Vineet Mehra on Marketing in the Agentic Era and How to speak the Love Language of the CFO

15:44Josh Spanier:Yeah.

15:44Vineet Mehra:How do you define it? How do you bring it into the culture?

15:47Josh Spanier:Yeah. Well, let's define it first at a macro level, and then I'll kind of lean in on one use case that I think is most pertinent in 2026 and beyond. So a full stack marketer, if you think about the job of a modern CMO in two ways, being a capital allocator to drive growth and being a systems architect, right? Those are the two things that I think a modern marketer has got to be. A full stack marketer means, I always say, I'm pretty dangerous across the funnel, right? Like I can get into it. I'm not the best at it. I have people way better than me at all of the things I'm about to talk about. But I know how to ask the right questions.

16:24Josh Spanier:And I definitely, no one can kind of put the wool over my eyes on things, right? I'm dangerous enough. So if I'm in a tech meeting, you know, I can have really deep conversations about the data pipes feeding the tooling that we've put in place to ensure that they're going to work. You have to be able to ask those right questions. And on the brand side, you have to be able to talk about social media and the latest trends and the latest creators, right? To be able to really, to get there. Now, you need a team of amazing people around you, which I'm thankful to have. But the full stack marketer can go from a systems conversation to a data science conversation to a conversation about the cultural zeitgeist and geek out on the ad platforms and the latest sort of launches that Google's made around, you know, automation of the media stack.

17:10Josh Spanier:Right. So that at the highest order is, I think, what the modern CMO has got to be able to do.

17:16Vineet Mehra:Double click into systems architecture. Sure. That's a relatively new phrase, and the audience may not quite understand what that is. I don't quite understand what that is. So what is systems architecture, do you mind?

17:25Josh Spanier:Well, I think it's more relevant than ever in this world of sort of the sygentic era, right? And let me dig into that. The first thing is, if you think about it, it's no longer just about leading a human organization, right? Like, I've got dozens of agents doing various tasks that humans used to do. So in many ways, you're not just architecting your organizational design with humans. You're now in almost like a cyborgish sort of way, co-piloting a lot of your humans with machines, right? And the role of machines is no longer just a tool, right? Like a hammer is a tool. The carpenter is like an agent, right?

18:12Josh Spanier:And so in many ways, it's that marketers aren't just using the tool. They're actually having almost assistants or agents beside them that are actually doing much more multi-level complex tasks. And if you think of your role as an organizational leader, you can no longer think of your role as just a leader of humans and designing a perfect organizational structure filled with, you know, sentient beings, you know. You now need to blend human and agents. And underneath that is the systems architecture, whether that be the tools you're using, where you're deploying agents first, second, and third, the data pipes that are training the brains of these agents.

18:56Josh Spanier:The reframe I'm trying to share here is think of organizational design as almost systems thinking, which is blending human, machine, agent in ways that have never been done before. And I think it's a huge skill set for CMOs moving forward.

19:12Vineet Mehra:It's super fascinating. And we shall be reading and learning more about it. I was at the Consumer Electronics Show a few weeks ago. And I know you hate going there, so you weren't there. But I was there. One of the things I did while I was in Las Vegas was I asked every marketer, junior, senior, and in between, where are you at on the AI progression? Are you sophisticated and progressive in all use of AI and marketing? Are you competent and sort of in the middle? Or are you a laggard? And six months ago, even three months ago, people were willing to say they were a laggard. They were willing to say the technology is not quite there, regulations, we're behind and we know it, but it's okay.

19:49Vineet Mehra:At CES this year, not a single marketer was willing to admit they're a laggard. And understandably, it's probably career limiting. But what was interesting to me is not a single marketer admitted that they were progressive, advanced, sophisticated, getting ahead. Everyone said, oh, I'm average. And it sounds like from what you're describing and your use of AI agents. I have two questions here. One, in Chine, you are sophisticated and progressive in your use of AI. You're getting ahead, right? So tell me about how you're using your AI agents. What is a progressive marketer to your mind actually using it?

20:24Vineet Mehra:And if there's a couple of examples of how you're using AI to really get ahead, I'd love to hear it.

20:29Josh Spanier:Happy to do that. Well, let me first address the laggard versus progressive thing. I always say to my team and to anyone who will listen, today is the worst AI we'll ever be, right? Right now.

20:40Vineet Mehra:That is true.

20:40Josh Spanier:And so it is moving at such a fast pace that just when, I think it's almost impossible to get past mid on this. Because just when you think you've got it figured out, you know, seven other point solutions are launching. And I think the honest truth is, is that all of us are kind of down the middle because it's changing so quickly. But we do a couple of things because this is a reality, right? One of our first AI wins, And I think it makes sense if you think about it, because in the early days when these tools were coming up, it was about, did you have the training material, like the codified ability to train agents to take on work, right?

21:18Josh Spanier:And one of the places in most organizations that things are codified and trained and written down perfectly in books is customer support, right? Because in many cases, you're outsourcing customer support, whether it's offshore, there is manual after manual after manual written. Today, more than 70 % of our customer support, not just chats, but also calls, are being handled by AI agents. And our NPS scores have gone up. It's been an unbelievable customer satisfaction driver, but also our costs have gone down. So a perfect sort of trifecta. And it was by far our first use case. But the reason I'm sharing that with you is not just to share the use case, but how I kind of think about this rapidly changing world.

22:04Josh Spanier:We would rarely on major initiatives like this, which is a huge part of our business, is customer support, you can imagine, in a financial services industry, bring in only one company. We actually, to this day, have two of the industry leaders in that agentic customer support space inside of Chime working together. And the reason we do that is we put all your eggs in any one basket. The field is changing so quickly that you get locked into one company. and now you're dedicated to their product roadmap. Well, we're so early, right? To use a baseball analogy, we're in the first couple of innings or to use a cricket analogy for you in the first couple of overs here.

22:43Josh Spanier:We're so early right now that to think any one company's got the roadmap figured out is kind of crazy. So one sort of tip and trick, right? Is to like say, hey, bring in two companies for your major initiatives and have them work at the same time with you.

22:59Vineet Mehra:Somebody who works for a financial company, channeling, yeah, spread your, diversify your risk. I get that. But let me just back up a second. You said that 70 % of your CRM is being used, using AI agents and that's increased your - That's customer support. Customer support. In financial, people care about their money a great deal, right? People get very emotional about their money, understandably. And for a financial institution, trust is, I assume, your number one thing. If your customers don't trust you, you are utterly screwed, right? Where's the human in the loop? Where's the red line? Where's the point as you went through this process is you started developing these AI agents.

23:32Vineet Mehra:You started using this technology where you said, we're not going to cross that Rubicon. We're not going to do that because that trust risk or the human element of this, because agents get things wrong all the time. AI gets things wrong all the time. It may have the right words, it has the wrong empathy. So you just talk about that process and how you work through all this.

23:51Josh Spanier:Yeah, look, it's a never ending series of iterations, right? It's not like there was this moment and we got it right. We're going going to be on this journey forever. But I think the NPS score increases, which is ultimately about member satisfaction, right? Who call in, you know, shows that we're on the right track. So it comes down to what the calls that are coming in are about, right? Or what the queries are about. In financial services, right? If your paycheck got deposited and you don't see it in your bank account, that's a really big deal. And so if it's a simple inquisition about like their balance or they're paychecked and coming on time, or maybe there's a password, right?

24:31Josh Spanier:There's some early use cases where actually an agent can handle those much better because you don't need a ton of empathy in those moments, right? People just want to call, they want an answer, and they want to get off the phone. But what that's allowed us to do, again, kind of like the creative example I gave you, is our customer support agents are now handling much more complex calls.

24:52Vineet Mehra:So everything you're describing, that makes perfect sense. And thank you for the explanation. I'm struck again by you're actually doing this. What is it about the culture of Chime? What is it about the behavior within the company over the years? Because I do not see other companies in other categories, even in the other financial services sector, embracing this next level ways of working, thinking, and doing that you seem to be doing. So what can we learn? What has been the unlocks that enable your company to do this?

25:21Josh Spanier:We have two, I think, two unique advantages and some things we're very consciously doing. The first thing is we're, yes, we are a financial services company, but at our core, we're a financial technology company. We're a tech company. And you're working at Google. You guys are clearly at the frontier of AI, both in what you offer, but also internally. Being a tech company at your core, right, really helps because the vast majority of your organization is made up of engineers and developers and people that are just going to be naturally curious, right, about technology, the boundaries of technology and where it's headed.

25:59Josh Spanier:The second part is we're a challenger brand, right? We are going up against institutions that are hundreds of years old. And, you know, last year we opened more checking accounts than anyone in America, right? We're a disruptor that's starting to really win some share. And being a challenger brand, a disruptor brand means you just your culture has this sort of mentality, right? Like, and I don't think that only should be for challenger brands. I think every company should have this curiosity, but it's easier when you have like a mountain to climb, right? Like you, you're constantly looking at every advantage you can possibly have.

26:36Josh Spanier:And that's deeply rooted in our culture. This idea of member obsession and disruption, right? It's very rooted in our culture. But third, then there's the management systems, you know, we put in place in this platform shift. So our entire management team came together and we said, look, we're going to make AI tooling because we don't know which one's going to be the winners here. Really easy to try in the company. So we work with our InfoSec team and we worked with our procurement teams to sort of lower the barriers to entry a little bit without obviously compromising on the important things on people's abilities, sort of really try out tools, whether you're in procurement or compliance or marketing or finance, you get people to become users and believers very quickly.

27:23Josh Spanier:And people start to see this is not going to take their job tomorrow, right? It's actually going to make their job more fulfilling, more interesting. And so that's been a really important thing as well, being able to lower the barriers to entry. So we maintain a level on those that are acceptable, but we've definitely allowed people to experiment a bit more.

27:40Vineet Mehra:Let me give you a hypothetical. You and I are both going to a CMO event in a couple of weekends' time, and you're going to speak with a lot of CMOs. Some of them are from more state, traditional companies. What do they need to do as individual leaders to get themselves up to speed? Changing the entire culture of a large corporation, that's hard. But as a marketing leader, we are all going through this AI transformation together. We're all learning at the same time. You've clearly embraced the moment and are pushing yourself. How do other leaders learn from that? What do they need to do to step up in this leadership moment?

28:18Josh Spanier:Yeah, look, I deeply understand these state companies and what it feels like to be a CMO in there. And I always come back to the same thing. It starts from the top. I mean, you're right. If you think you can walk into a company and change a culture of 100 ,000 people or 50 ,000 people or 20 ,000 people, that's unlikely to happen. at least not in the first kind of 24 months of your time there, it starts with you and being able to like, you know, use some of these tools in what is in your, you know, I've used this idea of a circle of control versus your circle of influence, right? And if you understand your circle of control that you personally, as an individual leader actually have, I would say, full discretion over, whether that's how I may be presenting documents to the executive team that I'm on, right?

29:06Josh Spanier:And just starting with, hey, here's some of the tools I use to like put this together, right? Or whether it's in your own department where you can actually increase velocity or reduce cost and then socializing those things with these tools. It starts with beginning your circle of control. Don't try to change the entire enterprise and start to prove it out. Understand what your company cares about, the financial metrics, the unit economic metrics, the cultural metrics they care about, and start to show that based on what's in your circle of control, you can actually impact those things and the change will start.

29:44Josh Spanier:But there's no silver bullet. It's not going to happen overnight.

29:47Vineet Mehra:I want to combine two things you seem to be very passionate about, AI and pizza, right? And so you said publicly that if you could, you would retire to run your own pizza business. So I just want to try and model, maybe we can brainstorm here. how can ai make pizza more interesting or better i like that from the things you've learned and seen and self-trained in yourself so clearly the i'll start and just say i have got i experiment with ai tools in my own time on the weekend to try and learn how to use them right across companies right obviously i start with gemini and google's products because they're the best but i'm using everyone's, but the notion of creating new pizza toppings, flavors and senses that you never think doing, AI could certainly play a role in that, right?

30:35Josh Spanier:A hundred percent. You know, AI can play a role in just about anything. Okay. So you've done your research, obviously. My dream in life is just to be a pizzaiolo and open a little spot one day. All this AI talk, I'll open a pizza shop one day. But anyone who knows pizza, and I'm kind of fanatical about it, knows that the dough is where it's at. And I haven't tried this yet, but I actually wonder, could AI create the optimal dough combination between flour, yeast, the dough, salt, all the things that make it well? Now, I'm a Neapolitan purist, so it's killing me to say this a little bit. But I wonder if an AI tool could really help you perfect that hundreds of year old recipe and take it to the next level and customize it to your guest's palette.

31:24Vineet Mehra:This is a really exciting thing. And one of the things I picked up from Shelley Palmer in my conversation with him is that as leaders, if we are not embracing using AI in our personal lives, our teams will understand and see that. And it sounds like that's some of what's going on inside of tech companies, that curiosity you mentioned, but embracing these things do better. So let me bring it back to Chime for a second. The best pizza dough is a combination of flour, salt, water, and a bit of yeast, right? Incredible simplicity, amazing output. And I'm just wondering, within Chime marketing, within Chime overall, within marketing overall, I feel like we add a lot of complexity in our toppings.

32:08Vineet Mehra:We add in data lakes and AI layers and tool this and tool that and stuff. And as a marketer, what does your marketing team overcomplicate with their figurative toppings and get away from the core basic ingredients that make Chime great as a brand?

32:24Josh Spanier:You know, you're so right. There's so much buzzword bingo, right, in our industry. You know, first, you have an amazing brand story that's relevant, that is narrated in a consistent way and sort of narrate it into the cultural zeitgeist, right? Not around it, not above or below it, but how do you connect your brand story into the cultural zeitgeist and connect it, right? That's job one. Job two that people, I think, is really allocating capital across both direct response and brand marketing. And actually, I think this is where I may pause and talk about where teams complexify things a lot. You know, for so many years, the last kind of seven, eight, nine, 10 years, there's been this sort of almost marketing civil war or like a Tupac, Biggie, West Coast, East Coast thing around sort of performance marketing and brand building.

33:25Josh Spanier:It's almost like West Coast is like we do performance marketing, right? And East Coast is we're brand builders. The honest truth is you can't performance market. And, you know, again, I'll come back to why I hate that term, but you can't performance market your way out of a brand narration and differentiation problem. And you can't use brand marketing to get out of a funnel kind of performance marketing funnel problem, right, where you can't capture intent. There tends to be a lot of just complexity around. But this is brand marketing. This is performance marketing. what I've tend to do is I've coined this term called performance storytelling.

34:01Josh Spanier:And essentially what I try to do is just kind of bring all of this together, which is the emergence of sort of performance marketing or direct response and brand marketing. And the reason I don't love the term performance marketing is because everything performs, right? Brand performs. It just does a different part of the performance, right? Performance marketing is actually just direct response. It's immediate acquisition of intent. And so both areas perform. And one of the ways we have got to simplify all of this is to break those silos and to understand that these two things need to work together, right?

34:34Josh Spanier:And I think AI is allowing us to do this in ways we've never been able to before.

34:38Vineet Mehra:Something that a lot of us in marketing say is we have to learn the language of the CFO. We have to speak C-suite. We have to speak in unit economics and this work and all this performance and brand coming together. Just translate, though. It's the, it's later than the quarter. Business is okay, but could be better. Your CFO wants results now. You haven't planned for that in the same way because you've got long-term ambitions and long-term budget setting and stuff. What happens in that moment? The CFO has come into your office or you've gone to the CFO's office and demanding immediate impact. What's the conversation then?

35:16Josh Spanier:Yeah, so, I mean, this is a very common occurrence. You get into, you know, so let's geek out a little bit on financial terms here, right? What I would say to a CFO. you get into these moments of sort of, you start talking about marginal return on your dollar, right? So let's say you're short the quarter a little bit and you have to make it up. There's only a few ways to do that, right? Either you get more dollars and sometimes the CFO will say, hey, if you spend a little bit more money this quarter, can we get this up? So in that moment, you have to be ready as a marketer. That's kind of option one to say, yes or no, I can.

35:51Josh Spanier:So there's been many times where CFOs come to me with this kind of a conversation, I said, no, thank you. I can't spend the money efficiently. That is an unbelievable answer to give to a CFO, right? Because you're not, you understand that if you add an extra dollar, what the marginal return of that incremental dollar will be. That's number one job. You've got to be able to understand at all times when I put an extra dollar in, what will my customer acquisition cost be of that dollar? Number two would be, do you need to reallocate capital from some of the longer term areas into the shorter term areas?

36:22Josh Spanier:And again, to do that, a lot of people will just say, hey, I'll take brand dollars and put it into sort of my direct response budget for the last month of the year and hit my budget and hit my numbers. You could do that and that actually will hit the quarter, but you have to be able to also model the next three quarter implication of that. Right. And that means you need the tooling to be able to blend your brand and direct response budgets into one area and have sort of the data science and the modeling to understand what the implications of that decision will be. So in those situations, I'll say to the CFO, sure, I can do that.

36:55Josh Spanier:Let's say there's no money. I'm going to reallocate. Just know that we may pay for that in this way in Q3 or Q4 based on the internal modeling tools that I've got. CMOs have to be able to sort of understand and connect their dollars to revenue, EBITDA and EPS at all times. And that's the only way you're going to be able to have this conversation with the CFO. The other part of it is, you know, we're talking about sort of the end quarter, but we can't underestimate sort of the upfront conversations that happen with the CFO every year, which is, hey, what do you need my payback period to be for the year?

37:30Josh Spanier:Blend it for the full year. What do you need my 30-day CAC to be versus the LTV to CAC kind of profile of our business, right? And that allows you to blend long-term with short-term. So this is in the context of much bigger conversations, but I couldn't agree more. I mean, the CMO has to be extremely CFO financial and balance sheet literate. And it's the only way I think you can have an effective conversation with a great CFO.

37:55Vineet Mehra:One of the things that I've seen and what's worked well inside of Google Marketing is you can learn the love language of the CFO, growth, unit economics, as you've been talking about. But it's also important that you coach and train your CFO on what they're asking for, because just dumping money in and getting wildly inefficient, that's a mistake as well. So there's a two-way thing. And so it's important that CFOs learn some of the language of marketing. It's not just a one-way direction. So I love what you said there. I think you're right on. I want to ask you some signal and noise questions.

38:27Vineet Mehra:We try and do this every week. So one or two sentences on these ones. Agency of record is dead. Signal or noise? Signal. Okay. Marketing departments with close to zero copywriters. Signal or noise?

38:41Josh Spanier:Signal.

38:42Vineet Mehra:Being good at marketing requires you to know how to code. Signal or noise? Noise.

38:49Josh Spanier:The agents will do it for you.

38:50Vineet Mehra:Okay. The impact of AI is kind of overblown in 2026. Signal or noise?

38:59Josh Spanier:Overblown. Noise. Okay.

39:01Vineet Mehra:Let's keep going here. The impact of AI in 2028 is currently underblown, if that's a word.

39:09Josh Spanier:Oh yeah, under Blanc. I think we're underestimating what's going to happen in 2028 and beyond.

39:13Vineet Mehra:Cultural relevance is the new CAC.

39:17Josh Spanier:100 % signal.

39:19Vineet Mehra:Okay, I got that from you anyway.

39:22Josh Spanier:Broccoli on pizza, signal or noise? Only if you've got some fennel sausage on it, then it's signal. Otherwise it's noise. Okay, that's great. Thank you so much for having this conversation. Thanks so much. Great to be here.

39:34Vineet Mehra:Before we wrap, I want to leave you with three takeaways to carry into your week. Here are your field notes from the frontier. First, the CMO is now a capital allocator. Marketing leadership is no longer just about campaigns or creativity. It's about knowing where the next dollar goes, how to measure impact, drive growth, and build trust with the C-suite. Second, AI changes organizations before it changes creative. The real shift isn't better creative or faster execution. It's redesigning teams with a balance of human judgment and agentic support. CMOs who think like systems architects will move faster, waste less, and unlock higher-order work for their people.

40:16Vineet Mehra:Third, taste still wins. In a world where execution is increasingly automated, differentiation comes from ideas, narrative, and cultural relevance. Perfectly executed bad strategy is still bad strategy. And taste is the multiplier that machines can't replace. Thank you for joining us. For research, case studies, and tools to help you lead through this next platform shift, visit thinkwithgoogle.com, where we track the signals, cut through the noise, and share what leaders are actually testing in real time. We'll be back next episode with another leader operating at the frontier. Until then, stay sharp and stay out in front.