Customerland

Who Owns Trust When Machines Choose

• mike giambattista • Season 3 • Episode 45

Share episode

The customer isn’t just human anymore. Agents are already learning our preferences, comparing offers, and making choices that used to hinge on emotion and brand memory. We sat down with Ben Wiener, global head at Cognizant Moment, to unpack how this rewires the entire go-to-market stack—from data foundations to creative expression—and why marketing plus IT has effectively become the business.

We dig into the real preparation work: consolidating and governing data so AI can act responsibly and effectively. Ben breaks down how organizations can fund the future by letting AI handle code cleanup, content migration, and workflow orchestration, then reinvest those savings into agent-ready experiences. We explore the move from omnichannel to omnibuyer, where humans, agentic buyers, and AI-assisted shoppers take nonlinear paths and demand different signals of value. That triggers hard questions about personalization, privacy, and compliance, especially in finance and healthcare, where trust is the currency.

What happens to preference when machines choose? We examine whether platforms, brands, or consumers will own the agent; how pay-for-preference could influence ranking; and why transparency standards will make or break adoption. Ben makes the case for brand-consistent AI personas that carry tone, empathy, and guardrails into machine-mediated interactions—like a great salesperson who’s consistent yet adaptive. Creativity doesn’t disappear; it scales. Relevance, speed, and ethics become the competitive layers that separate resilient brands from those left behind by algorithmic selection.

If you care about data strategy, customer experience, AI governance, and the future of brand loyalty, this conversation will sharpen your roadmap. Subscribe, share with a colleague who needs to hear it, and leave a review with your take: who should own the consumer agent—platforms, brands, or users?

SPEAKER_01:

There these would be the sort of line. Like there was there was the business, there was IT, and there was marketing. And depending on I don't know if you get angry letters, do we still get angry letters? I know I'll get angry here. Yeah, I get angry letters. Think about where we are now in the age of AI. IT and marketing are the business. You put those two things together, um, you know, and and and that becomes essentially the product.

SPEAKER_00:

Today on Customer Land, I'm with Ben Wiener, who is a global head of cognizant moment. And um just because this is the way things tend to work with an interview that I'm really, really looking forward to, like this one, we have to wrangle schedules like you've just never seen calendar dances like these before, but we finally made it here. So, Ben, thanks for joining me. Really appreciate it.

SPEAKER_01:

It's my pleasure. Uh, I've been looking forward to it as well. The fact that it's taken us a while to get this on the calendar has given me more time to read some of the stuff that you've written. And I feel you said we should have met 10 years ago. Uh, and I wish we had. I think that there's gonna be a lot of um shared insight around what's going on, and I'm looking forward to hearing what you make of this crazy world that we're emerging into.

SPEAKER_00:

Well, I'm uh I'm just a bystander, I live on the sidelines of all this stuff. You're actually deep in it, so I think your opinion actually counts.

SPEAKER_01:

You get to talk to a lot more people than I do who are doing this work, so I think it'll be a fascinating exchange of perspectives. Um, you know, my background, I started life as an ad guy, you know, mid-90s, uh, making really funny television commercials. Uh, and it was a great world, and it was a really simple world compared to where we are now. Wong duty? Long duty. Yep. Um, you know, I could spend, you know, the first uh 30 years of my career saying hi, I'm Ben Wiener from Long Duty, and getting the exactly that is the response I would get, which laughter. Um, and that would break the ice. And then hopefully we would create enough confidence that someone would entrust us with their marketing budget uh and we wouldn't put them out of business. Um, but at a certain point it became clear that there was a massive shift going on um for what it was going to take for brands to remain relevant. And the world was evolving from one where brands were defined by what they said, you know, what kind of ads they put out into the world, what kind of jingle they could write, uh, what kind of what their packaging looked like to a world that's defined by what brands did, sort of what they stood for, how they removed friction from a transaction, how they aligned with customer values. Um, you know, and that was obviously accelerated by you know the internet and digitization. I think the world that we're living in with AI is now the next great acceleration of that. And brands that figure out how to bring intelligence and identification um, you know, into their value proposition um are really going to be the ones that pull ahead for the future.

SPEAKER_00:

Couldn't agree with you more. And there is so much to unpack in just that last little little statement. But maybe just for context, can you tell us a little bit about actually let me back up just a half a step? Just a warning to our to the people who are listening to this. Um, Ben and I are in deep danger of going way off the rails in this conversation. So you'll have to just give us that kind of latitude. Um, we both think that there's a lot to talk about that's going to be a value here. So I'll I'll work really hard to keep us focused. Having said that, I think it's basically impossible. But just to set context, tell us a little bit about Cognizant Moment, um, how you are looking at the world right now, and then maybe more specifically the world of AI as it relates to brands. Um, because you've got a lot of thoughts there. I can already tell.

SPEAKER_01:

So uh Cognizant Moment was launched in January of this year. Uh, and what we did was we looked across Cognizant at the people and the platforms and the technologies that we were deploying that touched customer experience and brought them together into one organization. Um, because historically, our heritage at Cognizant is as a technology company. You know, we implement platforms. And so we had Salesforce, which is incredibly important to the marketing ecosystem sitting in one group. We had Adobe, which is important to the marketing ecosystem, and the creative ecosystem sitting in another group. We had design in another place, we had digital transformation strategy somewhere else. And that was a very sort of functional view of the world or a very platform-driven view of the world. And with cognizant moment, what we did was brought those things together so that we could focus on outcomes. Um, what's really on a client's mind? There is nobody waking up in the morning right now saying, God, I wish I could spend some more money on technology. And really, no one's getting out of bed thinking about um Salesforce or Adobe or Zoho or Braze or you name it. There's a proliferation of technologies that are all infusing um AI and customer experience. Everyone is getting out of bed saying, Who the heck is my customer? Because I don't know who they are anymore because I don't have first-party data like I used to. Or um God, how do I reinvent my business because I've got all of these legacy customers who want to engage with me in a certain way? But I need all these new customers who want to engage completely differently. And I don't have any more money to do that with. So, how do I figure out how to do what I'm doing today more efficiently so that I can invest in the transformation I need for the future? How do I sell more stuff? How do I make my customers you know more loyal? And so those are the questions that we answer at a strategic level, um, focus on outcomes. And then, you know, because we're cognizant, there is no technology that we don't deploy at scale with the certifications and expertise and specializations and scalability and onshore, near shore, offshore mix that you would expect from a systems integrator.

SPEAKER_00:

Is that all? Okay. That that's a lot, and you and you sit basically on top of that very large pyramid, at least as it relates to the advisory portion of that. Do I have that correct?

SPEAKER_01:

Yeah, the advisory, you know, and and the delivery. So I mean, we're you know, building, we're migrating, we're you know, all of those, yeah, you know the the the devil is often in the details. Um, and I think that's the challenge. Um, you know, agencies are really good at understanding the consumer and articulating strategy. They don't necessarily want to revel in the complexity of migrating 87 global websites. Um, I think legacy systems integrators are really, really great at technical things, but they're not necessarily having conversations at the business level, uh, or they're talking to the CTO, not the CMO, or they think about their customer as the technology organization, not the actual end user. So we bring the best of those of the thinking from both sides of that table together. Once again, yeah, this is um there used to be this sort of line, like there was there was the business, there was IT, and there was marketing. And depending on I don't I don't know if you get angry letters, do we still get angry letters? I know I'll get we're gonna get angry here. Yeah, I get angry. You think about where we are now in the age of AI, IT and marketing are the business. You put those two things together, um, you know, and and that becomes essentially the product or the product of the future, or certainly the way the product is accessed. Um and so that's a lot of what's driving the shift as well. Um, and in order for that to happen, IT organizations and marketing organizations need to engage with each other in ways that they haven't before.

SPEAKER_00:

There are an awful lot of consequences to that idea that marketing and IT bonded together, interlocked, are a business. Um, and I'm thinking of forget the technical challenges because those those are largely solvable, but the cultural ones are sometimes insurmountable or almost. I mean, I've I've talked to a lot of people who um system integrators as well, um, who are kind of like, look, if the pieces don't seem to want to fit together, we walk away. And I thought, well, they never seem to want to fit together because they're two disparate entities that don't don't really like each other a lot of times. So, how do you navigate that? I mean, you guys do this a lot, you do it at scale. So if you've got those kinds of, I'm gonna say consequences to uh reshaping a business to be marketing and IT driven, what does that look like? I mean, what do you what are you up against? And how do you how do you figure that out?

SPEAKER_01:

So um it's a balance of growth and risk, right? What what gets you fired? If you're a IT person, uh it's too much risk. If you're a marketing person, it's not enough growth. So the marketing organization, I don't want to say it's maverick, but it's really focused on how do we drive more revenue um through more marketing activity and allow us to flirt with the edges of privacy. Um, you know, and we're not the ones who get the phone call when the data breach happens. However, if the data breach happens, guess what? You've eroded brand trust and you know, marketing is doomed. Um, you know, IT see themselves as sort of protectors of corporate reputation, of operational stability. Um, you know, they're not necessarily as aligned to growth, but you know, now they need to be, uh, because there's no way to erase technological debt and equip organizations to be ready for the future and particularly to make the investments that need to be made in AI without more growth. So even if everyone's waking up with a different agenda or has a different sort of nightmare scenario, they need to come together. And I think there's an increasing recognition uh of that. Um, you know, that that's driving um a lot of what we're seeing as companies look to restructure themselves as well.

SPEAKER_00:

You know, you mentioned a few minutes ago that um, you know, nobody's waking up thinking, hey, I gotta spend more money on tech. Um, but there are certain things that are kind of universally keeping the C-suite up at night. I'd love to hear your thoughts on how, if, and how AI fits into those dream or nightmare scenarios in your experience, because you talk to these people a lot. Are people waking up kind of going, uh-oh, I better figure this out, or you know, my business as a business model goes away, or you know, is it a little bit less scary and uh you know than that?

unknown:

Yeah.

SPEAKER_01:

So I you know, I don't like to preach doom and gloom. You know, you can do a lot by you know, what do you what do we know about sort of storytelling, right? Raise the stakes, make everything existential. If companies that don't adopt AI in the next two years will be out of business, they'll be buggy whips. I'm not saying that. Um, but if you look at what's happened to traffic to online publishers, you know, because of AI-generated summaries, we can see that there are disruptions to existing business models that are going to happen pretty quickly and pretty profoundly that need to be reckoned with. Um, we have spent so much time and money and energy over the last two decades, three decades, becoming omni-channel, figuring out how to put our products or put our marketing, all the places a buyer might be, and making sure that we can connect with them everywhere along the journey. And now we have to shift from being not just omni-channel, but also being omnibuyer. How are we designing all of these experiences for a human buyer, for an agentic buyer, and for sort of an agent-enhanced human buyer, all of whom might be coming into the non-existent purchase funnel uh in a different way, and whose journeys are all nonlinear. So that's a really big picture thing to contend with. And if you think about the sheer amount of money that has been invested over decades building brand equity, uh creating preference, designing packaging that attracts you uh to a shelf, or building um a retail brand with a huge amount of trust goes away. Poof. When an agentic buyer with a very, very different algorithm, you know, uh, you know, that doesn't have that experience of drinking a perfect cold Coke on a summer day at the beach. Right. And has never seen that beloved Santa Claus ad, besides to suggest that today is the day I might want to try RC Cola because it's on sale.

SPEAKER_00:

Right. That's that's a nightmare scenario. As a recovery marketer, to me, that one would keep me up at night. It does keep me up at night, and I I think it keeps my colleagues up as up at night as well. Because to date, the solutions I'm aware of are all just possibilities. I don't think, and I'd love to hear your perspective on this. I don't see anyone who's got this roadmap to um selling into a agentic purchasing kind of worked out yet. I mean, if anybody does, it's probably folks like you.

SPEAKER_01:

I mean, we're all trying to figure it out. And the great thing about the future is that none of us is wrong yet.

SPEAKER_00:

Um I'm gonna quote that.

SPEAKER_01:

Um if I remember, I'll give you credit. That's brilliant. Um uh so we said creativity is the ability to hide your sources, so no need. Right. Okay. Um, and once you that that's a quote that has like 87 different attributions as well. Right. So you know these things sort of um who's going to own the agents that act on behalf of consumers is a really interesting question because there's always going to be a fundamental issue of trust. Um, there's going to be an advantage that incumbents have. I know Google, I've worked with, you know, I am so used to that search bar. Um, it's probably the website that comes up when my phone opens, um, and or my computer opens. And so um it seems natural that you know Gemni or Copilot or one of these hyperscalar AI platforms becomes, you know, where the identified consumer starts. But those of us uh with a certain amount of memory will also remember um Netscape and WebCrawler and sort of all sorts of other places that were default or Yahoo, you know, where we started journeys that aren't here anymore. So we can't count on incumbency. Um and then you look at sort of all of these independent uh you know, things like perplexity, you know, that have started AI native, um, you know, and are looking to establish themselves as being the next evolution of those things that we default to. Um, and then I think there's a sort of direct-to-consumer model as well, which is um, you know, if you imagine a world where each one of us has our own agent uh trained on our data, um perhaps following us around everywhere that we go, absorbing our preferences, that's a company that we haven't even thought of yet, or that's launching in a garage. Um, and it's you know, and it's going to be some combination uh uh of all of these things. Um the question of how do agents learn about us so that they can act most effectively on our behalf um is a really interesting thing as well. Um, you know, we've been living on this kind of knife edge of personalization versus privacy for quite some time. Um yeah, everybody everybody wants personalization, nobody wants to give up their privacy. The reality is the more you're willing to give, the more you're willing to get. Uh you're gonna be able to get. You know, we and so uh and AI only takes that to another level, um, you know, as as well.

SPEAKER_00:

Exponentially, actually, you know.

SPEAKER_01:

Yeah. Um, and then there's so some industries, highly regulated industries, um, on the one hand, we're seeing tremendously valuable use cases in healthcare, in financial services, um, for AI. And those are also areas where the data is incredibly sensitive and incredibly protected, either legislatively um or by companies that recognize they can only exist with consumer trust.

SPEAKER_00:

So with all the upcoming variations on agentic shopping, and and you know, I think a lot of people are predicting, you know, fully personalized agents shopping or acting on your behalf beyond shopping. So with all those different shopping AI-powered shopping variants, agents doing things on our behalf, and and I think everybody with perspective can see it's coming. And it just we don't know what it's gonna look like, when it's gonna look like that, but we know it's coming. Um, on the other hand, American businesses are notoriously short-sighted. We we work on our quarterly KPIs and our bonus structures that are you know making the the current numbers. So considering that, how is how is cognizant advising your clients right now to prepare for agentic shopping, not only uh getting in front of uh these shopping mechanisms, but I think even more interesting to me is the disinter the further disintermediation of the brand from the actual human. There's a lot in there, but I'd love to hear your thoughts on it. And maybe this is another two or three episodes we're trying to pack into one, but it's it's important.

SPEAKER_01:

Yes. I mean, so the first question you asked is how do you prepare? And um the most fundamental preparations that companies need to make have to do with their data. And that's bread and butter cognizant business. You know, we understand how to come into organizations to look at where data is sitting, how it's structured, what platforms it's on, uh, and make recommendations about ways to manage data more efficiently, more effectively, more cost-effectively, um, and manage data in a way that's going to be future-proof and help clients discern and navigate um the myriad platforms and places that are eager to take and host and store your data. And so, um, and that's why I think, you know, when I look across the landscape of um agency holding companies versus systems integrators versus specialists, like who's going to help drive this future? It starts with the people who have uh the ability to deal with data in a very sophisticated way, um, you know, and sort of figure out okay, here's what we're doing with Snowflake, here's what we're doing with, you know, data cloud. And so um those conversations are already happening. Clients may not love the amount of work that's involved, but ironically, um, that's where we're seeing a lot of early um identification is around building agents that go clean up data to get data ready for further agentification.

SPEAKER_00:

Right.

SPEAKER_01:

Um, that's sort of classic horizon one AI use cases, cleaning up code, cleaning up data. Um, you know, the next step is how do we fund investments in um you know the the the lay the agentic, you know, getting ready for agentic commerce or getting ready for agentic customer experiences. We have to go build something new. So we need to find a way to do the things that we're already doing um more effectively or more efficiently. Um so just like like a look look at the um like Adobe World, you know, the shift from AEM to edge delivery um is making the handling of content and the migration of content really, really efficient. So a lot of the time and a lot of the cost that used to get spent moving content around now gets done by AI. Those budgets get redeployed to start to experiment and answer the question that you just asked. And then ideally, um, you know, we start to industrialize the process of building websites, customer experiences for agents. Um, and that becomes more efficient as well. What we still don't know though is how agents are going to build preference. Right. So how much of it is what they learn from us, you know, how much of it is, you know, no, what what does Ben's agent know about Ben that makes it Ben's agent? Uh, versus how much of it is put into an algorithm. Um, and then to what degree are those algorithms transparent, right? Like um, in the same way that uh we are in the same way that a company pays for placement in a Google search result, are they going to pay for preference in an agentic algorithm? And how is that made transparent? And we don't have any industry standards on that right now, and I think that's another thing that will accelerate adoption of any kind of agent-enabled transaction, will be some assurance, either that there is no kind of pay-for-play in the algorithm or some transparency. Like, hey Ben, here's the this free agent, which is incredibly useful, is brought to you by the Ford Motor Car Company, Coca-Cola and Provence Insurance. Right. Among others, you know, and that's okay if I know that. Or, hey, Ben, you're gonna pay$9.95 a month for an amazing agent that's learned all the stuff that you taught about it and is not beholden to anybody. Um, and so, and it'll probably be different people will prioritize different things, much as they do now.

SPEAKER_00:

I um I look at this through a couple of different lenses, and and I'm in complete agreement with your your your outlook on this. But um, you know, when Alexa first came out on the market, the it was a listening device into aspects of your life that um prior to that just were completely private. You just there was no one that had any business or access to uh parts of your life otherwise. And then when um Alexa came out and its its counterparts, all of a sudden it could it could hear things that that it was just it was just never there before. So in the world of agentic AI, when we have these agents, they're gonna know things about us that uh could make us very uncomfortable if they became public or became used in any way.

SPEAKER_01:

Oh, and um, yeah, that I mean you know a slightly uncomfortable example, um, but not totally hypothetical um for um a cell phone company uh that has agentified a lot of its customer service. Um they've got sort of multiple chat bots going from the very traditional chat bot to the you know generative chatbot and now the agentified chat bot. And um, depending on what you've asked for, what you need to do, your query kind of goes to the level of technology that's required. Among the sort of highest value tasks is trying to convince people to not cancel a plan. And so if you have programmed an agent um to do everything it can to get a consumer to not cancel a plan, what's it gonna do? It's gonna start with you know, what kind of discount can we offer you today? And then maybe it moves up to the um, you know, um new phone that we can offer you, and then maybe it moves with the multi-line discount, and then maybe it moves to um we're gonna email all the pictures that are on your phone right now to everybody in your contacts. Right. Or just the nudes. Yeah, right. You know, yeah, we we're gonna take all we're gonna take your photo album on your phone, we're gonna run it through our algorithm, we're gonna find the things that you would never want anyone to see, and we're gonna email them to your mom. That would be consistent with the agent's programming to figure out how to get you to not cancel.

SPEAKER_00:

Yeah, that's the the nuclear option, but it's there.

SPEAKER_01:

And so, how do we start to build some of these guidelines? Um, you know, in. I mean, and so hopefully what you do, and then even worse, yeah, if that works, the agents will learn, like, oh, the most successful agents do this.

SPEAKER_00:

So that my employer didn't end up with a giant, you know, fine. Yes, exactly. Right. There's just there's so much there. Um, but I you know, I think it's gonna be a process. There's I think I look at it in terms of the tension between the human and the technology. It's not negative tension necessarily, but it's just that dance between trust and value, and it's always changing, it's always shifting and morphing. In fact, the the factors that that go into trust and value in one scenario are completely different in another, and they change over time. So that dynamic is very dynamic, and I think worth paying attention to.

SPEAKER_01:

So I think this gets to like what we talked about at the beginning of this conversation that you wrote about, which is um you know, this idea of emotion as a competitive layer. Um, and you know, we we we talk about user personas all the time. We're not necessarily talking about AI personas. Um, and right now we've sort of um built AI to have no personality. Um, you know, and at some point the AI personality needs to reflect the brand and become a differentiated experience. Um, and so um I think the hopefully the technology becomes more human when we give it permission to do that. Um, and then I think you know, one of the things that figure it's interesting to explore at a brand level is you know, brands are about consistency, which breeds trust, brands are about control, um, you know, and showing up, you know, in the same way everywhere and in a way that's consistent with values. On the other hand, when you engage with an individual salesperson, they both act within that brand framework. They're trained by ATT at the ATT store to give you an ATT experience, but the experience you get talking to Charles versus Talking to Susan is still different. And you might have a better experience with Charles than with Susan based on how they interpret and intuit your mood, uh, their natural skills. Um, and so how does that get translated um into an agentic experience or an AI experience?

SPEAKER_00:

It's um I'm looking forward to, and I haven't I have happened to have seen you know dozens and dozens of of uh attempts to uh monetize privacy and personal data somehow uh zero party plus or you know, pay for the absolute most transparent version of that, like you mentioned a moment ago. It'd be interesting to see how that plays out in an agentic world, um, because those variations are certainly going to be there.

SPEAKER_01:

There has been a lot of press coverage about the impact of chat GPT on brains. Yeah, we an amazing in the world of sort of functional MRIs, you know, we can see what's going on in people's brains, literally see what's going on in people's brains real time and how they are changing. Uh, and there's lots and lots of negative press around the impact of chat GTP on the brains of young people and on the brains of students. Uh, and on the other hand, you've got you know the University of California equipping um all of its students with chat GPT, the way we used to mandate people having um laptops. And is it because the University of California wants to make people dumber uh or turn them into inadvertent plagiarists? I don't think so. Um, you know, it it's yo, but what chat GPT or AI can do for those students is um ensure that your course selections actually meet the requirements of your major and actually get you out of there in four years, which is a remarkably challenging thing if you've parented a college student of late and tried to figure out, you know, how to assemble those classes in the right way. So these tech, you know, all of this has you know positives and negatives depending on how you want to use them. And so, you know, AI as your companion through college to write your essays for you, bad for your brain. AI as your companion to navigate all of the tedious, difficult procedural and bureaucratic elements of the college experience, fantastic. And so we're gonna have to decide, you know, what we do on that front. Um I love the advertising business. It was it. I mean, if you're a um slightly ADD liberal arts major who loves to learn new things and work with interesting people and can get passionate about you know adult diapers on one day and oil changes on another day and health insurance is another day, it's a fabulous place to have a career. Um, and I really, really worry about it. Um and so we we we sponsored um the VIP Clubhouse at Can Lion this year. Um, not a place where you typically see systems integrators showing up, but you know, Accenture has been there for years. You know, Infosys is there. Uh, and if you walk up and down the beach at Cannes, um all of the you know fabulously expensive installations that used to be the realm of WPP, um, or for that matter, Comcast, uh, are now Amazon, Meta, Salesforce, TikTok, you know, and next year it'll probably be Snowflake and PowerPoint. Um, but at the same time, at the end of the day, the super cool Brazilian creative people correcting, you know, gold lions are walking down the red carpet. And those worlds feel like they're just so moving further and further apart from each other. There's always going to be room for creativity. And, you know, I think you know, when you talk to creative directors, you talk to copywriters, designers, they'll talk about how AI can't truly be inventive. And I'm not so much of a technocrat, but I believe that, you know, the AI will get better. It'll get as the AI will get as good at writing books as Hemingway, it will get as good as Picasso, it doesn't need to, um, because 99% of what we do isn't all that inventive. So if it becomes more efficient, faster, more personal, or more relevant, it doesn't need to be spectacularly creative all the time. And that feels heretical, uh, but it also feels true. And unfortunately, I think the advertising business um has been going all in on its own self-uh serving myth of creativity for so long. It's missed the opportunity to own relevance from a technology perspective. And I think it's going to be very, very hard to get that back.

SPEAKER_00:

I just kind of want to let that thought hang in the air for a minute because I I think you're right. And um and I'm and I'm one of those people who believes like you do, like listen, it'll never it doesn't need to replace creativity as it can't. There's something just innately human about it that that isn't replaceable. And whether you can detect it or not, that's only half the equation. The other half is we have to express that stuff. It has to come out somehow. So it'll always be there somehow, whatever it looks like. Well, Ben, thank you for this. Um officially we'll wrap this one up, but we'll also commit to doing a follow-up because there just really is so much to talk about here. On top of the fact that you're just a really interesting guy. So let me let me put it out there and say open invitation. And if you're in New York, let's grab a coffee.

SPEAKER_01:

Uh, I would love to do that, and I will be in New York. So I will take you up on that.

Head Shot

Mike Giambattista

Host