BRIAN KENNY: If you want to know what keeps chief marketing officers awake at night, it’s this. For nearly three decades, winning online meant winning the search rankings. Brands learned how to optimize keywords, links, and content so consumers could find them and then decide for themselves what to trust. Generative AI is rewriting those rules. Instead of presenting 10 blue links, AI systems increasingly synthesize information, make comparisons, and offer recommendations of their own. That shift creates a profound challenge for marketers. What happens when an AI confidently tells consumers something about your brand that’s outdated or simply wrong?
Today’s case explores the emerging field of generative engine optimization and asks a larger question. When AI increasingly mediates what consumers know, consider, and buy, does the very nature of brand management have to change?
Today on Cold Call, we’ll discuss, “AthenaHQ: Building Brands for AI Search,” with case author Julian De Freitas and the company’s co-founder and CEO, Andrew Yan. I’m your host, Brian Kenny, and you’re listening to Cold Call on the HBR Podcast Network.
Julian De Freitas’ research addresses questions about how consumers perceive, adopt, and interact with AI systems. Welcome back, Julian.
JULIAN DE FREITAS: Always good to be here.
BRIAN KENNY: It’s another AI case, just another day for you, right?
JULIAN DE FREITAS: Another day in my life, yeah.
BRIAN KENNY: Great case. This will be fun. Andrew Yan is the co-founder and CEO of AthenaHQ and the protagonist in today’s case. Welcome, Andrew.
ANDREW YAN: Glad to be here.
BRIAN KENNY: Welcome to HBS. Following this session you’re going to be heading into an MBA classroom to listen to the students talk about your case and the decisions that you have to make. How are you feeling about that?
ANDREW YAN: Yeah, hopefully they have flattering things to say. But the reality of it is as a fast-moving field and being a startup in this field, there are highs and lows. And I think the case covers a lot about what it means to be evolving in this ever-changing field of AI search.
BRIAN KENNY: Yeah, you’re not joking. It’s changing so fast. And as a person who manages a brand, we are thinking about these things all the time and the impact that AI is having and will continue to have just on the discipline of marketing. It’s hard to know where it’s going. So it’s great that you’re coming here and I think you’ll really enjoy listening to the students and Julian talk about the case. There’s always some great insights that come out there and they’re probably going to learn a ton from listening to you.
So why don’t we just dive into our podcast. Julian, I’ll start with you. The case opens with an interesting example. An automaker discovers that a vehicle already on the market is selling well… That’s selling well, rather, is missing from a chatbot’s recommendations because the AI is relying on an old article, an article from 2021. I’m wondering why you decided to choose that moment to open the case and what larger problem does it reveal about brand management and the challenges of that in the age of AI?
JULIAN DE FREITAS: Yeah, so I chose this example firstly because it gets your attention. It’s not the sort of mistake a human might typically make, assuming some old article was still relevant. And of course, if you were in charge of that brand, it would be very concerning. But more importantly, I chose that example because it gets at a question that runs throughout the case, which is what’s changing because of AI search? And in particular, is the traditional wisdom around how to build a brand still the same or is it still the same, but maybe the mechanisms have changed? And in this particular example, in some ways it’s the traditional wisdom of have a consistent message, maintain control of authorship of your brand, but the mechanism has changed because the reason you had to be consistent in the past was that consumers have limited memory and they might be confused if they hear multiple things. And here it has more to do with how AI is searching for answers from these different sources to then decide what to feed you.
BRIAN KENNY: Yeah. How did you hear about AthenaHQ?
JULIAN DE FREITAS: I was lucky enough to have this brought to me by Allie in our California case center. Our case centers play such a crucial role in our ability to stay abreast of what’s happening and she connected the dots. She knew what I was interested in and it’s been a perfect marriage since.
BRIAN KENNY: Yeah, that’s great. We talk about AI search. I can’t tell you how many hours we’ve invested in the marketing community and understanding how to do search engine optimization over the past 10, 15, 20 years. It’s been this thing that everybody’s tried to get better at. And in some ways it feels like that was time wasted. I know it wasn’t and we can talk a little bit about that. But Andrew, let me turn to you for a minute. Can you, thinking about that incident that starts the case, take us inside the conversation with the automaker. And I’m wondering when executives deal with something like this for the first time, what’s their initial reaction? Do they realize they have an AI problem or do they think it’s a bigger brand issue?
ANDREW YAN: Yeah, that’s a great point. So with this global automaker, the reaction was, we don’t know what we don’t know. And there are these vestigial structures of our brand, which we didn’t even know were surfaced to AI, that AI is pulling from. And that’s the biggest concern from a brand perspective, from a CMO perspective, from a marketing exec perspective is how do you make sure that you have control over your brand? That’s a big risk. Of course, SEO is about how do you optimize your brand, but with AI, AI has an opinion and that’s a huge difference between SEO and this new space we’re calling generative engine optimization, GEO. Which is AI has an opinion about your brand, it has certain attributes it’ll associate with your brand. Similar to how your best friend, when they recommend your product, they might have opinions about the product. And AI has kind of turned into this sort of best friend recommendation, which knows a lot about you, has opinions about certain brands. And if you’re a brand manager, you care a lot about how is your brand represented on AI.
BRIAN KENNY: How did you come up with the idea to start AthenaHQ?
ANDREW YAN: Before Athena, I was a product manager on Google search and this was around 2024 when… Before the whole landscape has changed a lot since then. And I remember at Google, I was playing around with the earliest versions of Google AI mode and it would take three to ten seconds for it to get me to a response. And in the traditional search metrics, we’re talking about hundredths of a millisecond latency. And despite this taking three to ten seconds for me to get a response, I found myself gravitating towards using the preview version of AI mode. And that’s when I had this, oh wow moment. This is a different tech that is just objectively a better search experience. And you could kind of draw the line, connect the dots, that latency will improve over time as we all see. Latency has improved so much and it will only get better. And the moment latency drops, we see adoption rise, is a typical trend that we see with search. And we’ve seen that with ChatGPT, with Gemini, with AI Overviews.
BRIAN KENNY: Yeah. I go back to the days of dial-up modems. So I can tell you we become very patient over time. You had to wait a long time to even get a connection.
So Julian, let me come back to you for a minute. The case does a good job drawing the distinction between traditional search and that was the 10 blue bullets that I mentioned. We’re all familiar with getting that long listing of results and then you can pick and choose which one you want. AI takes the different approach, which is to synthesize that information and using some judgment like Andrew was just saying. How does that change the nature of the marketer’s job in that situation?
JULIAN DE FREITAS: Yeah, I think it changes it in a couple of ways. One is you need to change your mental model for what you’re doing, what you’re optimizing for. So before, as you mentioned, you’re trying to get ranked in a certain way. Now you’re trying to format content and be present on sources in such a way that this friend, or whatever metaphor that you’re using for the AI engine, wants to include your answer in its own paraphrased answer. So that’s a change in mental model. Relatedly, now that it’s talking conversationally about your brand, there are different metrics that you want to keep track of. So some of them are similar to SEO, like if you’re mentioned and the rank that you’re mentioned, but there are also things like sentiments. So is your brand mentioned in a positive way or negative way? And if it is, how much share do you have of that voice compared to your competitors?
And then as our opening example suggests, another factor is, is your brand being represented accurately? So it could be that the AI engine is drawing from a source that’s inaccurately portraying your brand. It could be that it’s hallucinating something about your brand. And so this is a new concern that you have to have on the top of your mind because of how that interface is changing.
BRIAN KENNY: Yeah. Andrew, what’s the value proposition that AthenaHQ offers to clients when you’re trying to convince them that this is an important thing for them to pay attention to?
ANDREW YAN: To put it shortly, we help brands see, act, and win on AI search. And there’s a monitoring part of how do you, like Julian mentioned, how do you see where your gaps are, where your weaknesses are? Being able to pinpoint with very pinpoint accuracy where those are coming from. Is it coming from this page on your website? Is it coming from your social pages? Is it coming from a third-party page?
And then acting is the critical part and that’s where Athena specializes in, is how do we help marketing teams market faster? And it’s about marketing velocity as well where different verticals are starting to move much faster than before with fast- moving consumer goods, for example, or software. These spaces are moving really fast. And this is an adjacent challenge we’re also tackling is when you’re marketing to AI, AI can change its opinion on a dime and you need to be able to change your marketing engine, so to speak as well and be agile. So that’s a whole other component that we work on.
BRIAN KENNY: The opening example is a really interesting one where it’s drawing on an article that’s old, five, six, seven years old. The information’s not up to date. We know that AI hallucinates. We know there’s the potential for it to provide information that’s not completely accurate or it puts an inflection on it that’s just not quite right. How do you combat something like that?
ANDREW YAN: Yeah, on multiple levels. So how we address this is we have a brand hub, we call it Knowledge Base, where we collect thousands of facts about your brand. And we’re able to cross reference that against the set of AI responses and see when AI responses contradict each other or contradict your brand hub. And from there we will detect where AI is hallucinating, where it’s deviating from your brand. And this really scales up, especially if you’re a multinational brand, you have such a scale of facts you need to manage as a brand that a human can’t do that. And it used to be in the past that you’d have a human probably in the legal department who would have a checklist of these are all the approved facts that we can have out. And to some degree that’s still true, but how do we augment that with AI to make sure that brand management is possible at a global scale?
BRIAN KENNY: Okay. We’re coming from a place where there was sort of one dominant algorithm that was Google that all the searches were derived from there. And now you’re getting into a situation where you’ve got multiple different AI platforms. I’m wondering how do we think about the brand managing its reputation when you have to be able to be managing multiple different sources of information?
JULIAN DE FREITAS: Yeah, it’s true. And I mean, the case talks about the distribution of sources, for instance, that the different models are pulling from where some like Copilot might pull from seven sources and then you have Grok that’s pulling from 20 something sources. And so there’s such a fragmented landscape. I think one concern brands should have is that they may be positioned differently across these models. There have been some preliminary studies showing that 55% of the time you might be positioned differently on one platform versus another. So being aware of that and is your brand consistently portrayed is one factor.
But I think also knowing who your target market is. So for example, if you’re selling an enterprise software solution, you probably want to prioritize Copilot because that’s where your customer might see you. And if you’re a consumer brand, let’s say that’s going to be less relevant.
Another factor would be just coverage. So if you don’t have a strong opinion on which particular model you need to be on, you might just track where is most of the search volume happening. Most of it still is on Google because that’s where the AI Overviews lie. And then we have ChatGPT and you can track it from there. And then I also think of this in terms of diversification. So you don’t want to put all your eggs in just one model or even certain sources that you know a model likes because as Andrew alluded to, this is a shifting landscape.
So you don’t want to lose sight of the bigger picture, which is have a consistent picture of your brand, be on enough of these models because at the end of the day, these AI models are trying to corroborate truths about you. And so that’s what you need to sort of keep your attention on.
BRIAN KENNY: And I would assume they want to be right. I mean, can we assume that the AI wants to get it right? Is that a safe assumption?
JULIAN DE FREITAS: I would say they want to be aligned with what humans would want and enjoy. And especially for certain technical questions, it’s going to want to be right. And then there are going to be other questions like what are the best trousers that it’s going to want to give you a useful picture that’s related to you in a personalized way.
BRIAN KENNY: Okay.
ANDREW YAN: And I can add to this too. Fundamentally, AI is looking for the shortest path to an answer. And whether that shortest path goes through your own content or if it goes through earned content, it doesn’t really care. And that’s the scary part from a brand management perspective.
BRIAN KENNY: It seems a little exhausting as a brand, again, I’ll speak as a brand manager to be able to manage all these different things. We’re talking about citations, share of voice sentiment, factual accuracy. As you’re thinking about giving guidance to your clients and to marketing leaders, what do you tell them to pay attention to? Is there something they should pay attention to more than something else?
ANDREW YAN: Yeah. So my overall philosophy is to simplify the field of AI search for our customers. What that means is that there are thousands of potential dimensions you could be looking at the data from, but from a marketing leadership perspective, we help simplify based on here are the axes that truly matter for your brand. Here’s a blueprint across, as Julian referenced, the AI models that matter the most for your brand, depending on your vertical and who you’re targeting and who your personas are. And based on that, we provide a customized blueprint for each brand. So it’s not a general landscape, here’s what it looks like for all brands, but if you’re a semiconductor company, here’s a plan specifically for your semiconductor audience. If you’re an automobile manufacturer, here’s a plan specifically for your audience and it expands.
BRIAN KENNY: Okay. So that’s kind of what you were referring to before. It’s got to be depending on who your customer base is, what your sector is, things are going to be a little bit different for you. Can you describe, because the case refers to the dark funnel, that sounds very mysterious. Can you describe what that is?
JULIAN DE FREITAS: Yeah. So I mean, first of all, for those in the audience who are not familiar with the marketing funnel, that’s the whole customer journey from being aware of your brand to eventually considering and comparing options and evaluating which one you’re going to choose and then having the intent to buy, and ultimately buying. And there’ve always been some parts of that journey that have been dark just because we don’t own every touchpoint that a customer might have as they’re researching the brand. So maybe we own the social page and our own branded website, but we don’t own some review site. So what’s fascinating about AI engines is that more of the funnel is becoming dark. And Andrew’s research and some other research I’ve seen suggests that that darkening is happening especially at the discovery phase at the top of the funnel and at the consideration and evaluation phase in the middle of the funnel. So if you think about it, that funnel is both collapsing into one touch point with the AI engine where you’re discovering and researching and comparing and evaluating, and all of that is not owned. So there’s a darkening that’s happening. And one of the symptoms that the case describes that this might be happening to you is that you’re seeing less referral traffic from touch points like an ad you put out, let’s say.
BRIAN KENNY: 100%, yeah.
JULIAN DE FREITAS: You’re seeing more direct traffic and then they’re staying in that session shorter with you before they buy. And so it suggests that they’ve already done the research elsewhere, probably on an AI engine, and then they’re coming to you.
BRIAN KENNY: But the fear that, I’ll speak for my marketing brethren here, the fear that a lot of us have is that people aren’t clicking through to your content. They’re looking at the AI search results. It’s nicely synthesized for them. They can get all the information they need there and they’re not necessarily clicking all the way through into your web presence. So I mean, how do you think about advising somebody on how to deal with that kind of a situation, Andrew?
ANDREW YAN: It’s a sign of the times, right?
BRIAN KENNY: Yeah.
ANDREW YAN: The shift to AI platforms in terms of search behavior is just as big of a shift as a shift from web to mobile, almost more than a decade ago. And we’re only going to see this trend continue. In terms of specific vertical KPIs and how that impacts, for example, in the field of e-commerce, we’ve seen that average order value for carts where the shopper is coming from AI is sometimes double or more than traditional traffic. They’re converting at higher rates across the board, across multiple industries that we’re seeing. And also they’re staying sometimes shorter, but sometimes longer on certain pages, specifically on these checkout pages, where it’s clear that the buyers have done the research elsewhere, probably on AI and they’re coming back just to check out.
BRIAN KENNY: Okay. So that’s the dark funnel. We don’t know where it’s happening, but somewhere in there the buyer’s making an informed decision about what they want to do. Andrew, I’m wondering how often you encounter a situation where someone’s digital presence, whether it’s their website or other sources of information, is conflicting in and of itself, right? We all like to think that the information that’s on our websites is accurate and up to date, but does this journey begin with me taking a look across the thousands and thousands of pages that Harvard Business School has on its website?
ANDREW YAN: So we do a lot of that for you. If you were to go through each one-
BRIAN KENNY: This is like a therapy session for me. This is great.
ANDREW YAN: Yeah. Maybe that’s my second career as a therapist, right? No, we build this brand profile automatically and it compounds through time. I can tell you a story about how other brands have managed this in the real world. So a very large financial institution used by, I believe millions of Americans, we found that one of their pricing pages about their mortgage rates was off, and it was using the old rate from a few years ago.
BRIAN KENNY: Ooh, that’s a biggie.
ANDREW YAN: It’s huge, right?
BRIAN KENNY: Yeah.
ANDREW YAN: Financial institutions have been fined millions of dollars for this problem. And that’s an example of you don’t know what you don’t know. This dark funnel is really very dark for a lot of CMOs. And in this analogy, we help turn on the lights, or we provide a flashlight where we’re not going to be able to see everything, but the flashlight, depending on where you point it, we’re able to focus on certain parts of the business that matter the most.
BRIAN KENNY: Julian, I want to talk about Reddit for a minute because the case points out, and I actually heard recently, as recent as this morning, I think when I was telling somebody about the conversation we were going to be having, that Reddit calls itself the largest LLM in media. Which might be true, but in my experience jumping around Reddit, it’s not all very helpful information. It seems like there’s a lot of noise. So I’m wondering how Reddit factors into this equation because it seems like it factors in pretty strongly.
JULIAN DE FREITAS: Yeah. The case describes that one of the most popular sources that these models draw from is Reddit, also YouTube, LinkedIn. And then you go down the list and then you start to see the usual culprits that we would typically think of as being authoritative, like Forbes. I think it says a few things. So one, at least historically, a site like Reddit is a place that you can crawl. And at the end of the day, again, these AI engines want to corroborate any answers. And so you can get some corroboration from a source like this. As you point out, it’s not all technical stuff on Reddit, but I think that’s also okay because in many cases, the types of questions are, what is the best this or what should I use for that? And so in those sort of cases, you want a chorus of opinions, rather than just one authoritative voice, even if it is very authoritative. I think if there are cases where the user is asking a very technical question, then you should expect that technical sources will be used for that, but that’s not happening all the time. But I think the scary part is that maybe ordinarily speaking, before AI search, you wouldn’t have gone on Reddit or YouTube, let’s say, as part of your customer journey. And now they’re actually influencing what you come away with in terms of what to do. And again, that’s a non-owned site. So it’s a different world that we are living in.
BRIAN KENNY: Right. No doubt. So let me ask you, Andrew, once you discover what’s truly influencing AI as it relates to your brand, is there an intervention you can do? Or is intervening going to cause more problems than it solves?
ANDREW YAN: I would say there are three types of interventions, broadly speaking, and that’s precisely where we specialize in. First is how do you, I call this vestigial content. How do you manage vestigial content, which you didn’t know was out there? Like in the example in the case, these content pieces which are providing outdated information or false information, that’s no longer true. We see this a lot with pricing specifications, and also if you’re a brand which has acquired multiple other brands and you’re conglomerating this house of brands, how do you manage that? So that’s the first type. Second type is optimizing existing content. Third is how do you figure out what types of net new content to be putting out to be optimizing for the AI engines? I like to use this term, the surface area of search, because managing a brand today on AI means that you have to manage a much broader surface area of search, which Athena centralizes. And we built Athena to centralize across multiple different channels, across social, PR, on page, all of these. In the old world, you have 10 blue links, majority of people would only click on the first three, and you’d only care really about the first three. Once you got to the first three, job done, we’re good here. Now off the shelf AI can pull from 20, 30, 40, and not to mention deep research can pull from hundreds. So as you see AI move more towards deep-research style of reasoning and analysis, the surface area just keeps on growing. The other nuance here is that not every source is weighed equally. Some sources are weighed more than others. So how do you piece together the interventions in the most quantitative way where you get the biggest bang for your buck? Because you don’t want to be doing a lot of actions without much payoff. If you’re doing a lot of actions, but they’re not the most impactful, then you’re just running around.
BRIAN KENNY: And this is another sort of additive thing. It seems like every time a new marketing technology or platform or approach comes along, it doesn’t replace the thing before it. It adds to the thing before it. So we’re piling on more and more things on top of each other. And here, one of the things the case points out is that marketers might need to be thinking about optimizing both for humans and how humans think about and ask questions, and also machines and what the machine is looking for and how it’s responding to those questions. Can you talk a little bit about that challenge for marketers?
JULIAN DE FREITAS: Yeah, it’s a challenge in the first place because humans and machines perceive the world in very different ways. Humans are very visual, we’re compelled by imagery and videos, and there’s experiential brands and emotional brands that move us. And then you have something like these AI engines, which, if they’re looking at a video, they’re actually taking screenshots of it and then reasoning over the frames. It’s not that they’re experiencing something when they watch the video.
And if they say that the video is beautiful, it might be because someone said that and they’re sort of repeating that. And I think the sort of tension is that let’s say you’re one of these brands that has historically invested in this kind of imagery. Now at the same time, you need to provide content that’s going to be easy for AI to digest because it’s not going to spend a lot of time on your website before it gives up. So that could mean FAQs, it could mean tables of comparisons and that sort of thing. And so the tension is if you’re catering too much toward one side or the other, are you hurting the experience, let’s say for people who previously really enjoyed the imagery and all of that that you’re putting on your website. So I like to think of this as a spectrum. In the middle, you’re trying to do both. On the one end, you’re saying, “I’m not going to change anything for AI. I hope that it’s just going to catch up and figure out how to parse my human readable website.” And then on the other end of the spectrum, I’m just going to have two different websites.
BRIAN KENNY: Oh, that’s interesting. Yeah, like a shadow website.
JULIAN DE FREITAS: Exactly, like a shadow website. And the case goes into why that may or may not be a good idea.
BRIAN KENNY: It seems like a terrible idea to me, to be honest with you. One is enough, and then having another website to manage would be super challenging. Julian and I had a conversation, I don’t know, I think it was last spring or so with a firm. He wrote a case about a firm called BrandBastion that manages automated replies to customers who are writing in to comment or ask questions to. I’m not doing it justice. But the question here has to do with automation and how much should we automate with regard to AI and how we think about the brand? And how much should we stay involved because judgment matters? And how do you differentiate between the two?
We’ve been talking a lot at Harvard Business School about the fact that we want our students to embrace AI. We’re trying to find ways to bring it into the curriculum through cases and through assignments and through the research that our faculty do.
At the same time, we believe deeply that human judgment will be the sort of mark of distinction between leaders who know how to navigate through these kinds of technologies and lead through them. And automation is great, but only to a certain extent. So long-winded way of saying how much should stay automated? How much of it should really stay in the hands of humans?
ANDREW YAN: That’s a great question. I’ve always been very humans-first approach. How I think about it is the way we’re doing marketing, both for our customers, but also at Athena is: AI is kind of the back office and humans are the front office and let humans do what humans do best, like conversations or exchanging ideas with other humans. There’s a lot of rote repetitive work in marketing and many other disciplines in general as well, which a human shouldn’t be doing that work. For example, when you’re optimizing your content, one aspect is how do you format the content? Our philosophy is no human should ever have to think about that. You should be thinking more about the messaging than about how am I going to format this? And we just bake that into our process. So using Athena, our customers don’t even have to think about formatting. We do the formatting for them, push it to the CMS, the content management system. It all happens without them thinking about it. And then they can focus more on the positioning, the messaging, this more strategic part, which is so much more valuable.
BRIAN KENNY: Julian, I’m wondering as you think about this, in the broader landscape, do we need to completely retool our organizations? Do we need to change the way that we’re thinking about roles within the marketing organization to address this? Or is this just a slighter shift where we have to maybe change our way of thinking versus the actual roles?
JULIAN DE FREITAS: I don’t think I would go out and recommend that everybody hire a head of GEO. In part because when you hire for a role, then part of what you’re communicating is, oh, this is a specialized function and it’s no one else’s responsibility. Whereas we’ve just been talking about how this combines SEO and social and optimizing your own content. So really, I think whoever in the organization is typically responsible for coordinating these various functions should be activated to lead this and get people looking at the same dashboard. So I don’t think a dramatic reorg is necessary, but these different functions do need to find a way to work together so that they’re facing the same north star.
BRIAN KENNY: Yeah. Okay. That’s great. So our time is almost up. This has been a great conversation as I knew it would be. I’ve got one question left for each of you and I’ll end with Julian since he wrote the case. So let me start with you, Andrew. If you were to look down the road, say five years and you’re advising that same client that was featured in the beginning of the case about what’s happening, what do you think the landscape would look like? That’s a little far into the future for AI, I know, but just thinking forward, what do you think brand marketers are going to need to be able to adapt to continue to thrive in this era?
ANDREW YAN: I think a lot about this concept of semi-autonomous marketing. Which is we’re based in San Francisco, I ride more self-driving cars now than I ride human driven cars. And in autonomous cars it’s a concept of autonomy, levels of autonomy. And it feels like we’re still at one autonomy with most marketing orgs that we work with. I think where we’ll end up three, four, five years from now is where humans are in the loop, but in the highest leveraged positions in the loop. And even with large clients where there’s more structure, more of a marketing force behind it, I think we’re going to see humans being this upper part of the org chart. And then there’ll be teams of agents that are perhaps managed by other agents, which ultimately rolls up to a human. And in this sort of organizational pyramid, you have these, for lack of a better word, lower intelligence agents, which are cheaper to use, which are managed by these higher intelligence agents, which are more expensive, but can plan better. And AI comes up with many plans of, “Oh, here’s a signal that we detected. Here’s something we can do based on that signal.” And it’ll aggregate all of these signals. But then the human ultimately will be the one who makes a judgment call at the end of the day to execute.
BRIAN KENNY: Yeah. Well, that sounds pretty cool to me actually. I like the idea of that. I like that vision.
JULIAN DE FREITAS: The humans at the top, right?
BRIAN KENNY: So Julian, let me end with you. I’m very simply wondering if you want our listeners to remember one thing about the AthenaHQ case, what would it be?
JULIAN DE FREITAS: Well, I think that much of the traditional brand wisdom still applies. So you still want to have a consistent message, you want to maintain authorial control over that. You want to undertake re-brandings very seriously and think twice before you do that. But what we’re seeing with AI search is that the mechanisms by which you get there are changing. And so unless you’re aware of that, you might be looking at the right north star, but you’re going to intervene in the wrong place and you’ll never get there.
BRIAN KENNY: Andrew, Julian, thanks for joining me on Cold Call.
JULIAN DE FREITAS: Thank you.
ANDREW YAN: Thank you for having us.
BRIAN KENNY: If you enjoy Cold Call, you might like our other podcasts: Climate Rising, Coaching Real Leaders, IdeaCast, Managing the Future of Work, Skydeck, and Think Big, Buy Small. Find them wherever you get your podcasts.
If you have any suggestions or just want to say hello, we want to hear from you. Email us at coldcall@hbs.edu. Thanks again for joining us. I’m your host Brian Kenny, and you’ve been listening to Cold Call, an official podcast of Harvard Business School and part of the HBR Podcast Network.
