AI is killing SaaS

Drew: I had Claude code rebuild a CRM for us, and I had it hit the HubSpot API and pull all data over, and I just sunset. I turned off our HubSpot subscription like yesterday.

Brett: Building great experiences. Revolutionize your business. Building great experiences. Innovative strategies. Building great experiences. Elevating for your customers.

Drew: Like.

Brett: Comment and subscribe.

Drew: Hey, folks, this is Drew with another episode of the Building Great Experiences podcast. Super excited for this conversation. Today. I've got my friends Brett and Rio on the. On the show. Um, I was recently on their podcast and we had a really great conversation about the intersection of customer experience, marketing, uh, go to market and AI. Um, which, as everyone knows who listens to the show, uh, there's a lot of really interesting things happening right now, and so we're going to unpack that together. Um, and should be a really fun conversation. But Brett, Rio, welcome to the show.

Brett: Great to be here.

Rio: Great to see you.

Brett: Glad we're keeping the conversation going because it was certainly a fun episode that we haven't published yet, but soon to publish on Signal and Noise, so it's awesome.

Rio: Yeah, that was a good one. And it was like, it was great to see you. I mean, for those of you who don't know, Drew and I, we go back, what, five, six years at least, right? We work together at consulting firm for a bunch of years. Did some really fun work together. So it was a really good to reconnect. Good conversation. I was really excited to continue it here.

Drew: Yeah, same. No, for sure. And, um, yeah, for folks listening, check out Signal and Noise. The first part of this conversation is over there when it's published, and you'll be able to kind of connect some dots. But I'm a big fan of storytelling. I love people's stories. It's just like my favorite thing. So I'd love to hear Yalls origin story. How do you got into this? Maybe a little bit more about you for folks who are listening, don't know you. Um, why don't we start with Rio and kind of rock from there.

Rio: Okay, cool. Yeah. Thanks, Drew. So it's funny. So before I jump into like, my origin story, like, Brett and I actually, we've got a long history. We go back, what Brett are 30. 30 years at this point.

Brett: Right? 30 years. Yeah, yeah. No, 30. Yeah. 1996. And we're in 20.

Drew: 26.

Rio: Right. Right after college. It's funny. So I. I went to college in Boston, and right afterwards, uh, a bunch of us, a bunch of dudes, we ended up renting this house. I think Five guys rented each bedroom. And then over the next like three, four months, Drew like all these additional people showed up. Can I live here? And it kind of.

Brett: Yeah. And it was on. It was on Electric Avenue. So the song went along with this place. Electric Avenue Somerville, Massachusetts Scummerville, Massachusetts. A little bit of a rough uh, and tumble guy sleeping on the couch sort of scenario. You know.

Rio: Actually one dude slept on the couch for six months. Like that was. Became his bedroom. But I, but I digress. It was super fun. So Brett and I ran into each other, a few mutual friends there. We kept in touch and we both ended up back in New York around 22,000 era, you know, like around then and realized we knew each other and became really good friends and worked together in a startup in an apartment. So lots, lots of history there. We can get into that. Maybe, maybe that's story for another time. But uh, my personal story. So um, actually like very few people know this, but it actually started after grad school. Drew. I was actually a journal for a couple years. Not a very successful one. I guess it's a. It's a tough business to make any money in. So I was doing that for a couple years when I lived this. When I lived in France. Was enjoying it, having fun. Got published in some decent, decent public decent publications, newspapers. But then just wasn't. It was. It was tough to get by. So ended up looking for a job and found a job. Working in digital media is where Brett and I ran into each other doing media buying. And this is like the dawn of digital media buying. Right. So got into um, like worked kind of as a media planner buyer for a bunch of years. After doing that ended up getting. Finding my way into consulting. Worked uh, at a big four consulting firm. And then we met. We were working at. We were working at the same firm called Slalom. I was there for about 11 years and then right now actually work. It's kind of. My career has gone full circle. Right. So actually work in a consulting branch called Credera where like a cons like we call it, it's a transformational consultancy like SI Consulting firm within Omnicom, um, the big global agency.

Drew: Oh yeah.

Rio: So. So. And I was.

Brett: They.

Rio: They said they wanted me to come here because of my media background. So it all kind of all kind of came full circle. I'm the leader for advertising and ad tech. That's a little about me.

Drew: Cool. Love it. Well, side note, just the, the. The couch surfing college days. I can relate more my college post college life. I was like, slinging T shirts on the streets of Charlotte, trying to. Trying to build a business and. Yeah. Anyway, it's kind of funny how people start in, like, weird places, and then we all end up in this kind of, like, later on. Like, it's almost like once you've kind of gotten past all the hard stuff, you're like, yeah, I kind of went through some stuff back then, you know, Like, I don't know. I don't know if you guys had that experience. Like, it's. I didn't start opening up about my origin story until maybe, like, five years ago. I'm m. Like, yeah, actually, like, I used to opening up about all the

Brett: failures and mishaps, uh, and trials and tribulation, but really, that's. That's what it's.

Rio: Almost no one's career goes straight. Like, everyone has these crazy careers. And that's. I think at a certain point in your career, you become comfortable with who you are. Okay, I've done all right. Yeah. I don't have to be. There's nothing to be ashamed of. Right. Every, like, very few people had the perfect path.

Brett: Right.

Rio: So I think. And it kind of makes you who you are. So.

Drew: Yeah, that's right, for sure. Um, well, cool. Brett, I'd love to hear your, like, what happened after the couch surfing. Right?

Brett: Yeah, yeah, yeah. So, Bret House. Uh, I'm the founder and CEO of HiSignals. I'll tell you a little bit about that in a minute. Uh, but I've been in ad tech in Martech, in B2B SaaS, and now AI for a long time. And Rio kind of told the origin story there. Uh, interestingly enough, he was a journalism major. I was certainly a liberal arts fan. In college, I was a comparative literature major. Uh, he's got a podcast episode with Joe Zappa, who, uh, it was a Comparative Literature PhD. So very rare to find folks that do that. But I went into college thinking that I was going to write the next great American novel and dressing like Jack Kerouac and listening to jazz music and reciting poetry, that kind of crap. And then I realized I graduated, I moved to New York City, lived in Tokyo for a little bit, moved to New York City and realized, wow, I need some hard skills. And so what really made a lot of that abundantly clears when I joined the startup with Rio. And so we. We joined another founder, uh, who had. Who had sort of started early, had launched a company called Pop Up Technologies, and had gone through sort of the boom and the bust of the Pop

Rio: Up Ads like pop up any of those.

Brett: Yeah, early days you know. And uh, and so I learned a lot from him because he had, he had gotten more seasoned just in the sort of hard and fast business life where I was sort of just getting my feet wet. And so I took that experience, sort of first experience in sort of building an ad network, building a media buying platform, performance marketing. Early days of the Internet. Admit we were still getting the Internet delivered through disc from AOL in the mail. I say that all the time but it's remarkable. I remember my first email address. It was a Juno email address. Not to totally age myself but Juno hasn't existed for like 30 years. So that was a long time ago. Um, and I took that experience and said hey, I'm going to go back and get some of those hard skills. Um, you know I did have dreams of eventually founding and running a business. Um, and I went back for an MBA at BU and then, and then went right from there into ad tech and so uh, really started my ad tech career at a company called Exalate that was a data management platform that was bought by Nielsen. So I kind of was part of that exit and trajectory and launched the Nielsen marketing cloud. And then I kind of did the same thing with a little bit of a longer pre acquisition tenure at a company called New Star that's in the identity space and they were acquired by TransUnion. So that was a real seeing how you could bring a product, uh, that was multiple products with a very broken sort of go to market playbook or lack thereof and actually bring this stuff together. Uh, um, you know, differentiate yourself in market and actually get acquired for 4, 5, 6x was quite a journey for me. And then finally got to the point where I said hey, I've, I've built enough of the strategic foundations and playbooks and diagnostics and have run multiple types of commercial and some product teams where I think I could run and build the business. So I launched uh, High Signals and we're, we're calling ourselves a uh, launch readiness acceleration company that helps companies tackle some of the hardest problems more, more quickly around how do they bring a product to market and commercialize it effectively without just nuance and guesswork and qualitative like I think this is right doing it with, with a scientific rigor and AI to back up some of the decisioning process so everything can speed up and you get to market with something that's more likely to succeed. So that's where I am now.

Drew: Very cool. Yeah, well, really excited to crack into this conversation. I think our last dialogue in December was, was fantastic. And what's kind of interesting is the way of AI. A lot's changed in the last, you know, six or seven weeks since we talked. Just, you know, a couple little things that I'll throw out there for folks listening. You've probably heard of what was once called claudebot. Openclaw is now, you know, out in the, in the ecosystem. Recently, um, acquired this last weekend by OpenAI. And that has been really interesting, especially when you think about go to market or really anyone who's kind of in this sort of this landscape. Because it's like the first, I would say, real example of the promise of AI come to life. Meaning you can give AI access to tools, a computer, your accounts, and it starts to just do things for you without you having to hold its hand or, or kind of like chat and then copy paste over here and then chat and copy paste over here, right? It's gonna first real live solution that, I mean just about anybody can set it up and run with it, right? And so what it gets me excited about is when we talk about the frame of customer experience go to market ad tech martech. Like there's a lot more that people can do as an individual, right? And you kind of take that and you say like as an organization there's a lot more that's possible. So anyways, with that kind of frame, I'd love to hear what you guys are seeing in this space. Like you recently went to CES and obviously you all are talking with a lot of executives. What are some of the trends that you're seeing right now when it comes to go to market customer experience ad tech martech.

Rio: I guess a couple of things jump out like right away, Drew. So I think that your point of AI, like you look at like virtual assistant, it completely changed that, right? That you know, the open call example you just gave, right. Like people been trying to build those for years, but it's finally like, wow, it finally got to the point where the technology's arrived. I'm seeing that in different places, right? I'll give you a couple examples. So one would be I'm talking to a big cable company now, right? Like they've tried to launch self service tools to small businesses to, to buy inventory, you know, by advertising on through, you know, for either broadcast or cable, right. For years they've been trying that. It's been, it's been very difficult. They've rolled out these self service tools but it's so complicated between the day parts, the inventory types and then not to mention even creating the content itself. That adoption's been, I mean you know one of my clients he told me is around 5% is crazy.

Brett: Right?

Rio: Which, but I think these recent advances in AI, specifically agentic AI is just going to create an environment where we could take a very complex process, put it, put this like agentic layer over it to make it incredibly easy to make the complex hard. And then on the content side could we use this to maybe make like production ready video? Right. That we can, we can, that these businesses can upload. So it's finally gotten to a point where it's good enough to do that. So these advances in AI are suddenly making it possible to do things that you couldn't have done it a million years before. Um, now it's early on we haven't actually that you know, that's theory versus actually getting things into production. But I think it's creating this opportunity for that. Another example would be looking at content supply chain. I mean that's been a thing for a while, right. It's not new but a lot of my work now is in paid media. Helping brands you know better personalize better target content, get the right tools. It was always kind of sat like paid media always kind of like when you look at the content always kind of sat by itself in a silo using the gentic workflows, using generative AI. Can we now have like hyper personalized content that's targeted to the individual, at least to very small cohorts create a kind of on the edge, on the fly like that's, that's becoming possible now and that never really was before. So it's opening up capabilities that we really. Of things we've wanted to do for a long time but we couldn't previously. We can now contemplate doing those things which is kind of amazing.

Drew: Yeah, go ahead.

Brett: Yeah, yeah. So, and I think some of the stuff that came out of ces, I mean I think two things was, that really came out from some of the conversation I have was one, this notion of GTM compression, whether you're a B2C company, that's, that's, that's, that's forced to, to. You have product life cycles, you have, you have uh, marketing uh, and advertising life cycles to get the product into consumers hands. That whole process end to end is being compressed, compressed. And it's being compressed not uh, just because you know your PE investors or your board is forcing you to compress it. It's being compressed because now we can, to Rio's point, accelerate the decisioning apparatuses that help you make better decisions on, on what creative you can use, for example, or, or what micro segment or hyper local audience you're going to target. And a lot of that's because of tools. And one of them that I'm implementing at high signals that allow for um, like buyer simulation and competitive uh, scenario, uh, planning at a speed and a scale and even simulated focus groups at a speed and a scale.

Rio: That's very cool by the way.

Brett: Yeah. That are exponentially beyond uh, and I've talked to some founders about this, exponentially beyond what focus groups on the B2C side ever been able to do where you can simulate buyers down to a very nuanced level and you can do

Rio: dozens of different ones. Right. Instead of like taking months to plan one and getting questionable outputs. Right. It's really cool.

Brett: Yeah. And it's so it's like, let's say you've got 10 target markets and five buyer Personas. You could, you can suddenly, you suddenly start to get a multiplication effect of exactly how you have to nuance your creative messaging, your advertising, if it's B2B, your sales motions, your customer success motions, things like that, how do you do that at speed? And if you've got a focus group that can literally test that at scale, thousands and thousands of data points that are coming back and then interrogate the outputs for quality, for clarity, for accuracy tied to ICP and all that sort of stuff. To me that stuff has all been promised and it's been in brand managers hands, it's been in product marketing people's hands on the B2B side. And it's remarkable how slow us as humans move. And we can't do that kind of stuff as rapidly as sort of the market needs to. And so I think you're seeing stuff that these cycles are going from, you know, it used to take half a year, four months, three months, a full quarter down to three, four, five days or a couple of weeks at most. I mean I think the hardest part, once you get some of the decisioning outputs, is sort of winning the trust and hearts and minds of people that you work with and then aligning the execution piece of it to say, hey, we're pretty comfortable with this decision, it's been interrogated and simulated and scenario plan, but how do we actually get buy in from the team? So it still seems like there's a breakdown, I think, with how humans are going to be implementing these things. And I'm seeing it, I've seen it through my whole career. But even recently, B2B teams that still struggle with the ability to make decisions at a really high level of fidelity, like they're not really making data driven decision decisions. And I don't care if you're the CEO, the Chief Product Officer, the Chief Marketing Officer, they're oftentimes like uh, I've got 25 years experience, finger in the wind. I've run a spreadsheet and a model financial model and I know the market and so but there's still, it's still intuition driven. It's not actually being tested at a super nuanced level. And I just think that that speed is, is now available in people's hands if they, if they employ it correctly.

Drew: Yeah, I completely agree. I think what's interesting about the testing that you all kind of pointed at is not only can you simulate these things but like, uh, I don't know if you all remember this, like a couple years ago there was this whole debate over like is synthetic data worth it or not? Does it drive, you know, good results or not? I think at this point it's, there's a lot of data to demonstrate like actually the synthetic data is not that bad and actually it's getting better every day and you can start to simulate these things with high confidence, high probability, almost more so than like trying to collect a data set that you're then like analyzing and, and sort of extracting insights from. So to your point, I think you're spot on like AI and kind of go to AI times go to market is, is getting much easier to get data that you can action off of and you kind of know with high confidence JI will tell you hey, high probability this is actually going to work because I've run it against these like scenarios with your kind of like agents that are your ICP and it performed right. And so now you're not having to like make a finger in the wind situation. You're, you're like yeah, well think about,

Brett: think about what Nielsen did that was that Nielsen post acquisition, what they did with sort of television ratings. They take a small subsegment of the tv, the Nielsen tv. People that push buttons to sound personate and extrapolate and then they have a watermark of the television and then they extrapolate that same seed audience which was pretty small. They kept expanding their pool. Uh, but still 100,000 people extrapolated out across the entire population of the United States is a huge extrapolation. AI is able to do that, uh, if you give it a little bit of seed data at a level of accuracy and a level of complexity in terms of its predictive capacity, that is just unheard of ten years ago.

Rio: Well, I think a lot of our methodologies are outdated like that. I heard the other day that the way we even calculate inflation actually send out surveys to just a bunch of random people. I think it's a small, like 10,000. It's like just very small number when you think of all the signals they have, right. That they could in theory use in order to really understand what inflation looks like. And I think, Drew, you brought up a good point, right? Okay, maybe synthetic, synthetic data is not quite, it's not quite as accurate, maybe, and it's getting better all the time. But like fact, you're getting it quicker, right? And then you could, you could then take that and then validate maybe using some, some, some actual real data on top of that. I think it's speeding everything up. And I think this also gets to another really important point generally too. I've been advising a lot of startups, AI startups, especially the past couple years, is it's kind of table stakes that you can, you need a good product, right? But if you don't have the good go. If you don't have the go to market nailed, good, best product doesn't always win. I've been advising one company. Their product's amazing, right? It is better than anything else we could find out there. But like, no one knows about it. Their go to market plan was a little off. Their ICP was. I think they had a wrong icp. Brett is one thing I told them, right? It's like you're focusing on wrong use cases, wrong person. You got to pivot this quick, which is kind of what you do and you learn when you're running a business. But I think those cycles are happening faster now, right?

Brett: Yeah, yeah, those decisioning cycles. And you don't have to make as many mistakes ahead of time. So like, is it a perfect example? One of our good friends, Judah Phillips, who've known, he runs Boston AI Week in Boston, which is not bigger than advertising week. He launched a startup called Squark AI. Don't ask me why that name. It sounds like a chicken joke. Squawk. Uh, the, uh, Ed talk about product m market fit. Like I saw him, you know, about a decade ago going through all of those struggles, and he's come to the conclusion that it's not the, uh, best product that wins. It's the Best marketing that wins. And it also has to do with founder market fit. Is the founder of that startup, him or her. Do they know the space and the value that they're delivering and the actual use cases that clients in the real world will adopt or be engaged by? And that's often the case is they're feeling around in space, sort of throwing shit at the wall and hoping something sick sticks, trying to figure out what the best use cases and some of these core capabilities could solve. A whole bunch of use cases. But, but his, his, his um, recommendations to me have always been simplify, simplify, simplify, simplify. Like if you've got a value prop, like I make QBRs better, like that's a great value prop. Just like, just focus on that. You make QBRS better.

Rio: Well, also be careful. Like the AI, uh, company I was like advising, they were in the healthcare space, right? It wasn't like in their initial way they're going to market Brett. They were actually competing with Epic. It's like, you do not want to compete with Epic. I mean like they're going to eat you alive. They have like 99% penetration or 85 and crazy. Like they work with everyone. And anything you say you do that they don't do, they're going to tell the client they will do. So like it's going to be in their roadmap. You'll never displace them. You should be complimentary to them, right? And you know, find something, you know, go take a different angle. So it's like founder market fit. What does your ICP look like? And how are you, like, how are you going to market? I think that's as important as the tech itself. Maybe one.

Brett: Yeah, well, and the advanced AI, uh, gives you the ability to simulate all this stuff. Not simulate it, but actually, you know, like I've done a ton of competitive SWOT analysis that honestly, in the old days, five years ago, you'd give to a product marketing team, you'd be like, do a SWOT analysis. And you're like, no, another academic exercise. I don't want to do that. And, and you do a SWOT analysis and you give the right inputs, right? And I've done this for high signals, I've done this throughout my career for other companies, more manually. And it's incredible the type of insight that it'll give you by scouring, scouring the, uh, the, you know, with deep thinking exercises or deep learning exercises. So it does give you the ability to kind of quickly, quickly pivot. Like I, I pivoted I pivoted a couple of times since the launch, and it's been December. These are changes that are happen in three to four days versus three to four months. So that's where things are really exciting.

Drew: Yeah, I, um, I completely agree with you guys. I think that at this point, what is going to win and what will continue to win is really strong ideas with really strong execution that get you distribution and strong brand and experience that helps differentiate in the market. Because it's like, all the tools and the technology makes this thing go faster and faster and faster. So, like, what will stand out? It's like, just because you build a day won't come. Right. Field of dreams is a lie. I say that a lot. Like, you have to figure out how to, like, get your product or your service in your customer's hands and make sure it's something that solves a problem that they have. Right. I tell a lot of founders this. Like, you want to be in a painkiller business, you don't want to be in a vitamin business. Like hell. To be in a vitamin business.

Brett: Yeah.

Rio: Uh, yeah.

Brett: And we'd be remiss to say that I can also be dangerous in this respect, because if you start with a bad foundation and you don't have that sort of decisioning, sort of discipline and logic and diagnostics behind what you're doing to make a decision around launching a product or launching a company, AI will, strategically, it'll just blow up the noise. It'll amplify the bullshit or the crap. Right. Which can send you down a, uh, rabbit hole.

Rio: It also tells you what you want to hear. Right. If you want to hear. Like, if you. If you say, tell me why, tell me this is a good market, you'll keep. And if you keep giving it positive reinforcement, it'll continue to, like, just layer on the bullshit. Right?

Brett: Yeah. Especially like the free LLMs and the sort of standard stuff that's available versus the more proprietary platforms.

Drew: Yeah, for sure.

Rio: So knowing how to prompt them and interrogate them and, like, you can go deeper and deeper. Deeper is very important, Drew.

Drew: So I want to share something with you guys. Um, share my screen. Uh, this is a. A tweet that came out from a guy named Brian Flynn two, uh, days ago. And I'm bringing it up because it's a very interesting kind of, like, thought, which is this idea of, like, go to market for agents, um, you know, for folks listening.

Brett: Can you read it? I mean, I can't even read it on my screen, but just at least get summarize it, Cliff.

Drew: Yeah, yeah, so, so that sort of TLDR here, um, and I'll, I'll drop this in the show notes for folks listening is basically this idea that as we all. I mentioned open claw at the beginning of the call, as we all start to have our own personal assistants who are kind of like driving everything, what will end up happening is those, those agents are going to make buying decisions on our behalf, Right. They're going to go out into the market, they're going to start making decisions for us, and it's going to be much more of a calculated decision based on like, probability. And also one of the things that this guy says in the, in the conversation is like, when it goes out into the world, it's going to be making decisions based on ROI for token usage. Right. So, um, just to give an example for folks who are listening who may not be kind of familiar with the token burn concept, every time you ask GPT or Claude or Gemini to go do research on something for you, it is making a decision to say, all right, I can go do this academic research on this topic, but if there's already a data source of structured data that I can hit and like, get it faster, I'm going to do that because it's a more efficient way to use tokens for my, you know, for the person who's kind of, I'm interacting with.

Brett: Yeah.

Drew: And then you take that a step further and you say, okay, like selling products and services. If there's like now this ability for an agent to hit an API to get like pricing information, value proposition, et cetera, et cetera, and help the, you know, whoever the user is on the other side kind of get to the, yep, I can buy this thing for you. Here's the best, you know, reasons for why, whatever. And they're just like, sure, go like I think it introduces. And the reason I bring this up introduces a very interesting go to market quandary. Right? Because up until this point we've been, um, optimizing for human behavior, human psychology, human responses. And like, now it's like, okay, what does it go to market look like when we're actually trying to optimize for like agents who are making buying decisions on behalf of a person.

Brett: Well, hold on, Let, let me, let me play devil's advocate because that's agent, that's agentic commerce. And uh, you're one. It's, it's like the Socratic method here that there's one big assumption that you're, that you're sort of assuming that we have to agree to, to say that this is an argument worth making is that agentic commerce is going to do what you said it's going to do. And, and I actually categorically think it's not. I don't think that agentic you're going to have an agent that goes out to, uh. Because I think it's already operating within the platforms themselves. You go to Amazon, you go to Walmart.com and there is a customized experience. It's called subscribe and Save. Right. And subscribe and Save could be identified in a thousand different ways. But the notion where you take the human out of. Think about the Internet of things and what a huge failure that was. Right? Oh, I'm just gonna, I'm gonna sell it. Say, tell my Alexa or my refrigerator,

Rio: 5G or uh, blockchain.

Brett: Yeah, yeah, my refrigerator is going to tell me I'm out of milk and it's going to order milk. And you take out the sort of human decision control loop of that process. I don't think that's going to happen exactly that way. I think it's going to happen within platforms to speed up and to recommend. But if it's doing it without human agency, um, I think it's a solution looking for a problem.

Rio: Okay, I disagree. Okay, here's why I disagree. I think that it'll slowly start. Will get more comfortable initially. Like you go to Amazon, you can only already use Rufus, right? And Amazon already gives you recommendations on uh, it, which is machine learning. Right. So imagine you have an agent and then initially you're going to say the agent. I'm going to travel to the city. Just find me some find. You know, I like to fly United. You have my frequent flight information. Find me some flights and come back. Maybe it's not going to buy for you right away, Brett, but I think it over time as you get more comfortable, first it'll interact with you. It'll be more like a dialogue back and forth. And eventually we're going to become more comfortable. Like 30 years ago, no one would. Buying something online was uncomfortable. Giving your credit card was uncomfortable. Apparently people used to fax in credit cards or email credit card numbers right to Amazon. When they first started, it was so crazy. Right. So I think it'll happen over time. But. But Drew, you bring up an interesting point I think beyond, you know, agenda commerce. And Brett, you know, maybe just how long does it take is the question. But like the assumption we all made when, you know, when a. Okay, agents Started were spun up as a concept, right. ChatGPT and all these other, these other platforms made it possible to think, okay, we're going to have agents replacing people at some point, right? First you started with Task now and maybe switch to jobs, right. The assumption was this will all be, it'll be much cheaper. I talked to a friend the other day. He's using openclaw. He had agents that were like going through emails, filtering inbound messages, crafting messages back, reviewing pitches. He said it was burning through almost $300 a day in tokens. So that's. If that's a lot, that's an expensive. It goes like a, uh, cost savings to, you know, maybe it's more expensive than a lot more expensive than someone you would hire to do that. So granted token prices will fall down. There's a massive build out of data centers now, these models getting more efficient. Maybe chips will get more efficient too. So I don't, I don't think the prices will be static. But it does, uh. But it's not free, right? It will cost you something.

Drew: Yeah. Yeah, I think so. I love these perspective. This is exactly why I brought this up because I wanted to have this conversation. I tend to, this is my opinion. I think that there is, there are going to be types of products and services that people are still going to want to do their own. Kind of like talk to us, talk to a real human and kind of like there's probably never going to be a day where you just trust an agent to buy you a car or buy you a house. Right? Like, I don't know that that ever will happen. But I do think that like when you think about software, right, uh, software is going to. I mean there's a whole discussion right now like software is getting eaten by AI and at a certain point, I don't disagree. I actually just. This is like a funny little anecdote. Um, I, two days ago, um, had cloud code rebuild a CRM for us and I had it hit the HubSpot API and pull all data over and I just sunset. I turned off our HubSpot subscription like yesterday. Um. Oh, wow. And like I didn't. I just had to do it. And what was interesting is, uh, what Claude made recommendations to me about was like, hey, um, we're gonna need the ability to do this. So like, let's get Supabase. So well now I have, you know, so Supabase is like a 10amonth subscription, but it's like cheaper than paying for like 12 or 15 people on HubSpot. Right? And it's like, hey, you know, if you want to send emails, we got to have this service, um, called Resend, which is like, allows you to send through SMTP for like a, you know, tennies or whatever per thousand. And, um, I'm like, okay, so like, it made the recommendation that I was like, I don't care. Sure, I'll sign up for these services. I don't really just get to get to the outcome. And so I think that depending on what people are buying and what they're doing, like software in particular, it's going to be harder and harder for people to kind of create a moat because agents are going to start to be one, building the tools for people and just like accessing APIs or MCPS to kind of like do the thing. And then they're just gonna. There's gonna be a currency exchange on the back end that allows me to buy it from you and I don't have to actually use your thing. And then so it introduces again, like this very interesting question of how do you sell to an agent who's making that? Like, you know, the agent could have told me there's a plethora of other, you know, Supabase, like tools out there. It told me, hey, we're going to set up a Supace. I'm like, sounds good. Like rock and roll, I don't care. Um, and that's the kind of thing. And again, I'm a case study of one. And obviously this is anecdotal, but it starts to beg the question of like, well, what should go to market teams and executives be thinking about, especially if they're in the software world, where their services, their software service is going to become a commodity. How do you get distribution in that world? Right, yeah.

Rio: Uh, that's interesting, Drew. I mean, I think that, look, I've seen people vibe code a CRM. Like, actually it did basic things like case management, contact management.

Brett: Right.

Rio: Um, you know, all the things you, you would need at a very basic level. It's certainly a small company for a CRM. It's not going to pass muster and enterprise. Right. So I think short term, our salesforce, these other guys are going to be okay.

Brett: Sure.

Rio: Because, like, you know, their data's locked down, they've got privacy, they've got security, they've got legal reasons. They want it. They want to have things managed a certain way. So. So short term, I think they're fine. But that does beg the question in the future, right, if, uh, if agents are doing A lot of the, the bd BDR SDR activities, maybe agents don't need a salesforce. Maybe they just need basic CRM. And if that's the case, maybe something that, you know, agents, agents can vibe code very quickly, that is basic CRM functionality is going to be okay. So I don't think short term these big SaaS companies get clobbered by this, but the big ones that are, that are big exploded. They've got massive, massive teams, uh, of account execs. They've got massive customer success teams. I think it, I think it challenges that the big SaaS model, right, where you know, it starts to eat into their margins, starts to beat down and maybe they can't even sell their light, their software the same way. Maybe it disrupts that. So I think, I think SaaS as SaaS as a thing is fine. But I think that the big SaaS models do get challenged as some of their, you know, some of their growth gets curtailed by this. I don't know Brett, if you have a thought on it.

Brett: Well, yeah, I think especially when you go down to mid market, Drew, I mean, you know, the enterprise I get, you know, they're big, they're complex, they're highly governed, they're highly siloed. Sometimes you do need ERPs and CRMs that can, can move across an organization and handle sort of the weight of that organization and the complexity of the organization. But mid market, like I'm not sure that you need the weight of Salesforce. And I've seen this problem again and again and again at sub $200 million sort of mid market, high growth organizations that have rev ops teams that are just way too highly staffed, right. Spending way too much time on just the most blocking and tackling type tasks of just, you know, integrating SaaS. I mean how many times have you seen Salesforce integrations? I've seen it multiple times in my career. And you'll have 10 people working on this for like six months without any like measurable output because so much time has to be done with mapping fields and data deduplication. I saw it at my last organization and it slows down your ability to drive measurement and analytics, right? Whether it's customer lifecycle analytics or churn analytics or acquisition analytics. And then you end up having teams which kind of becomes the biggest broken part of go to market reporting their own stuff from their own siloed viewpoint through their own platform. So you have the marketing team reporting from Marketo or HubSpot, the sales team and the chief revenue Officer reporting stuff out of, out of Salesforce. The customer success team is using another, another platform and then maybe the product team is using Pindo for product analytics. And none of this stuff is agreeing. And there's one customer behind all of these different data points and data platforms. Right? So that to me is where it really breaks down. And if you can have a simple system that's easier to orchestrate, it's easier to connect and quicker to market for mid market, you're in a much, much better position, I think.

Rio: And true. It's interesting too, like what Brett just described, that disjointed kind of not only internal because customers, uh, feel that right. If internally you're not using the same reporting systems, you're not using the same systems, you're not on the same page. And that ultimately does manifest itself in customer experience. You brought up CX before. So I think like when you look at those handoffs, you're going from marketing, right. To sales, when we, to pre sales to sales and then to success, customer success. Like there's a handoff. And so I think thinking of things holistically and like uh, starting with the customer working backwards. I mean, you mentioned, we mentioned Amazon earlier, they're great at doing that, right? Like let's think about them and let's design things like the way they, you know, the way we want them to experience it. Right. And I think that will change some of those decisions you make. But, and I, uh, don't know, maybe AI does help smooth a lot of this over, right? I mean even if things are that muddled under the covers, if AI can create that one agentic interface that does work directly with the customer and guides them through things and obscures the incredibly complex. Maybe that's a big accelerator.

Drew: Yeah, I think that um, the really interesting thing is I think humans are the bottleneck right now. I think we have a society like kind of a grown accustomed to our apps and our software. But what actually I think is going to start to happen is like instead of having a CRM, you're just going to have different sets of data that your agents are connected to and you can just on demand ask whatever question you have and it'll tell you exactly what you need to know. So instead of you having dashboards and reports and pipelines and whatever, it's going to, you basically just say like, hey, what are the top three deals that are going on right now? What are they? Where are they blocked? What do we need to do? Like hey, what? Customers are kind of stuck, right? Now like, you know, and it starts to make these things on the fly for you without you having to kind of go and like log into a software system and like look at a screen.

Brett: Yeah.

Drew: And as long as we can access

Brett: the data though, there needs to be some centralized system record, like I'm talking to a company right now uh, that we're thinking about partnering with that is a bi, like a business intelligence layer that plugs into all of those fragmented systems, the CRM systems, the marketing automation systems, the customer analytics or product analytics systems. And it takes all that unstructured data that may, you know, and pulls it all together and synthesizes it and gives you immediate insights around, churn around, last conversations around qb. If you're putting together a qbr, it brings all this stuff to the surface. I mean, oftentimes to your point, I mean it is ridiculous. Even relatively advanced SaaS companies will take weeks and weeks and multiple people talking. Nobody knows where the bodies are buried. Nobody, nobody knows what the systems of record is. There's three different instances of Salesforce and they can't, they can't, they can't figure out which clients buy which products at the very, at the most basic level.

Rio: In fact, I, I my prediction, the rise of AI native software packages and like the creation of this agentic layer will put a stake in the heart of the concept of a power user. You shouldn't need to be a power user ever.

Brett: Yeah, I agree. Uh, because, because of the conversational interface.

Rio: Why should your, why should your software have 50 knobs and controls and things you got to mash? It makes. It's, it's only because we were forced to do this because there's no other way and people were trained to use these crazy tools.

Drew: Yep. Yeah, I think, um, the, the completely agree with you. The software becomes like almost less important as the outcomes and the capabilities and it's like effectively whatever you can imagine or whatever your agent is, is imagining for you. Like just to give another example, um, I have like a CLO agent which has a team of kind of operators underneath it that I can kind of delegate to. And I, because I, you know, every day I pull up, I've got terminals and I'm kind of doing stuff. I said, hey, co, look at everything, the status logs and transcripts for the last like month and like make some recommendations on stuff that we should build. And it came back and I was like, hey, uh, you're building this like delivery management tool for your team. I've got access to Slack and email and Your transcripts. I'm going to make you a, ah, client health dashboard. And I'm going to make it based on like sentiment in the transcript data, um, you know, sort of signals in the Slack channels and, and like email indicators. And I was like, oh, uh, that sounds great. Yeah, do that. And so 15 minutes later I popped up and I had Dashboard showed me all of our accounts. Red, yellow, green, right. Two accounts that was like here's red flags that, you know, so and so said such and such. And it was like kind of a slightly off comment that may have, you know, signaled whatever. And I was like, oh my God. Like I've been trying to figure out how do we build something where like the people are putting in the inputs so we have like a health status of everything. Yeah, I'm like, I just built it for me without me even asking. Right. Like crazy. And I think that's the thing, like to your point, Rio, about like, no more power users. Like, power users, they're gonna just get whatever they need. They're gonna make it custom to them because they don't have to like, fit in your little box. Right?

Brett: Yeah. And how difficult was it to build that, that COO agent? And what did you build it in?

Drew: So it's all Claude code. And um, what's crazy is like when I started it, I started kind of baseline building like context layers. And now because it's like, I've given it enough context, we use Fathom as our meeting transcript note taker. I use the API, I pull down like 1500 transcripts. And then I had it like triage and organize them all into like buckets of different things. And then I have context layers. And then I built this like, coo. I have a cmo, COO T. Oh, and cfo, uh, and they all have like teams underneath them. And I can, I can literally just in terminal say COO command and it'll be like, boom, spawn sub agents, delegate and they'll go do the stuff. Like, truly, like I'm, I'm actually pulling in data from all these different systems using API and mcp. I'm not logging into any of those tools anymore. Cause I'm like, why don't I need to. Like, why am I going to go to your thing? Right.

Rio: You know, like, I actually think like, what you're describing, I mean, it's going towards like the end state of all this. I think this is the interface we're ultimately going towards.

Brett: Right.

Rio: It's like Star Trek. Okay. Computer. I mean, Star Trek predicted a lot Right. But I think that's another thing that predicted. Right. Is that is the end. Uh, I don't know when we get there, it probably will take a long time, but I think ultimately that is where we're going to get towards where we're interacting with machines. The way we interact with way way millions of years of evolution designed us to interact with each other. And then, and then because that's going to be the. Maybe we even goes direct to our brains, I don't know with implants, but I don't know if people will be comfortable with that. But I think like for all intents and purposes, this is the end state that we'll eventually drive towards over time how we interact with them. And all the things you're talking about, MCP servers with APIs, all of that's going to happen in the background and will be abstract. I actually heard like that some companies, a lot of their code base now is not actually code anymore, it's just language.

Brett: Yep, yep.

Drew: Yeah, it's funny, I read something and ah, it was like perfectly said. It's like English is the next coding language, like it already is arguably. It's like because we've abstracted the code to the level of English. Right. Where you don't have to actually like know how to write any of these, you know, Python and all this stuff. Like it just, you just speak in English and it just does it, right?

Brett: Yep.

Drew: Kind of wild.

Brett: That is wild.

Drew: Well, uh, this has been great conversation guys. Like I fully expected it was honestly way better than even I thought it would be. Um, you guys are super interesting. I always like to leave folks with like some practical application, like you know, people who are listening product business marketing leaders, you know, mostly in the mid market and they're all trying to figure out like how do we do this stuff, like where do we focus our energy. So I love your perspective on that.

Brett: You want me to start go for it? Uh, yeah. So I think, I think think of go to market one, define your terms. Think of go to market as less about messaging and more of building an operating system within your team that really fundamentally connects the promise of the product to the market, to the competitive reality, to the actual buyer types and get very nuanced that way and then ensure that that information like cascades through the organization. Because what I find a lot of in companies that I consult and work with and in my own experience just having worked within GTM and in product and marketing organizations is that there's often strategic drift with the people involved and the teams involved, right? And so you get one thing coming from the product leadership, uh, that doesn't necessarily get smoothly communicated to the marketing organization or the sales organization. And then by the time you have people delivering product to customer, it's a. I always say this in our podcast. It's like the Chinese whispers, right? They. It goes around the circle and it. And it's lost a lot of its original meaning or it's changed radically. And then you've got people delivering something that may not be anything like what the product team or the marketing team or the sales team promised the client. And so getting those sort of decision mechanisms and strategic mechanisms planned in, executed correctly, leveraging AI, which I think compounds your decisioning and, you know, predictive and critical thinking capabilities, and then being able to seamlessly get that through the organization to ensure that everybody's got the same playbook and that you don't get this breakdown along really what is one journey? Because oftentimes I find customers have multiple journeys within an org and it's extremely frustrating and they, and they end up being unhappy at the end of the day if it's not. If it's not delivered correctly.

Drew: Love it. How about you, Rio?

Rio: I guess my advice would be people tend to spend time doing things they are used to doing or like doing maybe some combination thereof. So I think that as a business leader, you might find yourself doing that naturally as a, As a course of habit. Um, stop. I mean, I advice to be. Stop doing that before you do anything. Think, okay, can I, can I delegate this to AI? I mean, it used to be kind of delegated to someone I'll hire. But you can, like, everything's sped up now. You can go open a prompt and you can write a really nice prompt that, uh, ask the right questions, provide your information. You could probably get some output that would be very helpful. So I'd say before you focus on the most, the highest value things, the things Brett and I were talking about, Drew, the things about, like, what does your ideal customer profile look like? What is you, you know, what concrete steps you need to do this. Like, how are you going to. Like, how are you going to. Where are you going to spend your marketing? The real decisions that need. The things that need you to make a decision. There are going to be some things that really do stop at you. Focus on those things. Have AI do all of the rest. I think that is going to be the biggest unlock, uh, trust this. I mean, obviously to validate everything. Right? But these tools are amazing. You should be using them as much as you can. That'd be my advice.

Brett: Yeah. And Drew, I took it from you in our last conversation. As you said, forget about 10x. I mean, it's. Each one of us has the opportunity to be 30, 40, 50, 50, uh, 100x, uh, and I'm seeing that in my own. In your example with Claude code and being able to build out, uh, simulated leaders, C levels across all the core functions of the organization that you can then go. And I'm thinking about building Steve Jobs as my boss.

Drew: Right.

Brett: But you can sort of simulate that and train it and then use it to make you smarter and better at your job. Right. Easier said than done. There might be some technical components to that, but I think people have got to start trying and start really leaning in and you'll realize that a lot of it is conversational and doesn't require super technical capabilities.

Drew: Yeah. I, um, I'll leave you with this, like, little story I just recently read, the Goal. It's like a. I reread it. It's an old book, right? 34 years old. And it talks about the theory of constraints. Right. And the, um, in your business. And I had Claude go do research on the book, extract kind of the key insights. And then because of the context it has about me and my company, I had it basically tell me, like, hey, act as though you're kind of this, you know, you're. You're advising based on this, this framework. And it came back and gave me a whole bunch of really, really great takeaways. And that took me. I mean, like, anybody can do that, right? You can do a ChatGPT, you can do a Cloud Gemini. And like, those are the kinds of things that, when I'm talking to business leaders, I'm like, like, just start using it honestly. Like just. Just like, uh, once you. Your mental model shifts to like, what's possible with it, it's like, sky's the limit. Right? But a lot of the. I think the. Frankly, there's a lot of folks who are just afraid or they are sort of like, well, you know, it's not really my. I'm not a techie kind of person. You're like, like, don't need to be that.

Brett: That's a scar. That's a scarcity mindset versus. Versus an abundance mindset. Right? And. And it really is important. And you get a lot of fear from nine out of ten people I talk to. Is AI the first thing you hear when you talk to AI? Uh, just in cocktail, you know, gatherings or whatever. Is it hallucinates. Can't trust it. Haven't. Haven't used it that at the end. So a lot of people are just dabbling and there is a lot of just general misunderstanding on how to leverage it to its abilities.

Drew: Yeah, for sure. Well guys, this is great, uh, great conversation. I appreciate you being on the show. Where can folks find you online?

Rio: Signalandnoise AI. That'll be up actually Brett, this week. Right. Um, that's our creator's website where we have all of our podcasts going back. Going back about a year. We also have our articles. Right, Brett? Articles that Brett and I write as well as like a curated group of executives who contribute to that are publishing on our platform as well. But also Apple podcasts, Spotify and YouTube. Signal and noise. Um, you can look it up. You can look Signal Noise, Brett Rio. It'll come right up. And we're available on all those. Always love more subscribers and check us out if anyone wants to appear on the show or just chat about stuff. Hit us up anytime. Drew, it's been a pleasure.

Brett: Yep. And ah, you can find me on LinkedIn. Also hisignals.com. there's a lot of signal going on in the names of the companies that I'm involved with. High signals.com, signal, signalnoise AI which is launching early next week. Both are, you know, we're founders or I'm m. A founder and a builder as well. And uh, so love what you're doing over there, Drew. It's Stealth X. And uh, uh, yeah, looking forward to talking to you again.

Drew: Yeah, for sure. Well, awesome guys. Thanks again for being on the show. And uh, folks, you can find me online. Drew burdick.com and on every social platform. Drew Burdick, uh, Drew Hberdick. So appreciate it guys. Good conversation. See ya.

Brett: Like comment and subscribe.