Designing brands for Machine Customers
Your customers aren’t just human anymore. They’re machines...and they’re already making decisions about your brand. Agentic commerce is here, and it’s predicted to account for $300B–$500B in sales by 2030. Brands that design only for humans are already silently losing billions.
In this episode of The CMO Show, we talk to author of [descriptor] book, Machine Customers, Katja Forbes. They are joined by Gurvinder Singh, CMO at Altimetrik, which is putting Katja’s principles into action. Together they unpack the next tectonic shift in marketing: AI agents as buyers, validators, and gatekeepers.
From AI agents rejecting suppliers over expired certifications to entire networks blacklisting brands in seconds based on poor data hygiene, this conversation exposes a blunt truth. Marketing is becoming machine‑readable, binary, and unforgiving. It’s data. Verified. Structured. Machine‑legible. If machines can’t read your brand, they won’t trust it. And if they don’t trust it, they won’t buy it. This episode shows you how to navigate the shift, what to change, what to prioritise, and how CMOs can lead as machines move from channels to customers.
####
This episode of The CMO Show was brought to you by host Mark Jones, producers Niall Hughes and Kirsten Bables and audio engineer Ed Cheng. This is an edited excerpt of the podcast transcript.
####
Mark Jones
If there's one thing that marketers and CMOs have always cared about, it's the customer. That is, the human customer. But of course, we're in the age of AI now and AI agents. And of course, everything claw. And maybe if I say OpenClaw, you might know what I'm talking about, which begs the question, what happens when our customers become machines? Thanks for joining us on The CMO Show. Great to see you again. My name is Mark Jones and I'm your host. And The CMO Show is brought to you by ImpactInstitute in partnership with our friends at Adobe.
Today, an interesting conversation. We're going to look at customers that are both human and machine. Of course, I'm talking about AI, but we're going to come at it this from a slightly different perspective, which is, understanding these machines from the marketing lens. How do we work with them? What are the different types of machines that are shaping the future of marketing? And how do we think about it from the perspective of the ability to influence customers that are obviously machine, as I said, but what are they thinking and feeling? What decisions are they making on behalf of human customers? And what's the role of the marketer in all of that? So, fascinating and interesting conversation. Joining me is Katja Forbes. She's the author of Machine Customers. And Gurvinder Singh is chief marketing officer at Altimetrik. A nice double shot. Let's see what they have to say.
Well, hello, both. Thanks for joining us. Really appreciate you coming to speak with us on The CMO Show today.
Katja Forbes
Yeah, pleasure to be here.
Gurvinder Singh
Thanks for having us.
Mark Jones
Now, so I'm really interested to explore an interesting idea around the idea that brand has human customers and machine customers. It's an AI conversation, of course, but we are heading into a world where AI is beginning to make purchasing decisions for us, and that is a whole nother question for marketers.
To kick it off, Katja, I saw a piece about Australians abandoning Valentine's Day as the romantic recession hits, as it's called. Sounds dramatic. Do you think this is a good example of machine customers in actions? What do you think?
Katja Forbes
Well, I think there's a couple of things going on in that abandonment article. One was a little bit of cost of living, of course. But the other aspect in that article was a sense that if you outsource the decision-making to an AI agent about what they get for you for Valentine's Day, does it somehow make it less valuable, less thoughtful? If you're asking ChatGPT, Claude, whatever your favourite large language model is, "What shall I get my husband, wife, boyfriend, girlfriend for Valentine's Day?" and you don't put the thought into it yourself, maybe it's not as special. But the delegation of the decision in purchasing is very, very real.
Mark Jones
Well, I think there's a real difference between real flowers and digital flowers. So we could start maybe from that point of view, Katja. Gurvinder, thank you also for joining us. And you've got maybe a different perspective on this. But I know from Altimetrik's perspective, you guys are tracking this pretty closely. Do you want to set us up around this conversation and what's going on for the CMO?
Gurvinder Singh
One, I think the shift is real, right? And I think the difficult part that we are facing right now is that we have to manage both perspectives. I think rather than forgetting to send the flowers, even if a digital agent is helping me send them, I would be okay with that. And I think the same thing is happening from how our customers buy from us. And it's not either/or. I think we are literally trying to change the tyres of a running train, not even a running car. So I think both aspects would matter right now because we can't do either/or. The transition has not yet happened, really.
Mark Jones
Now, you're a CMO at Altimetrik. So it's kind of interesting, you're sort of speaking from a couple of perspectives here, right, in terms of the work that you do as a CMO and then this kind of broader perspective. What was the aha moment you had that led you to realise the significance of this shift where we have two sets of customers, if you like, so human and machine. What was the moment for you?
Gurvinder Singh
Yeah, I think two things. One, we are a AI engineering company. And what I also realise is that the kind of work that we're doing for our customers also needs to be applied in-house to how we do our marketing and the engineering needs to apply to my team also.
And my journey started with reading a book by Don Scheibenreif from Gartner. He wrote a book, along with Katja's book, which was a book about When Machines Become Customers. And the funny story is that he actually put a thing on LinkedIn promoting Katja's book, and that's what got me to reading about what she was doing. And I, trust me, lost my sleep for at least two to three nights because that whole thing about there are these agents or bots or machines out there which are going to take a decision, and are they able to read signals that I'm sending out there?
So, I think it was a practical approach. As I think about it, that how do we implement this because this is the reality we're living with. I would say more so in B2C than B2B, but I think that transition is going to happen very soon. So, that was my reckoning moment after reading her book.
Mark Jones
Katja, congratulations. Step one, get people to read your book. Step two, keep them awake with your ideas. So, well done. That's achievement unlocked, I reckon.
Katja Forbes
Thank you. Thank you.
Mark Jones
Just for the listener, just tell me a bit about the relationship between you guys. So, obviously, you've met through this book, and this has sparked a series of pieces that you're doing, some work and so forth. So, tell us what's going on.
Katja Forbes
From my perspective, Gurvinder has been an advocate for me. So the relationship has been one of advocacy. He has told people online about losing his sleep. And I've been super grateful for that because as an independent author and somebody who is so committed to trying to help people do this well, to navigate this paradigm shift well, having a really strong voice that advocates for it has been really powerful.
Mark Jones
Yeah, nice one. Well, let's then dive back into the story we picked up around Valentine's Day and think about it. You tell a story in your book about Tyler, an AI shopping agent that didn't just stop buying from a supermarket, but it blacklisted the whole chain, and then told all the other agents that it was blacklisted. Tell us about this story. And for me, that's a pretty sharp moment, I think, for a lot of CMOs because now we've got agents, well, with feelings, in inverted commas, right? Opinions, decisions that scale and you potentially lose control. What's going on here?
Katja Forbes
Yeah. So, the Tyler story... So, Tyler is real. Tyler is actually now real. I have built Tyler with OpenClaw. But the story that you're referring to is where Tyler goes out and tries to buy something on behalf of its human, me, in that scenario, organic spinach. But then Tyler goes and queries the supply chain of that supplier and finds that its organic certification has expired. And so it can't actually verify that it is organic spinach after all.
So, what Tyler does is, transaction halted, violation reported, and then proceeds to share it with the agent network to say, "This is a non-trustworthy supplier. It's told us lies." All the other agents then drop it. And then Tyler goes off to buy my organic spinach somewhere else that is certified. So, huzzah, I get my spinach. But the point being that data integrity is really binary. It either is or it isn't. There is no fluffy space here for us to greenwash our way past an agent because you can't. And I think that the cost of those kinds of data errors, even if you are still certified, but somehow your credential is not machine-readable, the cost of those errors is going to be significant.
Mark Jones
Gurvinder, what's your take on this? We're talking about a consumer experience, but obviously, this is going to be a B2B issue very soon, if not already.
Gurvinder Singh
Yeah. There's actually an example that I can share. First of all, I love reading books and I love blogging about books, which is what got me to promote Katja's work. In a similar instance, a friend was releasing a book. And of course, when you're releasing a book, you have to go do all the hype, promote the book. And right on the day of the launch, he got an email from a bot from Amazon saying that "The cover page is plagiarised and we are pulling the entire book out of circulation." And then there's no human that he could talk to. He was interfacing with a bot or with an agent on the other side. And I'm talking about this happening, like, six months ago.
So, I think some of this is already playing out. And what we are thinking of as a company is working on how to ensure that we put the right guardrails around these things. You cannot stop technology from going there and starting to take those decisions for you. What we as a company want to do is to ensure that we put the right guardrails around it so we can help our customers, help their customers in terms of making some of this a reality, which can actually help improve customer service and not go the other way in terms of creating a gap between what you're selling and what you're delivering.
Mark Jones
I feel like I want to go down this rabbit hole just a bit more. What do you do? Do you get your agent to talk to that agent to say, "Stop it. Everything's fine"? Or where does this go?
Gurvinder Singh
I think the answer is, I wish we had one single answer to it. This whole thing is going to get complicated. What will also happen is... I don't know if you heard, at GTC, there was a launch of NemoClaw, which is something that NVIDIA is launching. I heard there's something from Perplexity also that is coming. So, you will see this whole race towards some sort of a claw.
For example, the biggest question that I'm asking is, what's your claw strategy? Because believe it or not, it is going to come. And in my mind, how I think about it is it being an agent that simplifies my day. There are so many things in a day that I do which is not the best, most productive use of my time. So if this is something that can help me take out that additional time, and if I can give that time back to myself to think more about strategy, to think about more of spending time with my family, that's how I look at it. But then is it ready today? I don't think so. I think there's still work to be done in terms of putting the guardrails.
For example, I was listening to another podcast and there was a point about Google, for example, should be the number one to come up with an offering, because guess what? A lot of my stack, my Gmail, my Calendar, the Maps, everything is on the Google platform. So you will see a lot of companies come up with this, some sort of a claw strategy, which we will see. Whichever is able to win the trust point, the trust is the biggest point, will be the one that'll succeed.
Mark Jones
So, first we had lots of different GPTs. Now we're going to have lots of different claws. Katja?
Katja Forbes
Yeah. What Gurvinder is describing is what I call the foundation level. So, there are distinct altitudes in this of where people are operating. The foundation level is where we have a lot of that fragmentation, the payment rails providers in there, Visa, Mastercard, Worldpay, Stripe. We have got the different GPTs, the different large language models all trying to get their agentic commerce off the ground. And because it is all about at that foundation level, discoverability, machine readability, what happens is, once everybody kind of gets to that level, then we're actually making ourselves interchangeable. So if everybody's discoverable, everybody's machine-readable, what differentiates us?
So, the more sophisticated conversation is that other conversation that Gurvinder is talking about, which is the trust layer, so that middle layer of operational trust, which looks like trusted handshakes, transactions that you feel confident, fraud prevention, those kinds of things that we expect to be in place for us to want to participate in this ecosystem and for the machines to behave as customers in a trusted way.
But I think that the thing... if everybody's equally discoverable, everybody's equally trustworthy, again, what differentiates us? And from the experimentation that I've done, there is a stratospheric layer here where not a lot of people are talking or operating, which is about the values that your organisation signals out into the market from a data perspective and a machine-readability perspective. And do you stand for, is it independently verifiable that your organisation lives the values that it espouses?
Because if I, say, want to buy myself a really great quality but also sustainable mountain climbing jacket, then I'm going to put that out into the marketplace. And the one that has the most data about the way that their sustainability works and their circular economy participation works is Patagonia, and they have it in their Footprint Chronicles as data. So, my agent, Tyler, is going to go find that information and be able to discover it, trust it, but the thing that sets it apart is the fact that it matches my values, whatever those values are, and is independently verifiable in machine-readable format.
Mark Jones
Before we get to some of the pointy issues for CMOs in all this mix, there's some layers you're speaking about here in terms of the infrastructure layer, Visa and Mastercard, of course, and then you'll have the trust layer, you've got the consumers and the suppliers in the traditional commerce world. The thing that strikes me as I'm listening to you is that this is not too dissimilar to the credit scores we have in the banking system. But what's happening now is that they're being exposed more broadly at scale and it's being made more visible, I think, at least from an AI agent perspective. Is that a good way of thinking about it?
It's sort of this... Because trust has to be measured in some sort of way, is what you're talking about. But obviously, in the human world, somebody walks into a retail store, for example, and you look at them and you make a human decision as to whether this person looks safe or not, right? Or in the B2B situation, you have meetings and you get to know the people at the different company and then trust is built in a sort of a human way. Plus, I guess, as the relationship goes on, you could say there's certain data points around how often we meeted together and how quickly you responded and all those sorts of intuitive business signals that build trust. This seems to be very much a concreting of that. And I'd love both of your take on this. Is this what marketers now need to be thinking about? There's a whole technical layer to transactions now that has got to be baked into a pretty complicated system.
Gurvinder Singh
I think the way I see it is that, one, of course, yes, you are no longer a CMO. I also believe you are a CMTO, where you have to understand this technology layer. For example, one thing that I'm running in my team is that everyone is making an agent. So I'm making a CMO agent of myself, which should be able to help me qualify certain things that can help me save time. Similarly, my website head is making an agent that is helping us test the ability of what we are putting on the website to be put on social.
So, this whole thing about creating agents, about as we get more used to living and breathing through this change that we have to go through, I think this will become more of a normal. For example, today we've kind of understood the whole SEO algorithm in a way. Similarly, we will go through a big change where this whole AEO or this machine legibility is also going to become a part and parcel of how we operate as CMOs.
Katja Forbes
I also would caution CMOs as well to make sure that they don't get stuck at the technical layer, because then you just risk becoming utility, you risk becoming interchangeable-
Gurvinder Singh
Good point.
Katja Forbes
... as the pipes that get things done. So, the things that I've looked at, particularly in the marketing space, so there's two elements. One is, if we look, in the 1950s when we did marketing, we talked a lot about the features of the product. "Oh, here is a fridge. In the door of the fridge, there are 12 places for you to put eggs. The fridge handle works like this." And then through '80s, '90s, 2000s, we figured out in marketing, "All right, actually, we can get feelings into this. We can get emotions into this. We can start brand storytelling." And that's the way that things have moved on.
But now with, I guess, the data nature of AI agents requiring that information, we're kind of taking a step back to that feature-specific area. And what we actually need to do in marketing, and this is a new thing for marketing to learn how to do, is to quantify emotional sentiment as part of the description of the products and services that AI agents can then access.
And the other place where I think there's real gold for marketers as well, which is something they've never, ever done before, is figuring out how to capture the intent behind the AI agents that have come to ask them for their products and services, figure out what they've been instructed to do, and use that intent as a way to understand the machine customer type better, to understand the human behind the machine customer, whether human in a business or individual consumer, to actually then be able to have a more powerful conversation.
Intent capture, I don't think that we're particularly doing this yet. It's something that in the organisations I'm working with, I'm trying to teach them how to do. There's considerations here, is an intent that I express to my AI agent, is that a private conversation? Is it protected? So is there information that I'm sharing that maybe I don't want you to know about? So, I think there's a bunch of ethical quandaries and questions that we have to ask there, but those would be the two really big new things for marketing, quantifying emotional sentiment to change how you do marketing, to include machine customer types, and capturing the intent in order to change what you say, how you say it, and who you say it to.
Mark Jones
So, Gurvinder, you could speak to this, because I'm interested in an example. Help us understand this idea of encoding emotion into agents, because I would hope the agents are going to do that largely at scale in an automated way, right? But at the same time, there's got to be a creative input to that.
Gurvinder Singh
So, I think the first point there is, can we trust the data? Because the amount of bias that is there today, the ability of the agent to get trained and the fact that some of that data is coming from a specific set of people that belong to a specific region in the world or a specific strata of the society. So I think that's going to be a very key factor, which will kind of delay or where we will take more time in terms of how much we can trust these agents or these machines to be able to understand our emotions because they are able to understand the emotions of a specific gender or a specific type of people coming from a specific region.
Mark Jones
I would love a case study, another scenario, even. Can you maybe elaborate a bit more on that?
Katja Forbes
So, I've got one on the consumer side, if that's helpful. I have a collaborator in New York, a gentleman called Geoff Gibbins, who runs a company called Human Machines. Geoff has done a lot of work with a major cosmetics brand around understanding how they're positioning in the market when Gemini, Claude, ChatGPT go and search for their product category. What he's discovered in the research that he's done, we're talking 250 brands that he's researched across and more than 40,000 AI queries, and proved out that the thing that determines whether you get cited or whether you get recommended is what the rest of the internet says about you.
And so, when we're talking about quantifying emotional sentiment as data that is machine-readable, that can contribute to that elevation from citation to recommendation, it's the ability to find reviews and then quantify how somebody might have said something made them feel beautiful or it was really effective and "I felt like a better person when I used this product, that product, or the other product," or any of those kinds of emotional narratives around products.
So, we all know that most of the large language models use Reddit as one of their very strong sources of content, which is a place where people like to go to say things about what they think and feel. And so, figuring out how to do the same type of thing intentionally through marketing, rather than relying on all of the things that just get out there from an advocacy or a detractor perspective.
When I experimented with this myself, I asked Perplexity to go and find me a comfortable pair of flight socks. Now, Perplexity doesn't know what comfortable means. It can go and find me flight socks. It can find things that match the parameters, compression that I'm after, et cetera, et cetera. Ugh, this is a bit of a old-lady example, isn't it? But anyway, flight socks are really great. They're very, very good and keep you from getting DVTs. Anyway-
Mark Jones
Everybody needs them.
Katja Forbes
... point being, it doesn't understand comfortable.
Gurvinder Singh
You need to share the outcome with all of us. I need those socks as well.
Mark Jones
So building narrative trust versus algorithmic trust, I think, is a fascinating thing for marketers where in that example, there's the machines that can start to learn over time or the AI agents to learn over time what comfort means for you in that example. And so, that will happen. But obviously, on the marketing side, we're wanting to influence both the customers directly and the agents to say, "These are comfortable socks," in this simple example, "And here's the reasons why you can trust us to do that." So, there's quite a few layers that are emerging in here. And I think, obviously, we're at the very early stages of knowing how this would apply even across different industries and segments and B2B as well as B2C. So, it's very, very early. But quite clearly, trust will be the key there and who we trust and what's the agreed way of trusting things.
I wanted to jump to the SEO to AEO part of it, because I feel like that's the next topic we should briefly touch on before we wrap up, which is that this is a very interesting take you've got here, so answer engine optimization, Gurvinder. Do you want to explain what that is, how it relates to this conversation? It sounds like GEO to me, quite frankly.
Gurvinder Singh
SEO was about getting on top of the search results. What was my rank when somebody Googled my name? AEO is about answering engine optimization that today, if somebody would go and ask a question about... In fact, I did that right before joining in. I went to all the four models and I asked them, "What do you know about Altimetrik?" And it's funny, one of my channels that I use frequently said, "First thing that I want to tell you is that you are the CMO of the brand." And the others... So I think the-
Mark Jones
It's like, "Why are you asking me? You should know this."
Gurvinder Singh
No, but I think the point is that what'll happen now is that you have to be a part of the answer. It's not that you're a part of choices that somebody is making. So, if there's a question out there that which is the AI engineering company, or who can I trust in terms of if I have to outsource, or which are the top players out there, I think the point is that the difficult thing about AEO is that in a search engine, at least if you were on top 10, yes, the number of chances that you were able to get in terms of getting through would come down. But here, if you're not a part of the answer, then you're clearly lost out.
I would say SEO is a science which we clearly understood. We know all the answers. But AEO is something we are discovering and those standards and norms, which is where I like what on... The first thing that I like about what Katja put in her book is that she's able to put some of those frameworks and those methodologies which all the companies need to sit down. And we need to, in a very disciplined way, go out there. For example, I've turned my SEO guy, he's now responsible for AEO. Why so? Because he understands that part and he's able to figure out how do we move from this to this. This is not going away and that is coming forward. So how do you balance both?
I think there are five signals that matter the most, right? One is, do you have structured data and schema markup in terms of what you stand for? Second is, is this machine-readable? Can this value proposition be read by a machine? Third is a third-party validation. Like, to your point, Patagonia is known as Patagonia because of all that coverage that they get in terms of how the brand stands up for what. Fourth is, you have to be very specific in terms of claiming. It cannot be a vague thing. If you're claiming to be something, can you be very specific? Like you were asking us to be specific by giving us some examples. And then the final one is consistency across all the sources. No matter where it picks the signal from, whether it's coming from Reddit or Wikipedia or YouTube, is there a consistency in terms of that signal?
Mark Jones
I like the simple steps there, and that makes a lot of sense. I wanted to kind of cast forward a little bit and get a sense from both of you. When we start talking about a machine customer experience versus a human customer experience, what will the difference be and what does it look like to have a great machine customer experience? I feel like you've just actually started to lay out some of the steps that we can take towards that, but I'm also interested in the experience end of it, if you will.
Katja Forbes
When we talk about machine customers, I think at the moment, perhaps in our heads, we think AI agent that goes shopping for me. Machine customers is a very broad landscape, and we will see, in the very near-term future, machine acting as customers, delegated agents, Tyler going and shopping for me, but also factories that run predictive AI and predictive AI as maintenance that say, "My rotor's about to burn out. Please go and buy a new rotor," autonomously through their ERP systems. So factories ordering for themselves.
We're going to see cars acting as customers. Mercedes-Benz cars can already buy third-party apps through their platform, can negotiate for their own charging, pay for their own parking. So we're going to see physical AI. Anything that's autonomous will be able to transact. So we're talking Caterpillar trucks in mines that need new tyres that already operate on AI to get them around their locations. I think that we're also going to see robots in home before the end of this decade. Robots are going to be able to transact on behalf of their humans, on behalf of themselves.
So, what we need to do from an experience perspective is figure out, "Okay, what are the types of machine customers that are going to be wanting to buy our products and services? And for each one, how do we have to treat them intentionally and differently?" Because me selling tyres to a Caterpillar truck in a mine is different to me selling face moisturiser to a delegated agent for an individual person is different, again, to me trying to answer a procurement ask from a smart city with a multi-agent network. So, I think that the intentionality and understanding that it's not a one size fits all.
Mark Jones
Katja, just on that, your book has five types of machine customers, right?
Katja Forbes
Yes. Yes, that's correct. Delegated agent, autonomous buyer, multi-agent network, co-buyer scenario, which is a lot of where we are right now with people asking large language models, "What will I buy?" And intermediary broker, the machine customer that sits in between the human and the product, kind of like Amazon's Rufus or Walmart Sparky or those sorts of agents that sit on the platform. So, yes. And for each one, you have to treat them differently.
Mark Jones
So, Gurvinder, maybe you can wrap up for us on that. There's a comment you made earlier about CMOs being system designers and being strong on the tech. Those five different types seem to play into that quite nicely. What's your thought on how this is going to play out?
Gurvinder Singh
CMOs were known to be the chief storytellers I think we are kind of becoming the signal architect now where my ability to, one, still maintain the current status quo. Because guess what? The kind of contracts that I have to run, which run into multimillion dollars and multi-years, are still dependent on that human angle. So you have to still balance that part. But then the five types of machine customers that Katja referring to, how am I able to build signals which affect or influence those? And like I said, the structure that I said that it's my ability to convert my signals into being machine legible. And second is, are there buyers out there that I may not be even aware of?
I think I love how her book starts with someone missing or there's a flight that gets cancelled, and there is this agent that is going and booking the flight for me. It has all the rules. And that's what kept me awake. And I kept telling people that the fact that I may be the airline, which that agent was not even able to get through, even though I had the best customer service, I had the best pricing, kind of keeps me awake. So, I think our ability, CMOs have to play a big role in terms of understanding. One, you have to, first of all, appreciate.
Katja Forbes
The last piece of advice that I would leave CMOs with is, don't have your gaze firmly fixed on our Western markets to signal where this is going. And in China, over the Chinese New Year period, we saw Alibaba run the most incredible campaign where they offered 25 yuan coupons to everyone to order bubble tea. And what they did was they said, "You can have this 25 yuan, but you must use our model, Qwen, to order the bubble tea. So, what everybody did was they went to Qwen, the model, they put in their location, Qwen found them a bubble tea shop that was close to them. Then they customised their bubble tea to be exactly what they wanted, and then bought it, and it was delivered to them. Now, this did 10 million orders in the first nine hours and 120 million orders across the life cycle of the campaign. Alibaba spent 3 billion yuan on onboarding people to delegate a buying decision to an AI agent. That's about 431 million USD. So, this was a huge campaign. And the onboarding, I think it's the most expensive onboarding campaign in all of history, but it worked. think we really do need to pay attention to other markets other than Western markets to see where this is going, because this is agentic commerce at epic, epic scale.
Mark Jones
I've really enjoyed this conversation. It certainly expanded my mind and got me thinking about all sorts of things. Gurvinder, I hope I'm going to be able to keep sleeping, though, just to your point, because all these ideas, a lot to work on. So, Gurvinder and Katja, thank you so much for joining us on The CMO Show. I really appreciate it.
Gurvinder Singh
Thank you, Mark. It was a lot of fun. And stay awake, but then remember that this is for better human fulfilment at the end of the day.
Katja Forbes
Yeah. Thank you so much. It's been a delight to be here.
Mark Jones
Well, I hope you enjoyed that conversation and maybe, like me, left with a lot more questions than when you began. The thing that stands out to me from this conversation is thinking about the transactions that will be happening and, in fact, are already happening at scale through AI and how these different agents out there operating on behalf of human customers are making all sorts of decisions, and in the future, will be making even more purchasing decisions based on the values that they hold to be true, the trust signals that they're bringing in from all the different sources that we spoke about. And then, of course, the role of the marketer in all of that. The traditional notions about how we influence, shape, and nudge consumers are changing, not just because there's humans and machines, but because we've got to think about this whole different interconnected landscape. Absolutely fascinating and something that we need to keep digging into. So, that's it for this episode of The CMO Show. Thanks for joining us. My name is Mark Jones. See you next time.