And basically align with those limitations with the CFO to make sure that we’re talking the same language, right? And ultimately, how we spend dollars in marketing needs to be able to translate into how we record expenses in the P&L. And ultimately, all of that revenue that is being attributed to marketing needs to show up in the P&L, right? So, that’s where efficiency curves, marginal decreasing returns, profit maximization points, all of these concepts come into life in order to really have that healthy conversation between marketing and finance to align on what is success and how that translates into a positive bottom line.

[00:24:39] Jason Hemingway: I think there’s so many good things about what you’ve just said. I think that relationship with the CFOs is one of the things that I think people need to work on. The moment you start a new job as a CMO or even a senior marketing leader, it’s getting there with the CFO because essentially you’re on the same team. It’s not the battle. It’s just, you both want the same thing, which is to grow the business. And you both want to be able to prove the value of what each other are doing, or rather, what the spend is. And if you can make a decent business case and you can justify, you know, the P&L and show where you could operate some growth levers, I think most CFOs will appreciate that and actually help. And if you’re stuck for what to measure, they usually have some good ideas to kind of go away and think about it, don’t they?

[00:25:22] Llibert Argerich: Larry, our CFO here at Thumbtack, is one of my most trusted thought partners. So, I continue to go to him. It’s like, hey, this is the result of an experimentation on incrementality test that we’ve done. This is what we’re seeing. These are the challenges. What do you think? And having his point of view from a company finance architecture is invaluable, right, and vice versa. We build the trust in the measurement and the transparency of the limitations. Again, there is no perfect attribution. But being very clear and transparent on what we know and what we don’t know allows for these great conversations that allow for us both to compromise and say, okay, this is the best information we have. And this is how we make decisions moving forward. This is how we invest the dollars. And then this is how we’ll track them and make sure that it all lines up to what we are set to achieve.

[00:26:21] Jason Hemingway: I love it. Yeah, nothing is perfect. And I think it’s just establishing that kind of benchmark of understanding, so yeah. Look, we’ve gone quite heavy there talking about finance, things like that. So, we’re going to line things up with what we call the mid-show moment. And every guest we get on the In Other Words podcast, we ask the same question. And it’s all about language. What’s the most memorable language fail in your career that you’ve come across or seen?

[00:26:47] Llibert Argerich: Well, this one’s pretty personal because English is not my second or third; it’s my fourth language. And I’m notorious for misusing idioms. For the longest time. I’ve been saying, “Killing two birds from one shot.” Because in Spain, we shoot birds. We don’t throw stones at them, right? So, I’ve been saying that for the longest time. And people are very polite and never corrected me until one day I heard somebody say, “Killing two birds from one stone.” And I was like, oh, crap. I’ve been saying it wrong all along. And that applies to many, many different English idioms that I somewhat either translate from Spanish into English or misunderstand. So, yeah, I’m notorious for that. Misuse of idioms.

[00:27:38] Jason Hemingway: I mean, that sort of stuff happens all the time. I mean, I think I even do it in English, actually, and it is my first language. One of our team said the other day; I’m not going to name them publicly. They said, “It’s like an elephant in a China shop.” And that’s mixing two. So, a bull in a China shop is the expression. Or an elephant in the room. So, it’s done all the

[00:27:57] Llibert Argerich: time.

[00:27:57] That’s something I would do for sure.

[00:27:59] Jason Hemingway: I love it. Right. Let’s get back into it. And we’ll start thinking about AI. And you mentioned, you know, Thumbtack recently making its marketplace available through Claude, ChatGPT and Alexa Plus. And really, to sort of meet those customers where they’re trying to seek advice nowadays, you know, on the AI tools and platforms. A homeowner can ask an AI assistant about, I don’t know, a leaking pipe and then end up hiring a Thumbtack pro without ever really visiting your site, you know, coming to the Thumbtack website. So, what changes in customer behavior made you think that it was the right time to invest in that kind of strategy?

[00:28:37] Llibert Argerich: AI is changing the world, as we all know, right? And it’s fundamental for us to adapt, for any company to adapt to this new behavior. So, at Thumbtack, we’ve done two things. First is really re-platform our entire platform with AI, right? So, we’re leveraging LLMs to make sure that people come to Thumbtack and can ask questions, have a chat, and understand what’s the problem. If there’s a stain in your ceiling and it’s brown and looking odd, and you think you need a painter, the LLM will help you understand that, actually, maybe it’s not a painter; maybe it’s a plumber; maybe it’s a roofer; something’s quite wrong, and then connect you with the right one. So, really enabling, having that platform, that intelligence, really power the platform on Thumbtack’s skill. And likewise, that’s the behavior that people bring to different chat interfaces, whether that’s ChatGPT, whether that’s Gemini, whether that’s Claude; they have these questions, right? So, one of the fundamental insights for us at Thumbtack is, like, homeownership and being a business owner in the home services is full of uncertainty. You don’t know what’s wrong. You don’t know how often you should maintain something at your home. You don’t know who to hire to improve or remodel your kitchen. There’s a lot of questions. How much is that going to cost? And if you’re a pro, it’s like, where’s the next job going to come from? How busy am I going to be next week? So, those are uncertainty questions. And people take this uncertainty and need answers, right? And what AI is phenomenal at is at providing very rich contextual answers, right? And these answers happen to be on ChatGPT or in these different platforms, right? So, for us at Thumbtack, not only building that intelligence within our platform is important, but recognizing that more and more customers are using and having conversations in this platform. So, earlier on, we reached out to all these companies and had API partnerships, right? So, we take our Probase and our API and then provide that to the LLM models of different providers to make sure that when they provide answers, they connect with Thumbtack and say, “Hey, these are the Pros that we know are vetted.” They have background checks. They have licenses and so on. So, the LLMs have the right source of information to provide the right recommendations that will serve the customer’s questions best.

[00:31:17] Jason Hemingway: I love it. And I think that’s kind of sort of absolutely the right approach, those APIs to those. And it’s an interesting sort of strategic move and probably a very first-mover kind of idea. And you’ve got this idea of trust being really important. Everything you’ve said about that relationship between the pro and the homeowner. I guess one of the things, trust is non-negotiable. You know, when you’ve got someone coming in your home, it’s kind of very personal, you know, you’ve got to trust the platform that’s organizing that for you. So, as, kind of, that becomes distributed across all those AI platforms? How do you think about maintaining the brand or the trust from the brand whilst meeting those customers where they are on that journey using AI tools?

[00:31:56] Llibert Argerich: I think one of the things that I’ve realized in the last year and a half is that the advent of AEO and GEO has really given content marketing and content in general a much bigger role to play, right? So, content’s always been important, but in the world of AI, it’s fundamental, right? Because in SEO, you needed to have the right content in your platform, right? So, you need to be able to appeal to the Google bot, and it will scroll and say, okay, yeah, this is a relevant authority in this particular topic. So, we’re going to put them on top of the search results. So, AI has exploded that. Right? So, the trust, the authority, the expertise really comes across the content, but not just the content in our website, but the content distributed everywhere in the internet, right? How people talk about us on Reddit, how people talk about us in different social media channels, the presence we have on YouTube, right? And all of that is what builds expertise, authority, and ultimately conveys trust. So, how our content marketing and distribution strategy for Thumbtack – but for any brand out there – has become extremely important, right? Because that’s how LLMs decide whether to quote you on a particular question that a customer asks on their interface and whether they’re going to show results from either content on your website or pick up that Pro, that supplier from the feed that we’ve integrated and show the result there.

[00:33:49] Jason Hemingway: Yeah, it’s interesting. I suppose it ties very nicely back to your expansion of the channel mix because you just have to. It’s all part of the same plan, I guess. You also mentioned Seattle, and I think that’s a really interesting campaign that you did. And you’re using AI across all these different operations and marketing channels. For the Seattle summer campaign you mentioned, you hired a local artist to create illustrations by hand. Talk about that choice in the context of AI.

[00:34:17] Llibert Argerich: You know, using AI or using humans to do work is not either/or, right? It’s both. Right? So, as we think about telling stories, we think about our narratives is like AI provides a set of capabilities that didn’t exist before in terms of variety and proliferation of assets and different versions, so we can test much faster. But at the same time, we still need creative, flaring humans to do a lot of the work, right? The ideas and so on, they don’t come from AI; they come from people. And in this particular activation, since I was mentioning earlier, we wanted to really be relevant to the Seattle marketplace, right? So, how people thought about their homes and what they care about. We connected with Eleanor Doughty, who is a local artist, and she did the design and creative for the facade of the public library in Seattle. And that was great because, like, yeah, this is a recognizable designer artist in Seattle. And we got in touch with her, liked her ideas, and she connected with our vision to really provide that local flair and understanding of what it is to own a home in Seattle. And then the campaign started from there. So, again, it’s not about all AI or all human; the two things are not in conflict. They’re highly complementary. And we can still do an amazing job by leveraging both, right? So, ultimately, it’s what message, what narrative, what story, and how that connects to your brand. And then leveraging different assets and opportunities to do that, right?

[00:36:06] Jason Hemingway: guess it means that the dividing line is less about whether to use AI and more about where human judgment and AI kind of mix or fit. Which decisions, from a marketing point of view, are you comfortable delegating to AI? And which do you think still require that personal side? Is it the creativity? Just what’s your thinking on that?

[00:36:25] Llibert Argerich: The answer is going to change now and two years from now, right? So, now we’re at this point in time where it’s all about training the models, right? It’s very easy to create skills on Claude to make small actions, right? And getting, I don’t know, an email digest or what’s going on on my Slack or even creating some video assets on different tools, right? So, that’s easy today. Then the next level is like, okay, so getting agents to understand our data, right? And it’s not simple. It requires a lot of work and specialized work. Every database from different companies is specific. So, that database needs, in order for an agent to understand that database, it needs a semantic layer. So, the agent can understand what’s going on here. So, then it can actually propose valid actions and insights, right? So, as we learn and trust that, then the next layer is like, okay, so now the agent understands my data, understands what’s going on in a given channel. It’s giving me the right recommendations. So, now let’s allow the agent to go and make some of these actions and take some of these actions in different channels. So, connecting via MCP to Google, right? And then allowing that agent to go on Google and change creatives and change bids and change optimizations, right, is the next layer. All of that requires a very robust evaluation layer, right? Because again, the most valuable asset we have is our interactions with our customers. And you just can’t unleash agents and say, “Okay, yeah, I trust that it’s going to do its job and connect to customers.” Because hallucinations are real. Sometimes misreading of the data is real. Sometimes you need to correct. So, the human layer to make sure that we’re training the models properly and we allow them to incrementally take more and more decisions is one path. And we’re early in that journey versus where we’ll be two years from now. The thing that’s never going to go away is the thinking about the opportunities, thinking about the strategy, right? So, what I’m excited about is using AI to remove most of the manual friction that exists in today’s marketing workflows, right? So, how valuable is it for someone to connect to AdWords and manually update creatives and manually change bids? I mean, yes, it’s important, but those are things that we can totally delegate because they’re not necessarily value-adding. And we can focus our humans in really thinking about, okay, what are the next set of customer problems? How do we create more value? How do we test different versions of value propositions and then use AI to really provide that scale, right? I was mentioning earlier the hyper-local nature of the marketplace. There’s no amount of people I can hire that can effectively communicate and balance the hundreds of thousands and millions of markets within the marketplace, right? And that’s where AI really becomes invaluable, right? Setting the right agents, providing the right guidance, making sure that they make the right decisions, and then unleashing those agents to make trillions of decisions on a daily basis to optimize these different areas. That’s kind of like the direction. But as I said, it’s a journey. We’re early on in the workflow automation, and our humans are with fields. So, the way I describe our marketing practice is like marketing at Thumbtack and growth at Thumbtack; it’s empowered and enabled by AI, but it’s for humans to drive value for other humans, right? So, the human element’s never going to go away, but AI is going to help us increase the amount of value we drive.

[00:40:56] Jason Hemingway: Yeah. And look, I think all of those things are plus-one to all of those. I think, you know, ultimately accountability is a big thing, right? You know, there has to be someone who’s accountable for what’s going on, you know, and you need the guardrails. You need trust to build up, you know, in the same way he talks about trust with the customers. We need to be able to trust what the output of the AI is. And that always requires, at the moment, somebody to sort of think about how that all works and be accountable. Things go on, I guess. So, look, we’re coming sort of towards the close. And I guess two final questions before we do a little quickfire round, which we always do. So, you’ve had over 20 years of building, I don’t know, growth across lots of different countries and cultures and languages. What have you learned about your customers that you wish you’d understood at the very beginning?

[00:41:42] Llibert Argerich: Every customer is unique and special, right? And earlier on in my career, I was thinking about segmentation and groups of people and how people behave in similar ways. And to a certain extent, even today, we’re still thinking about, like, Gen Zs versus Millennials, right? And it’s like, well, but those are like massive audiences, like hundreds of millions of people in that bucket. You cannot simplify. It’s very easy to simplify, right? And ultimately, what technology, AI, is allowing us to do is to go into having one-on-one conversations with our customers, right? And really, that’s the value. Like, if you can’t really understand where a given customer is today, right? In the home services industry, for Thumbtack, it’s like, where is their life stage at? Are they single? Are they married? Are they buying a home? Are they getting ready to have a baby? What’s their home situation? How many rooms? All of that information allows us to have a very personalized conversation with our customers. And the more we know, the more we’re able to add value, to meet them where they are, and remove the uncertainty from the journey, right? So, again, what’s broken? Why is this not working? How often should I change the filter of my air conditioning? How much do I pay for the renovation of my kitchen? All of these are questions that burden homeowners everywhere. And our job is to fix that, right, to remove the uncertainty. And I wish that, yeah, 20 years ago, I had the clarity and the tools to be able to deliver that level of personalization.

[00:43:35] Jason Hemingway: Yeah, I mean, going back to one of the things you said earlier, I think the world is changing quickly, so that all that was not possible 20 years ago is absolutely possible now. And segmentation was probably a thing that you needed to do. And actually, segmentation is a good shorthand for just thinking about markets, I guess, in a sense. But now you can do it like almost micro-segmentation. You can also do those things like journey-based messaging, you know, where is that person in the journey with your brand? Are they trying to solve a problem, or are they trying to find a service? There’s all sorts of ways you can do it. And like you say, in two years’ time, things will go even further. So, it can be more relevant, I guess, to that person’s individual experience. So, lastly, we talked about search and things moving towards AI conversations. When you think about language and across different languages, what do you think brands must get right in order to remain that kind of recognizable and trusted kind of organization?

[00:44:29] Llibert Argerich: At the end of the day, I think, like, the most important thing for a brand is to have a clear positioning, right? So, what’s the value that we drive? And very often, companies are led by financial outcomes, by things that we want to deliver for us, the business, but we should never forget that any business exists to deliver value to the customers, right? In our case, it’s homeowners and Pros, right? So, ultimately, the most important thing is clearly aligning decisions, whether that’s marketing investments, whether that’s positioning messaging, whether that’s product roadmaps, whether that’s features that we want to launch, right? New businesses that we want to go into is, like, how is that helping us help our homeowners and our Pros, right? And we want to be that partner that really enables and helps homeowners and Pros to accomplish more. So, everything is centered around that existing pain points, opportunities to drive more value across the board, and that’s how our entire decision set is really focused on, right? That obsession with the customer and driving value for both sides of the marketplace.

[00:45:47] Jason Hemingway: Now, I guess in different markets, it’s that obsession with making sure that those cultural nuances that you talked about at the very beginning and the language rules and the cultural idioms, as we talked about at the very beginning, are all correct, relevant, so that you actually show that you understand that customer is not the same as a customer somewhere else. So, right, we’re into the last three questions, quick-fire questions, as short answers as possible. I sometimes say one word, but I’ll let you off if you do two or three. What does language intelligence mean to you?

[00:46:17] Llibert Argerich: I’ll say back to what I was mentioning earlier. It’s certainly connecting the value that you want to convey with the value that your customer perceives. Being language intelligent is being able to translate what you want people to see and how they actually perceive the value, and have that narrative be as clear as possible.

[00:46:39] Jason Hemingway: Love it. A global brand that’s getting AI distribution right other than yourself.

[00:46:46] Llibert Argerich: It’s hard to say, but I always like to think about Airbnb as one of my companies that I look up to. I think the work they’ve done across growth, content, branding is exceptional. And I always have them as a reference point because I’m a fan of the brand.

[00:47:10] Jason Hemingway: Yeah, love it. Love it. And what podcast or book would you recommend to another CMO or marketing leader or business leader?

[00:47:18] Llibert Argerich: Aside this one? I mean, like the podcast I was on that you mentioned, The Marketing Vanguard, I think that’s a great source. I also listen to the AI Daily Brief. I find it fascinating. Sometimes it just goes way over my head because it’s very technical, but it’s a great way to just get, like, a rapid-fire in 30 minutes of like what’s going on in the world of AI and learning things that otherwise would be very hard to learn.

[00:47:46] Jason Hemingway: Yeah, I agree with that. Well, look, all that remains is for me to say thank you so much for being on the podcast. It’s been a brilliant conversation and, you know, wish you all the best for the future. And hopefully, you know, you can come back in a couple of months, a couple of years and tell us what’s changed.

[00:48:01] Llibert Argerich: Yeah, I love that. No, thank you for having me. It was great talking to

[00:48:05] Jason Hemingway: you.

[00:48:06] Llibert, thank you for a brilliant conversation. You’ve taken us all the way from Andorra to the increasingly unpredictable world of the modern CMO. And you shared how language and culture and AI are really reshaping the way brands grow. So, thanks again.

[00:48:21] And that’s it for another episode of In Other Words, a podcast from Phrase. I’ve been your host, Jason Hemingway, and a big thank you to Llibert Argerich for joining us today. If this conversation got you thinking about how your brand travels across markets and languages, head to phrase.com

[00:48:37] We’d love to help you figure that out. And for more conversations on leadership, growth, and how businesses connect with customers across markets, please subscribe to In Other Words on YouTube, Spotify, Apple Podcasts, or indeed your preferred podcast platform. Thanks for listening, and I’ll see you next time.