The Most Efficient Marketing Org In The World? (35 People, $350M) | Tim Rutten, CMO @ Backbase

Tim Routin is the CMO at Backbase, one of the biggest digital banking platforms in the world. He's running a crazy efficient go-to-market organization, completely built around AI. I have a 35FT in total, running a book of business of $350 million and up AR. To learn that efficiency, Tim had to throw out the old org structure and rebuild the entire GTM function from the ground up. In today's episode, he shares a ton of what he's learned through that process. Make sure you have a landing zone, a good architecture, really understand how to do this, so you can make no-engineers dangerous. We also get into a bunch of other fun topics, which GTM workflows and tactics are going up sleep and how GTM roles need to be restructured for this new reality. Certain parts are being hollowed out, and that is not the career anymore. The career is. I loved this episode because, as you know, I'm massively AI-pilled, and Tim really knows the stuff. I took a ton of notes and gave me a lot to think about, and I think you'll really enjoy this one. Welcome to the Revenue Leadership Podcast. A quick PSA, there's a change coming to the Revenue Leadership Podcast. I've moved my show to a new YouTube account, so if you enjoy the show, I recommend you go subscribe so you don't miss any upcoming episodes and a bunch of other content that will be on there. You can find the new channel at youtube.com/@kyle-norton. The channel name is Kyle Norton, the Revenue Leadership Podcast, and that's where we're going to be posting all the future episodes, clips, and a bunch of new stuff as well. If you're wondering, the audio podcast will still be in all of the same places that it's always been on Spotify, Apple Podcasts, SNPs, and so on. If you listen in audio-only form, no need to do anything, but if you watch on YouTube, now you know where to find future episodes. While I'm at it, I would love if everybody could support the show by liking, subscribing, reviewing, leaving a comment on whatever platform they listen on. It helps new viewers find the podcast, it's really helpful for us, and I always forget to mention it. If you love the pod, or even like it a little bit, do us a solid and go review it, comment, share it with a friend, and that would be most appreciated. And now let's get back to the show. Today's guest is Tim Ritten, CMO at Backbase, where he spent the last decade helping build one of the biggest digital banking platforms in the world. He's recently launched its AI Native banking OS, and Tim has been building the same kind of AI Native operating model inside the GTM function from the ground up. And so we're going to talk about what AI Native leadership actually means day to day, why individual co-pilots or table stakes, and why we're really past them now, and what it takes to redesign a revenue org around AI principles and combining agents and humans, and we're going to unpack some really specific examples that I'm excited about, so we're going to get a pretty tacky along this one. So Tim, thank you for joining and appreciate it. Yeah, very welcome, Kyle. I'm excited to be on the show. Maybe for starters, the more you say AI Native, the more people will probably get a little bit iffy or maybe even a little bit annoyed by now. Like three or six months ago, it was the cool thing in town, and by now people are probably getting a little bit hesitant. And I hope that through the conversation, we're going to show a little bit more an example, at least to be able to show some examples on what it really means and why it's not, it's not a buzz word, not at all. Yeah, I'm excited to unpack. These are episodes that I always enjoy doing where we can really get into specific use cases with people who have been deep in the weeds, building stuff themselves. I have a bunch of questions about what you previewed to me on our intro call. So let's just start there. Like what does AI Native mean to you? Like it is the buzz word to sure every founder wants an AI Native revenue leader. People want to be that. So how would you define it and what does that actually mean to you? To be fair, I think the starting point of this whole AI Native wave that we're currently in on how to organize and even our value proposition is an AI Native banking almost. What does that actually mean? It's more of a principle of architectural most. Like how did you build up your organizational model? How did you build up the rules that are working in that organization? How are they leveraging the technology that's around to do their work and to do it in a different way that is actually a native meaning? If you and I would start a company today, Kyle, I would do it vastly different versus starting that company two years ago. I would start with primarily, okay, so there is AI. How do we organize ourselves and such a way that we get the maximum value out of that technology? And again, it's a container term. So there's a lot of different technologies and a lot of different ways to apply it accordingly. But the essence of AI Native is that your systems thinking by design from day zero. So if you didn't think about AI powered, AI first, AI, what have you, they're not going to get the full value or it is not going to get the full value out of what the technology actually can do for you. So AI Native is really, it's really a thing. It's really a different way of looking at the same problem. I'll give you an explicit example. There's many centers, many AEs, BDMs, marketeers that use their own GPT instance, BIT clause, BIT, GED GPT or Gemini, and they do certain workloads through that interface. It probably helps you do your work better, faster and sometimes even repetitive, but it's not systematic. It's just for the individual and it's as good as the way you've organized and let's say set up your instance. It doesn't help the organization as much. So if you're the senior leadership of that group that's all playing with these different types of tools, how do you make sure that the quality is there? How do you make sure that silly work is taken out? How do you make sure that you have an end-to-end view on how the operations are actually working? And how do you make sure that you take out the stupid work? That's not what the individuals in the organization are going to do by themselves. So thinking AI Native means, top down with a very clear mandate, how are we going to completely rewrite the organizational and operational model, and that literally starts at, we're going to throw away the org chart. We're going to look at the actual work that we need to do and we're going to re-invision it from the ground up. Versus, let's say, stitching it on top of something that we already have running for many decades and there's politics and hierarchy and what have you. So that's what it means to me. It's very transformative. Backbase is an older company. It was a pre-AI company and so you're going through this transformation and rewriting things from scratch. So what has that meant for the org chart? I think that's a really interesting example of thinking in this way. So how did you look at the jobs to be done in the org chart and rebuild things? Yeah. So the first decision I took together with my CRO was if we followed a hypothesis that I just shared or at least the philosophy of you either go all in or you don't go at all. Like in the middle, you're not getting the full value, but there's a lot of upheaval in York with these tools and everyone needs to do their little thing. So we basically said we're going to go all in. So we're going to assume for a moment that this whole agentic workforce and rewriting your processes in the org chart that actually holds true. So let's just go all in. Let's take the bet almost. The low was a little bit more informed, of course, than what I'm sharing here. From that moment onwards, I could look at the organization as a system as an engineering problem almost. How do we make sure that our value propositions, our ICPs, our segmentation, our go-to-market strategy, the intel that we have from dialogues, in actual conversations in the field? How do you make sure that all that context is properly organized so we can start driving certain activities that immediately cross all boundaries of all teams? And that was literally the call number one. If myself and the CRO are fully behind it, we're going to break the barriers of the typical team setup. Because if you drive go-to-market holistically, sales, marketing, partner sales, customer success, it's all the same. They are all working against the entity of the customer, the prospect, or a partner in the other scenario. And all motions around it, whether it's driven by the marketing craft or the sales craft or what have you, they only to operate on the market. And that basically became the go-to-market OS, which we'll talk about a little bit more, which is literally a system, an application that is running on our premises that runs the full operation, front to back. There's almost no exceptions anymore. What that does to the org chart is very interesting. I was reluctant to change it in the beginning because there's so much change already. People are very much fatigued with what's out there currently, all the hype, all the big buzzwords, all the massive boats on LinkedIn and people are replacing the marketing teams or the sales teams with an agent, which is not true, just so you know. We decided to slow down on the let's re-orc and really flip it to a supposedly an native model and actually bring everyone in on the learning journey. And I think that part was critical of making sure that we didn't lose our people. I think that's one of the main insights that I'm coming back to over and over again. You will need the UMK capability to drive the whole system into the right direction. Is it the same people though? You need people to drive this org chart, but not everybody is going to make that transition to being a systems thinker and like a love of being in tools. Have you had to change over much of the team to try to get to this place? I would say that right now we're organizing more around logical units of work and collaboration, which means that I'm actually collapsing certain functions. So on the marketing side, as I'm the CEO, classically you have a brand function, brand a creative, then you have content and digital product marketing, the field, ABM events you can go on. like it's all these different desks that have a little piece of the pie. That has been collapsing every other quarter into, yeah, let's say streamline buckets with one single leader. So eventually my leadership team also is going to have a way simpler setup because you can get a broader remit and basically bring responsibilities together that can logically be clustered now because you literally can. Let's say the posts that are out there talking about the fact that certain roles are being challenged. It is true. So they are being challenged and some are just honestly redundant. It makes no sense to do desk research anymore. Why would you? Just fire up an agent, utilize XA, maybe combine with fire crawl and you'll get the best damn research out there. There's just no better. No human can do it better. If I think about Princess the BDM role, certain parts are being hollowed out and that is not the career anymore. The career is how do you strategically penetrate the top accounts and how they build a relationship using the infrastructure as giving you the right signal and the right, let's say, detail on a certain account that's warmed up with exactly the right angle to pick the phone up for. Or you know exactly where the particular stakeholder is, maybe this individual is speaking at a keynote event, make sure you are there with all the ammo of the research that you have coming from the infrastructure. Even though that sounds very logical and easy to say, still within the actual individuals that hold a certain role as being challenged, they feel a little bit scared. It is very overwhelming to see this whole machine rolling into the org and literally taking out the work. Humans aren't great with change. It's not in our nature to be, it's not in our nature to be great at dealing with upheaval. We're adaptable as a species, but on an individual level, we crave consistency and stability. What you see in many organizations is this thrash of resistance because it feels like change is being done to them and not with them. How have you gotten everybody to come along on this journey and change the nature of their roles or be collapsed into different units of work? What's been most important to get people to come along this journey? What I did last year and we're talking end of Q3. This is the moment that I myself got into the terminal, fired up cloud code and basically the universe opened to me. This is before it was holding happening on the LinkedIn channels I would have you. Because my CTO basically said to him, you have to get into it to get the understanding of what's coming. Ever since I was completely sold and it switched my mindset almost immediately and I applied the exact same logic to my marketing leadership team and also very soon after to the rev-off team. Like team, I'm going to get you a premium seat with Cloud Premium, which means you can go into the terminal, fire plot code and just do whatever you would like to do. There's no boundaries. There's no target on it, no ROI, whatever. I just need you guys to see what I'm seeing. That's exactly what happened. Within two weeks, people were building prototypes like crazy. They were connecting certain APIs and got analyses out that, you know, originally you couldn't because of Sys' boundaries or integration work or what have you. And something very interestingly happened. I was, of course, doing my part, which on the content side and strategy side was just top notch, best quality I ever was able to deliver as an individual and next to working with my teams, of course. But the exact same thing happened to my marketing leadership team. So all the six, seven leaders that were running a classic operation saw the light like, okay, so this is coming. This is what it can already do today. And they completely rethought how they would love to operate their team going forward. Not having the answer per se, but at least the click was there. That this whole AI conversation is not just a high level, boardroom type of thing to cut costs. It's a very real thing and it's actually incredible. If you embrace it early and you start becoming a leader that has this in their back pocket, just like you and I can work with Microsoft Word and Excel and Google Sheets, whatever. It's like a standard skill. At this moment, you probably still have it on your resume. Like I'm super AI native and I can do a cloud. I think it's six or 12 months. It's not there anymore. It's just normal. If you don't have it, then you're really working in the wrong world almost. So that was the first part where I internally got a lot of momentum. As in people started pinging me on Slack from the full organization, like, hey, Tim, going to get access to your repulsory. I heard you guys are doing some crazy stuff. And it just kept on going. So that an initial getting people excited is not me telling them it was actually me opening up the door and saying, hey, you're going to get this expensive account. Go for it. And then we just basically went through the whole revenue organization. So we had around 20 or 30 people overnight that we're doing this type of work and it just seeded like crazy. And what were the big structural challenges that needed to get solved? Because you give everybody terminal, you give everybody cloud code. And then invariably they hit a bunch of these roadblocks where, you know, the quality of the outputs, not what you want it or they can't do certain things. And some of those things need to be fixed at like a structural level. What were some of the pillars that you found you needed to put in place to get to get the maximum benefit? You mentioned context engineering as one of them. But what were those other key things that you feel like needed to be in place? I would say that the real response to this question is, this is not a great approach to scale it at all. This is a great approach to get people fired up and see the power which by now by today it's 100 times more powerful. So people just immediately see it get the ah-ha and off you go. The main move I was making is get people changed ready and actually make them the champions to drive it forward. All I was not the answer. If anything, I wanted to get people out of the terminal ASAP again. So around two to four weeks when we hit that mark, we basically put it to a halt and we've got enough momentum and mental buy-in from the full org to start properly investigating. Okay, how do we bring this to life so that the full organization can run at the same level of quality? So we can actually tackle the challenges that you would typically have like data consistency, who can access which data, who can actually create, read, update, delete, sales force. How do you really bring this together? Not a regular terminal instances and plot co-work instances. That is simply not scalable in the context of how at least we run the business. So it's just the initial breakthrough which all the way ended up in the boardroom and the management team like, okay, what you're doing in that corner together. Can you please present it? I think it's very meaningful and maybe we should actually consider doing it across the business. It basically gave me enough confidence to aggressively invest in a landing zone where you could actually start building GT MOS, like a proper system where people log in with the rights and entitlements that they have can do their daily work. Whereas the maturity of the work is actually driven by other agents or automations or workflows. Which of course is a bigger step to take, right? That takes time. It takes also a bit of brevity to go through that door versus just going for the frontier labs that have, let's say, seed licenses and basically give individuals to a degree super power. So this is what led you to build the GT MOS as the core harness and interface for all the GTM work. Is that correct for us? Okay. So maybe just let's start diving into the operating system. So what did you build? GT MOS or Go to Market OS is basically our internal term for what we built in-house to basically have a very robust landing zone where we can build any type of application, any type of agent that you can work with, be it conversational, be it trigger based, to actually bring it into a landing zone so that we could deploy it properly to all of the people that we are collaborating with across the Go to Market organization. So if you think about it, if you put it on the screen, on the left there's a sidebar which is actually mapped against capabilities and themes and programs. And then on the right you properly have an interface that allows you to do the type of activity that is part of that particular team or part of a particular program or part of the particular reporting angle that you're looking through. What is really critical here is that the creation of that landing zone was actually the biggest hurdle. How do you get compliance sign off in a non-regulated business but working for regulated banks? That's of course where we are playing in the industry. Make sure that security compliance is completely okay with the landing zone that you've chosen for and want to go for because it will get connected to your CRM in our case that Salesforce. It will get connected to external data. We use a lot of let's say research tooling and newer network search tooling like XA. You bring all that data in. How do you make sure that that's completely, let's say, signed off in the sense that you're not doing anything silly in terms of the business or risky? The next thing that needs to be completely safe when it comes to authentication. So we actually leverage our enterprise authentication services to get people signed into GTMOS. So we know who they are and we give them access to their specific apps and agents. And from that onwards, it's just like you're using chat GPT or a cloud honestly. It's actually using the same frameworks, but it's completely ours. So where typically you would, let's say, stand up your chat GPT or cloud desktop environment or your core work environment, you start connecting services, right? You start connecting Salesforce or HubSpot or your email or your calendar. We best. do that at a enterprise level. So we have these tools under the hood, fully connected, fully guard-reel and metered. So you can't go crazy and pull everything from sales force in one go. So that everyone in the organization can do the work that's relevant to them at the highest quality level. Well, we can continuously keep it tap on what's actually happening, what's working well. And also some tools just change. Like tools that are great for outreach today might be really not great for tomorrow. And we can easily swap out the tools and the actual team members working on top of the platform or the agents that are consistently working on the platform. They just pick up the new tool. That's it. In essence, you're looking for an A&A dev stack or an A&A dev landing zone or I think for cell calls at the A&A dev cloud. So where do you deliver and deploy applications and agents and everything with it that is completely leveraging the A&A dev way of thinking. For cell is a fantastic example for this. It's our standard on the GT MOS and we leverage the full stack of our cell, which every other week is getting more and more complete to hit our objectives and additions. So when we start it, you could easily leverage the cell to indeed deploy your application, which is next JS with TypeScript and React and let's say notice the package builder. All good. That's not really special. It's been around for many years, but it also introduced the AI gateway. So you basically have one API that gives access to all the alarms out there for which most of them are zero data, a retention policy driven, meaning you can bring more confidential data to the table, but you can do it at scale. It also gave a proper framework or a reference framework architecture if I may, that basically helps you structure. How do you set up your agents? How do you structure your prongs? How do you give agents tools? How do you manage those tools and version them? How do you make sure that you have all these ingredients properly set up? So you're basically engineering whatever the agent has access to, yes or no, and what the quality of those, let's say elements and parameters are. So it for cell or basically getting more and more into that stack as an example, so workflows became a new capability in their stack. I think it's two months ago, completely native, which means it also in your build. You can actually build workflows with natural language. So it builds agents, it builds workflows, it builds tools. Those are all the critical things that you need to fire up an agent or any automation workflow. And before you know it, we're now six months ahead into this game. I think we're now at 30 to 40% of completely automating and rewriting the way we operate as an organization. With Versel, do you get access to some of the capabilities out of the other tools? In Codex and Cloud Code, now they can spawn off sub agents to do a bunch of other tasks. Do you get access to all of that through the gateway? Or do you have to build those primitives inside GT MOS? So this is what Versel actually delivers, because the gateway is the gateway. You basically have all APIs, all other lands to hit them and get a response. So you need something in front to actually do that type of work. And that literally is the AI SDK from Versel. So that basically does all the spawning of agents, sub agents bringing them back, collating the outcome and then taking it forward. It's also not per se only in that context because for much of the work that's happening in GT MOS, it is actually not per se AI. So if you look at the way we do rewriting, which is completely consuming in headless from Salesforce, we built the reports with Cloud Code in a cursor setup, within this architecture. However, if we then deploy it, it's just an XTS application, which consumes the APIs from Salesforce. By now, it has a bit of a data tier in the middle, like a semantic fabric that basically has all the latest and greatest data from Salesforce. So we can do it at rapid pace and performance. So we don't have hundreds of people hitting the Salesforce API, but that's about it. So there's nothing agentic or nothing per se AI in the actual interface that the individual user see. By now we're applying it on top. So we build it with AI, of course, like agentic delivery. By now, we have all the data all reporting completely in place and then you can start to bring in, let's say, reasoning loops to have a view on pipeline quality. Pipeline discipline. How fast is the velocity of different regions, different types of opportunities? Where do we see based on the patterns where a deal is being called? But unlikely to happen because we see these patterns in, let's say, the conversations with the actual prospect. So I think also there, there's a bit of a misunderstanding that all of a sudden you need to do everything AI in the sense that everything needs to go through the LLM. I would say that almost almost 70 to 80% of GTMOS is not LLM based. It's actually very deterministic, but there's automation and workflows below certain activities. And there's a running that do certain activities as well. And they hit, let's say, LLM through an AI gateway. So also there, it's what's the actual balance of what you actually utilize. Interesting. So how many tools did you deprecate? How many tools could you like stop using because you've got this solution in place? So last week we deprecated a tool for 15k per year. Beginning of last, no, this is Q4, we were able to not renew six ends in this particular scenario because we completely built a signal engine bespoke. There's not black box, but white box. I know exactly what is happening there. We define the scoring rubric all the time. Like we're really tuning the model ourselves. And that's by the way, it's our model. It's just such a, let's say, mind-opera at the moment you have that capability to build it, to bring it to life, and it connected to the other building blocks that you have a vision for initially. And then you get take off because you have some foundational stuff in place. It's actually, it's going to give you alpha. Like go to market alpha almost. So these are two examples already. And there's many more coming. They're on the list. So before the end of the year, we're taking out at least 100k extra. And so when you're building new tools and applications, you're doing that in the VERSEL experience and then deploying it in GTMOS. So the actual build can be in any type of IDE, like your developer environment. We leverage for the majority cursor. A cursor for the reason that it has a fantastic debugging flow or debugging capability, which is using their composer model, which essentially goes to the code very fast, puts debugging items into the code and then tests it and fixes it completely autonomously. But for the very same account, we could just build in the terminal with Cloud Code on. So basically, it's the decision of those that are building within the GTMOS repository. But from their onwards, it's not more advanced than it's literally committing against a GITA repository. There's all set of checks that are happening, security checks, build checks, quality checks. And once that's, let's say, pull requests reviewed, it automatically gets deployed in VERSEL. Also does certain checks and then actually promotes it to production. So there's this whole pipeline by now that is, yeah, make sure that we have quality that we bring to the table with every pull request that we approve. How many people are working on this full time? That's the funny fact, because full time, it's actually only one individual today. And then part time, it's individuals like myself and two to three other builders that are spending sometimes enough to do them with the majority evenings and weekends, because we're just really in there, really excited to be able to build. But finally enough, that's where it ends today. And already you see that type of takeoff on what are the chiefs within the context of backbites. We are in markets for more builders, but if you think about that profile, it's like the wildest profile today to recruit for. So I think the beauty is mostly to make sure you have a landing zone, a good architecture, really understand how to do this. So you can make no engineers dangerous. That for us has been a major step. It's you and two to three other people building part time. Are those like revops folks or those like business owners who what's the profile of those other individuals? It's myself, but I have an engineering background because I basically was running a web design agency for 10 years when it was still interesting. Very helpful indeed. Also ran web hosting businesses. So I have a bit of a system's engineering view as well. So basically I have a head start. I know the business. I know the marketing craft. So with that me briefing someone to build on my behalf is actually already inefficient. But secondarily we have a GTM engineer like a proper one who's basically running the full show full time. Our director of revops is building. Let's say part time. Our director of content marketing is building part time. And we're now trying to get one or two more business people that also have the neck of systems thinking and building because they've been doing their own little prototypes for for the past few months. We're trying to onboard them as well in the architecture so we basically get even more critical mass to make faster progress. And it's showing already. And so what is onboarding somebody look like? Do they have to learn computer science primitives and like start to understand how these systems operate? Is it about just learning to use the tools and giving them a cursor account and showing them how to like build an e-vow or a tester? What does that onboarding look like to you? I think first of all the people that you try to onboard they should be naturally energetic to dive in and learn regardless of where they are in there. Let's say learning. cycle. The second point would be under the assumption that your architecture is sound and it's properly set up for, let's say, agente delivery. You don't need to be fully computer science level at all to build certain capabilities and requirements. Not at all because the repository itself will indicate to, in this case, Claw, which is our favorite LLM currently for building. Hey, this is how the architecture works. Hey, if you're making UI components, use this library. Hey, if you want to have access to Salesforce, use this particular tool. No other tool is allowed. So basically, you have all harness in which the building happens, which means that you're almost abstracting away all the complexity for any business, let's say, person to just interface with Clawed, build requirements, start planning, which you will do completely within the context of the architecture and off you go. So in my view, especially within the coming three to six months, there's going to be a moment that anyone could certainly build within our architecture. That does assume proper architecture, proper setup and pipelines and how requirements, plans, test-driven design, domain-driven design. How all of these best practices actually come together and are abstracted away for those that know the business that would like to build. Needless to say, still, you need to be comfortable in building, not an expert, but comfortable and interested. And then you can learn incredibly rapidly. Yeah, I mean, you just ask whatever question you've got to your LLM of choice and really explain the frameworks we've chosen in Y and explain it. You know, they could explain the entire code base to you just in a conversation. And it's also hard to look at all of this. So when I said you go all in or nothing at all and you remain in the middle with everyone has their own little instance and they build prototypes and you send HTML files to each other. It doesn't scale. But why the ah-ha is there is that people that are running a role or an operation have the best ideas because they are the ones that have certain challenges and they basically get mind blown and think, hey, what would happen if I would do X, Y, Z? Let's do a prototype. Let's try to stitch it together. If you cannot have them land on GT MOS, you get that type of firepower to build an engineer and actually get it to production. And it's the actual business leaders. I mean, that is that's IP to me. That's special. That's how you have a competitive advantage in how you go to market in my world. That's exactly the thesis. Overinvest in a great architectural landing zone so that the people that know best what we could actually automate and you know, maybe have complete your genetic processes in place as well, that they're actually the ones on the buttons that they can do it instead of knocking on the door of GTM engineering, which I don't think is a functional zone. It has longevity, assuming what I'm sharing is true, which I see enough evidence of. So, say more about that. Why don't you think GTM engineering is a role with longevity? Look, it will have longevity because with one individual you can go very, very far, but it will become the bottleneck. So if I look at my team right now or the setup we have right now, there's one proper full time allocated engineer, but the only way we're going to continue scaling is the circle around it that is prone to build and can leverage that architecture. If you don't get that circle in place, you need five to 10 GTM engineers and it will become very, very slow. Why? Because they will not know the business. They are not running the man generation. They're not running outreach. So you need to brief them. Then it goes back and forth. And even though it's agent delivery, so it goes supposedly a little bit faster, you get meetings, you get slide decks, you get status updates, you get all the mum and native type of behavior. So you either know the business you can build or you don't. If you don't, to me, eventually, it will not fly because the interesting fact is that the competitor next door, if they get that formula properly cracked, they're going to go 10X compared to you in terms of speed. So I do believe in a GTM architect, like a heavyweight senior player that knows their stuff, owns and governs the architecture and the pipelines and everything with it. Maybe you have to to make sure that you always have redundancy and quality control in forward principle. But also, I think those are the two I would invest in heavily and then I would make sure the outer circle is enabled to build and go have speed. So the GTM aren't to protect is basically building the harness and the platform. And then the person building the tooling are the business owners, the agents, the workflows. Correct. Correct. And think a GTM architect is always also more fair to what I'm seeing right now in actual daily execution because engineering almost seems like, yeah, they're just classic building. No, no, they're architecting the system in which it's being built by agents. A hundred percent of the GTM OS code is generated, which is already on its own is a crazy fact. It's a different riff. On sort of like our approach and what I've been sort of espousing, which is a centralized version of AI and go to market. Because what I've generally seen is either people give everybody a cloud license and they just say, go nuts and then you end up with a ton of skills locked in people's individual cloud or chat instances and there's no scale to it. And oftentimes those things aren't built very well. So I've preferred to build with a centralized team where people go and access that expertise that's going to build things and then deploy them generally to the entire rep population. But this is a different version of it where if your harness is good enough, you can actually turn all of the leaders in a business into builders. But understanding how to qualify or certify to get to building is still an interesting question. It's still quite centralized because it's only a handful of people. But don't get me wrong. I'm not making a statement that everyone should be a builder and get into the repository and have cursor and start building. Not at all. You need a subset of top players, maybe five to eight over time that are committed to the whole deployed environment and they're basically looking after their business lines and making sure that certain capabilities get built out. And then a lot of, maybe not power, but another innovation power gravitates to those individuals because they're going to get the hang of things and because they're so close to the business and the fire, they also know where they're going to make the business most happy because maybe it's even themselves directly or it's very close in their proximity. So as an example, our director Revops is building is fully responsible for the reporting, let's say, parts of GT MOS. The reports, also the dynamic interactive and automated reports that we have by now, they run the full pipeline generation and pipeline management cadence. Nobody's in sales force anymore. It took him three months in total. To build the whole reporting stack native into GT MOS. Correct. But then fully in the way that we actually operate, including automated signals, automated emails with end of week status reports, pipeline drift, what have you? Like the full operating model, which you could have a vision on, let's say, year or two ago, that you would implement in many, many years. We now implement it in three months. Just imagine that individual and that role completely fired up. The sky is the limit. Whatever you can think of, to get it with the team, I just saw them having their QBR, they will deliver next week. And it's rolled out, it's deployed, and it's going to change the way we operate all of our go-to-market teams and sales functions in the different regions and territories. So with that one individual, we got that 10X impact. I don't per se need to in the Revops domain. One is enough. Same thing in my marketing work. And how big is Backbase total? 2,000, 2,200 individuals. So still quite a lean team for that size company? I would even argue that I might even have the most lean marketing work at this scale, I have 35, 50 in total, running a book with business of 350 million and up ARR. So if you divide that head count wise, we are very efficient and very effective, I would argue. If I look at go-to-market efficiency or the magic number, so trailing 12 months, new ARR divided by sales marketing spend. Every dollar invested, we generate 80 cents in ARR. And ARR is five years. So it's actually eight times five is a nice 4X impact. You're in Back-to-Back meetings all day. You're trying to stay present, but you're also worried you'll forget the decision, the action item, the important next step. And that's where granola comes in. Granola is an AI-powered notepad for meetings. You jot down rough notes like you always do, and in the background, granola transcribes and turns them into clear useful notes when the meeting ends. No bots joining your calls, no distractions, just a clean notepad that helps you focus. Use code "The Revenue" for three months of granola for free, so you can head on over to granola.ai/therevenue to learn more. Again, that is "The Revenue" for three months of granola for free. Check it out at granola.ai/therevenue. What else did you look at besides Versel as this hosting platform? What are the other options that somebody could think about to build this harness operating system/UX? We were looking into Google GCP and Vertex, so that we are on the Google suite as a company, so it was Pretty. logic for logical for us to look into that stack. However, they were standalone components, and the SDKs that I'm built with and the ploy were not as, let's say, fluent as we found with Virsell. So can you actually, in your agentic building environment, have the agents work against the environment to deploy certain things, to set certain prawn jobs, to deploy the database migration script and execute all of it, run the text that wasn't as mature with GCP at that moment. And then very soon after we just landed on Virsell because the prototype was just flying already. And the main challenge we had to go through is say, can we get a signal from this? Because it's not a very enterprise grade stack, at least not at that moment. But by now, Virsell is certainly becoming or very rapidly becoming an enterprise stack. So that was really those are the two options at the table. There were maybe a few others in the long till, but they didn't pass my table. And did you look at buying any solutions like this in the market? Like, are there really solutions in the market that purport to be this harness and system? For sure. So if I rewind back the time to January last year, when our CEO said, we're going to go risk on, we're going to go all in a little AI internally and in a product proposition. I started doing my runs like, okay, what's out there currently? What could we potentially consume? So I had many pitches, I had many demos, and with all of these demos and pitches I felt this makes no sense, pricing wise, like hundreds and thousands of dollars. And it only did one little vertical of the actual vision I would love to get to. So in essence, I would overspend on IP that was not really changing the game. And it was almost overnight expensive because it was AI. So there was just a mismatch between what am I looking for? What am I going to pay for? And by the way, you're a startup and clearly you have no attraction. So no, thank you. So with the majority of the ones that we were testing out, we felt there's no mode for these guys. So why would we buy into it? And the more we went through that cycle, I felt you can think as a CMO and stay hands off. And I don't know, I've got this portfolio to manage. And I'm a leader. I'm like, no, we're going to get into the terminal because I think we can actually build certain things on our own. I probably get to a breakthrough level within three months. And that's exactly what happened. And the first problem that we tackled was, look, there's a renewal coming up from the six ends platform. It doesn't work for our motion. It's a great company, but it doesn't work for our motion for a variety of reasons. Let's just not renew it. So we terminate it and we have an issue because it informs our campaigning and informs certain parts of the engine. So within two months, we had to fix the issue. We need our own single engine. And that's exactly what the team stepped into and started building. And I think we got a lot of conviction out of that build after those two months. And then we just kept on moving forward in that direction. So you said that you reorganized the team around logical units of work. Can you explain that and how is the team organized today? So when I took over, I reorganized the team in brand creative and product marketing, like the strategic column, more of an HQ, how do we bring it globally to life. And then in the middle of the organizational model, we have digital, doing the always on, the digital campaigning, the content function and the ABM function, which are essentially the regional marketing teams. And then all the way on the right, we have marketing operations, which we actually clustered into sales and marketing operations, like the Red Alps move. The bucket in the middle, that one is collapsing. So digital content, ABM, to me, are one thing. Because you you classically have high level demand generation activities happening, and then low level ABM execution happening. And that divide or that connect between has been very hard to execute in a human led model. With an AI led model, it can actually run as exactly one thing. So that's why we're seeing that collapse happening. And so then you'll basically have two pillars. You'll have brand creative product marketing and you'll have digital content and ABM and those will be the two, correct, like nodes of the team. And my current vision is brand creative is really a pillar because that is a certain craft and it's very much out there, different types of levels of storytelling that need to happen there. I think product marketing is mostly evolving into the go-to-market portfolio function. How do you bring all of these different go-to-market products, value propositions, and so forth together? Well being super close to the customer. And actually, that's going to be 80% of your job. You're going to be on the phone all the along with customers, executives, buyers to win loss analysis. So whatever you bring into your go-to-market execution engine, which typically was slide decks and documents, so on have you, you're generating that with the intel that you have. But then you better get the intel. So that's a certain twist on that craft. To go deep on, you need executive exposure. If anything, you almost yourself need to be an executive to be able to level with these people because we sell at the port level. And then in the middle and need to your point that collapses, if you ask me into, it's basically a dimension in EVM function, ran as one. And not multiple layers with multiple activities, multiple road maps. I don't know, it's one single road map and it's one engine. One customer journey, one funnel, or different channels. Correct. Which also at the same time is very demanding. At least for those listening, if you're in the position that you can run that full portfolio across all elements of those crafts, how do you drive digital properly? How do you drive content properly? How do you drive AVM properly across the full stack? And strategically leading the team and having depth across everything while doing is in an A&A this way. I don't know if there's many people out there that have done it before, but if you're excited, you should call me. And they're continuing to collapse and continuing to have one person be able to orchestrate more just because of the tools you've put in their hands. Well, absolutely. And maybe to give an explicit example, currently we're probably towards end of July, I'm going to have it available to the teams, where basically scanning the 3000 accounts that we go to market for. It's a very specific dam that we have and they're mid to large scale banks. With the GT MOS, we're now in a position where we can actually scan those accounts practically 24/7, map them against their ICP, where the ICP we can also refine all the time based on what we learn in the market and where we see deal velocity and what have you across all segments. And using that, we can actually put these accounts in certain buckets, strong fit, moderate fit, cooling down, no fit. So from their own words, we have a subset of accounts specific to an ICP match with our solution lines, which we can then outside in research, which also is almost instant. And I'm talking hundreds of accounts still, if not thousands, to understand where are they investing, where are they on their journey, combined with the signal that we have from our own signal engine, are they interacting with our channels, with our sellers, with our partner ecosystem on have you. And using that information, to me, it's like super proprietary data, we can campaign against them appropriately. Either at the highest level, let's say level more one to many, but then cluster based. So we don't do one to many to three thousand. We do it to 30 accounts, based on a thematic cluster that we've found. And basically that continues throughout the whole flow of our demand generation engine, which has nothing to do with how the team is structured. Like nothing. It's basically outside in how do I want the operation to function? By the way, first principles, what is logical to do, well listen to the market, find whatever is out there, combine it with the conversations you're having, which are practically stored in our CRM, 21/7. And start making a hypothesis and philosophy on what themes and what clusters are these accounts investing in. How does that then map against our value proposition? And then all of a sudden, I've got these three thousand accounts nicely bucketed, nicely themed, mapped against where we can actually enter with the highest propensity to buy. And then I have campaigns that all of a sudden become super targeted, super specific. There's no way that I could have done that or that I can achieve that just going through the human collaboration model. It's literally impossible. You can't research three thousand accounts at that level of quality. You then can't bucket them, you know, with individuals in their spreadsheets. So it's really systems thinking and engineering to solve them, to go to market issue if I may. And then running this every week for around 500 dollars at max, I think we're going to get it to 50 very soon. That's just out of your world. I want to break down some of those components. So like pillar one is you're just constantly listening to the market. And so what are you listening for and what is the system by which you're like bringing in that information and synthesizing it? Just like give me some of the architecture there. If you think about the type of account we sell into, they are mid to large skill banks, and we sell to the C suite and we influence the C minus one layer and sometimes C minus two. But basically it's C and C minus one. We're using tooling like exa, which is neural search. So it's not just a regular search API. It's actually a genetic search, meaning you give exa the objective to search for, let's say, the senior leadership of a given bank, let's say JP Morgan Chase. And it just goes out there and it reasons and it challenges itself on the outcome until it has a let's say high probability quality outcome. And it goes lightning fast. So that's one of the tools where if you think about account research, account mapping, finding the DMU members and so forth, it's second to none at this moment. There's probably a reason why Anders and Horowitz is big as an investor into them right now. It's like the new, the future Google, if you ask me at this moment. But then again, you need, you have to have the philosophy and an assistant to land it in, otherwise you're going to just do a account by account. So we do it in large scale sweeps. So we scale up the process. We hit many of these accounts with a collado led research. And then given that you have, let's say, a thousand accounts and a thousand, let's say returns on, let's say that research data, you can start finding patterns. You can then apply a scoring rubric. Like, okay, what are the parameters that are actually driving the match with the back base value proposition? So it's like a regression analysis almost. And typically get four to five scoring parameters, which you can then dial up and down to put accounts into certain themes and into certain buckets. Then all of a sudden, you've got your own, not mental model, but let's say data model, where you can just play, where should these accounts land, and why? And as we be the anglia you were looking for, like what do you then research? We research strategic projects. What are the merit investments at board level that they're going all in on, which are named, which are budgeted for, and have an executive against them. Like classic sales one and one almost, it comes close to rent in a way or a medic. But basically that's the game for me at least, in my motion at back base. So what ends up happening, let's take JP Morgan-Jase, they have, let's say, five to 10 thematic investments, and a typical executive will have one or two of them in their portfolio. And some will be cross-functional so as multiple executives. We're now able to execute this at a level where, on a theme level, a particular bank can be in multiple clustered campaigns with different executive titles. So I'm actually campaigning against JP Morgan-Jase from three different angles, with three different personas, three different campaigns, three different, and so forth. So I think that almost collapses the whole one to many, one to few notion, because this is actually one to few already. Because you're making smaller clusters that actually globally hold true. So in a way, I'm rethinking how the mansion can be done in a world where building and analysis and research is almost for free. It isn't for free, don't get me wrong, but there's no limitation. Just explain what back base does quickly, because I want to ask some specific questions. So back base is a white label banking platform. And recently we announced that we're basically the first ANAID banking operating system, which means any bank that needs to either transform their operation to become more effective efficient or needs new web channels, digital channels, mobile channels, or wants to bring agents to life, they need a certain stack to do this on top of. And that is back base. So back base typically enters mid to larger scale banks that have hundreds of legacy systems that they need to tie together to actually serve as the customer or help the employee in the customer call center, do their work, or the branch representative. Elastically, that's all stitched together. And if you've got these broken apps with back base, you have one single operating system that basically orchestrates the customer experience, the agents, as well as the human employee that's interactions. With that in mind, what are the themes that you're watching for? So you're watching the entire market for the strategic projects that are mentioned in the news or analyst calls or Twitter and trying to understand who has projects that are relevant for the pillars of your product. Is that what I'm hearing? - Yeah, that's exactly it. I have to say though, this is the obvious one. You take the analyst report, and you take the press releases, analyst reports will have you. But on top of that, there's a lot of signal around the individuals and the executive team. Take their LinkedIn, take where they are speaking, they are on podcasts, they are on certain podium. Try to find those nuggets, which is a combination of the exit tool, a fire crawl and a few others. They basically get that surround sound view. Like what are they talking about all time? What are they passionate about? What's their focus in that sense? And what is beautiful with an LLM? Once you have all that raw data in, the reasoning is exactly the power of the LLM, basically trying to find correlation and how can they basically score and wait? What do we believe holds true? It will also give you garbage if you don't set those rules and scoreings right. Once you've got that engineered, yeah, I basically have a pipeline of Intel that's second to none. - It's a little bit like Palantir with external and internal data, trying to stitch it all together and then find the probable actions. - Honestly, I think anyone can build this. However, I think it's the vision and philosophy that's gonna allow you to stand out, because eventually there should be software vendors that are gonna do exactly what I'm describing right now. 'Cause it makes sense to me. First principal's thinking, of course this is gonna happen eventually. So I think right now there's just six to 12 months, maybe 18, let's see how fast the industry goes, where this is an advantage. Because landing this in a larger organization is just, it's a people problem. You have enough buy-in, do you have the right people to actually do this? Is this not an on the side thing and it gets ignored? And you really funnel investment and power too, which yes, no. So, I think the actual channel is mostly the human hurdle, not the building anymore. - If you wanna know where GoToMarket is going, you just have to look at where product development is today. And that's where GoToMarket's gonna be in 18 months. And so the like, almost ad nauseam conversation in engineering is how do you build this dark factory or software factory? And what is the context engineering, the harness, the loops, all tied together in this system to be able to produce code at these massive output levels. But all of that maps perfectly to GoToMarket. And you're one of the very few companies that I think have endeavored to do this with this level of sophistication. I only know like two other companies who are sort of there, Ramp, who's talked about it publicly with Glass, where they have spent massive amounts of time and money, building the harness and an experience and using Salesforce in a headless manner and replacing their BI platform with homegrown tools. And I do think this is where people all end up. - And interestingly enough, the actual investment that I have been making, we have been making, is very, very low compared to the impact that it's making. I can't share the number here publicly, but it's incredibly digestible if you have the right talent to work with internally and right attitude, aptitude, culture, buying what have you. But let's say funding-wise, I'm really amazed to be honest. It's the best ROI I've seen in this full year. - Where is the ROI predominantly coming from? Is it piped-gen, conversion rates, per rep efficiency, what have been some of the biggest output areas? - So if I talk about ROI, the obvious one is, hey, we've been able to strip out certain infrastructure components, and that's just an immediate P&L gain, because the cost is not there anymore. The other one is time, like literally preparing quarterly business reviews, globally, regionally, per AE, incredibly costly. I think all the sellers in the conversation here, they recognize that most like it is. It's just, it's pulling teeth to get a deck together. That is completely automated with us now. The only thing they need to do, anyone who does a QBR and a sales column, they basically get their talk track right. Like, why are we looking at this data? Why are you calling you three not fully locked in? What's the challenge? And how are you gonna fill the gap? So they're now using their brain to strategize, okay, what's the next step? First, it's spending two to three weeks on marketing to prepare the QBR to begin with. And that was percolating across the full go-to-market function, just the gravity of we need to hold each other accountable. I think that gain on its own is a multimillion gain already in time, productivity and focus and happy sellers, like literally. Same thing towards our investor, right? They want all the detail, all the double clicks, everything. Now it's there. It's real time. Whatever you would like to have, we can actually deliver it to you almost instantly, which is neat to export to the PDF or the CSV. And all of you go. I think their operational excellence and reporting and accountability was a very big push that practically solved for. And then thirdly, I would say the interaction with the market and that's really the man-gen, pipeline generation, account-based marketing. How many one-to-one accounts can you execute if you only have 14 at this moment, account-based marketing managers in the regions? My reasons are just ABM. They just do ABM and ABM all in. Two years ago when I took over the role, they were all up until here with workload. Like I have too many accounts need to handle. There's no way for me to do five accounts at the same time. What have you? Right now, one-to-one account, and I think of them with you to workflow that we have running from account research to mapping a DMU to then doing the strategic narrative on the account, the messaging track, the engagement, strategy, the frequency, the actual content, the outreach plan, whatever, 15 minutes. Right? And it's ridiculously high quality. So 15 minutes of the ABM manager going back and forth with the model to. Yeah, it was typically together with the AE. So they actually go through it together and the system drives the full workflow. And then the output that is there is to all of them like, yes, let's tune maybe a few things in terms of language, because that's always, I think that's a good thing to humanize a bit more, although with the prom card rolls we have the majority of that already in check. But within 15 minutes you have a super strategic plan on a one to one account level. Prior, that was weeks of work. So right now, one to one campaigns to me are almost cheap. Like we can do a one to one. It's not an issue. Also with the outreach. Also with the warm-up of the account. Not an issue. The limitation number comes. How many can you personally as a human digest to also drive a relationship building? To do social selling. To basically have real dialogues in the comments and do outreach over email, have conversations. You can now do this at 50 accounts at the same time. Right? So BDMs and AE's are in my role at least, becoming supercharged networkers. Less busy with, we need to do all this account research. I need to deliver an account plan. It needs to have this format. So I'm just putting it into this format. Like you're putting the most expensive people in your payroll that need to deliver quota in slide X. So if you've seen an increase in pipe gen per person, or pipe gen overall. Yes, we are massively increasing the effectiveness against the market as we speak. I would be happy to come back towards the end of the year to really make the call and say we did it. But the initial uptake I've seen since the beginning of this year while bringing these GTMOS workflows live. It's clearly on the rise. There's also a little seasonality right. Q2 for us as events season. So it's always high. We should be tapering off now summer holidays kicking in. But yes, I see way higher activity, way higher penetration and way more new business conversations. And then I'm not even yet having fully life. The vision that I just shared with you of outside in the man generation and cluster campaigning and everything. So I think I think personally we're on to something at least there's a strong hypothesis and belief we have the infert to do it. It's basically weeks of building refining getting into production and seeing where it flies. Yes or no, my take is that it's very likely to fly with the last like ten minutes here. What do you think is the most interesting use case to share something that you're like eager to talk about. ICP is always a little bit the holy grill who owns the ICP who can make the final call and where we go to market and why and then sales is actually in the frontline marketing is not really in the frontline as much. Does it sit with product marketing does it sit with solutions engineering is it is at the CRO CMO level. We're just systematizing it. So an ICP to me is nothing more than set of parameters against which you can score an account. And that's influenced by your intelligence meaning you've got experience you get certain nuggets you win certain deals you lose there's dynamics around that ICP that change every other. I would say month by now. So making the call to systematize it and build an ICP engine which is when you think about it it's a set of parameters with qualifiers disqualifiers and certain dissatisfiers. So if this holds true the account is out I'll give you an example if they just acquired a competing vendor solution. They're just out of the mix we're not going to go for them at this moment in time until committed right or if they are a mid tier organization and they just got announced to be acquired by another probably larger institution. And we're not yet in dialogue with them ignore because they're going to be busy with the MNA movement. So there's a whole set of pretty direct and clear parameters that score an account which change all the time because we just announced the acquisition of Casisto it's an agent banking company conversational banking with that we can actually go lower into a specific ICP because we have a different solution capability we can bring. So what you need ICP campaigns are being tuned and off we go. And to me it starts at the yeah if you think that as a factory line the ICP to me is very strategic after that you can do account research within that ICP context and you can campaign building within that ICP context and so forth. And I think also there's sometimes you just have to crack the hardest problem for us last year that was the signal engine which is literally driving our engagement matrix like is an account called does it show first signal or is it highly engaged in a sales conversation. That's the horizontal axis the vertical one is no opportunity unqualified opportunity qualified opportunity and customer. And if you visualize that you basically we're playing a game to move non engaged accounts all the way to the right bottom of that matrix so that they're highly engaged next to becoming a customer. It also allows us to see hey there's a qualified opportunity but they're not engaged with us anymore so that that is a fake unqualified opportunity or we need to do effort to actually bring it back in drive the dialogue or maybe take another ankle and penetrate more. So by solving those hardest problems first signal engine ICP engine then you can get to the exciting stuff so the exciting stuff is almost the hardest stuff first. And after that you can get to the exciting stuff and I'll give you a final example. Why didn't I yet build an outreach orchestration agent that just hits all these accounts and hits them under LinkedIn and sends the right message at the right time because I need the foundations to be super super strong. And only then with the right strategic nature and the right intent behind it I am willing to consider to do certain agent outreach or certain nudges but limited because I think it can hurt your brand like incredibly fast. At the moment you turn it on and it's just too loud or too annoying or too multi channel it's you're out of the mix. And we are selling to the C suite C minus one we're in essence in the same room with McKinsey so we're like serious as a brand. So that's where foundation first be super strategic and then based on how we work the account to where the account sits in his life cycle. There will be gentle moments in time where automation takes over and we actually get a hundred X the interaction with the market. That's what I'm working towards. And what are those nudges for back base in if you know selling into the C suite at a bank. You know it's not cold email. Not at all they don't check their email spoiler alert. They're also not per se on LinkedIn although there's this very interesting tactic I love to share with you it's called love letters. Not she has it's it's crazy. I think everyone smells always want to share this one but basically love letters are a publisher raid of the executive team of the account. So you really champion the work they're doing the philosophy that they have the strategic moves they're making as in hey for instance DBS and Singapore it's a fantastic very digitally native organization and you know we want to get into the C suite and have a dialogue with them because we believe we can deliver value. So we basically do a bit of a long read love letter where we spec out our strategy why we believe it's a great strategy which by the way it is. We directly tag the executive team members and why we champion them and why we believe you should follow them when they are on their journey. And if you do this where we're doing is now for give or take six months let's say 40% of the actual executives respond because you give them a massive confirmation of their work you took the let's say. You went through the hassle to write a real piece some of it is researched and driven by AI but it's actually they're a human written to make sure that they come across properly editorial and people feel special they feel hurt they feel seen even the executives. And in the mix of back base that works because you know the higher you get up until a value chain ego and you know prestige and confirmation they they are important especially in the banking context. So that little tactic gets us into executive conversations call it's crazy. What else you wrapping into an ABM campaign to get to that seniority of person. And the podcast so outreach we do with the bank and reinvented podcast which is fun enough fastest growing podcast of 2025 we were able to create a platform and a forum for bankers to talk about transformation. That wasn't really there so I was also lucky in a sense that was so niche that it started working. And we've been upgrading that motion to only have C suite conversations and nothing else. So right now I'm saying a lot of no to people who would love to be on the show but it's C suite only you need to be the operator you need to have the number or you need to have the you know actual heat to transform the bank. I'll give an example I did now to each last week to the chief AI officer of West back in Australia she just got confirmed immediate response. Would love to be on the show I'm going to be very busy for now but let's just connect with my EA and let's get it booked in and I cannot do that scale the two examples that I'm giving you and that's really one of the things I'm very passionate about are super lightweight. They're super cost effective. Versus the classic let's go ABM put a massive elephant of a campaign on to the account and run it for three months and then look back. And we're doing is weekly depending on where an account sits whether accounts are moving or executives are moving minor outreach. And I think also with the brand building we've been doing for one and a half years straight now, we're getting way more top of mind transition so people actually feel like, hey, yeah, cool. I'll briefly respond like, like, why I'm here at Kyle. And then I would imagine you've got like a big research brief for the podcast that's all AI. No, I don't. And of course I do with sweep on Hey, who's as executive in whatever, but I want a real dialogue. Otherwise, I think it's a really uninteresting show. If it's scripted or if it's directed and we're only going to talk about this, don't touch that element. It's same to what you and I are now doing here. We uncovered a few concepts in our prequel and then we actually dive deep on whatever is let's say most resonating on the day. I do it with the bankers too because the moment they get energetic, I know this is where I'm going to go. And then I got a good show and it's going to be super interesting because they're clearly passionate or they're going through something where they want to share something that they're passionate about. And that's not how they initially get into the show because they're uncomfortable, but the moment they get to that space, we're good to go. But we held the research. How do you find what that thing is? Are you doing a pre pre pre pre pre episode? Okay. Pre calls. It's like what's most top of mind to you is it's one question. What is most top of mind to you and why? And that's almost the episode. And then you can always fan out and come back and draw certain conclusions. But I listen to many, many podcasts and I try to find the formula of why am I hooked on certain like the moonshots one? Why am I hooked on that one? Because it's always stretching your mindset or it's always energetic and it's always a different angle. So if it's scripted and it's just banking and hey, we did a mobile app for the past four years, yeah, nobody really gets excited about that anymore. Yeah. I tried to be a lot more prepared and organized when I started the podcast and I would like send a docket ahead of time. But then people would start putting notes in the docket of like the answers they wanted to give and those episodes were the worst. And so now somebody's like, hey, I don't have a, I don't have like an outline for the podcast. Like are you going to send me something and I have to reply and I'm like, no, I'm like, I'm not going to send you anything. We're going to spend 10 minutes at the beginning framing out some ideas and then I just want to go because those tend to be by far the most interesting because I do this job every day. So I don't need to do a bunch of research to know what interesting question is to ask the whole point of the podcast is like, I'm just going to ask the questions that are interesting for me and I'm going to like count on those being the things that are interesting for listeners and so. Yeah. So coming back to your question, what is really breaking through in the ABM context? The example of a simple love letter and the example of a podcast and around that there's a lot of activity and little nudges outreach warming up, connecting them to an event. We have closed door community, closed dinners where only C suites are invited. So we make them feel very special all the time in every touch point and that's lit. That's it because maybe it's very simplistic thinking, but I have 3000 banks. They typically have a C suite of five to eight individuals. So it's 3000 times five to eight. I need to know those people. So my main motions are for those people and the rest is just to support it. And then you can get creative and being a C suite myself is helpful because I know what's coming into my inbox is typically not relevant to me or it's not even standing out or it's it need automated. So I think it's a wonderful time to have this tech, have this type of setup and the remit to actually just be creative and create breakthrough. Very cool. All right. I want to move us to the quick fire to get you out on time. What do you think separates a good CMO from a truly great one? Deep business and subject matter expertise understanding of the company that you are representing that really is the thing that makes for breaks in my world at least the CMO seat. What advice do you most commonly give to like first time CMOs or first time marketing leaders? You need to be well rounded. So if you come in just looking at your little swim lane or your craft that you are leading, you're already making the number one mistake. Market tiers are coming in as a specialist not as a generalist. If you ever want to let's say climb the ranks and become an executive that has a P&L level conversation or strategic conversation with your CEO or your CEO, you better get out of that mindset very fast. You're a business leader. Number one. Number two, you're a craft leader. And those two need to go together and if they don't then indeed the CMO will not have a seat at the boardroom and it will be just AVP marketing that is tactically executed whatever they cook up at the other table. Yeah. What's the hardest lesson you've had to learn in your career? Knowing what I didn't know and going through the pain of not knowing, trying to figure it out and hitting the wall multiple times until I got it. I didn't have to happen every other year Kyle. What's the best thing you've read in the last year or two? Sounds silly, but there was this book from Tiago Forte, I believe, about how to structure your second brain. I'm one of those idiots that is really into okay, how do I structure my thoughts so I can work with them appropriately and store accordingly? This is pre-AI by the way. But Tiago Forte's book was about structuring an obsidian vault, which is a graph database in structuring notes. That book really triggered simplistic thinking and organizing of everything that's going around in my life and in my role. Have you been able to make that work? I've tried this twice. Yeah, it works. But you have to be forgiving to yourself. So what I actively manage is projects and areas that works and that's it. So the other part would just like a big dumpster fire part where all my old stuff is and with you know, the false search, I can find them. But projects and areas works very well. Tim, this was awesome, man. I'm really excited we could get together on it and get into some like real weeds. I'm so impressed with what you've built. Like, I talk to people about this stuff non-stop. I'm constantly talking to people either on the podcast or outside of it to learn about what the most cutting edge folks are doing and I haven't seen a lot of people that have built this level of sophistication. So big congrats. And thanks so much for sharing all of this because I learned a ton. I have a ton of notes that I'm going to be taking back to my team and I'm sure folks listening are to learn a lot, learn a lot themselves. Thanks, Kyle. Happy to share. I hope it was helpful for anyone listening in and unfortunately it's audio. So I couldn't demo it or show it. That really speaks to the, to the heart of mind usually. But then again, we're also just on a journey. So we're still learning and I think that's also the main takeaway. Yeah, maybe we'll do a webinar or something and you can download some stuff later. I'll see you lots of people. Cool. All right, thanks, Kyle. Thank you so much. Thank you for listening to the Revenue Leadership Podcast. If you enjoyed it, don't forget to subscribe and you can find a link in the show notes. And be sure to leave a five-star review, share it with your network and please join me next Wednesday for another great conversation.