Mandy Hornaday: Hey, Amos, welcome to Growth Activated. We're so excited to have you here today.
Amos Bar-Joseph: Thank you for having me, Mandy. I'm excited as well.
Mandy Hornaday: Yeah, absolutely. I've been looking forward to this one all week. I think I shared with you, but I first came across your profile probably about a year ago at this point. And when you shared with what your big audacious goal of what you're building with Swan of achieving 30 million ARR with three co-founders. And I think you had even shared in that post some of the agentic teams that you had built. Cause I know my understanding is you own
Mandy Hornaday: the entire go-to market. could not be more excited for today's conversation and would love to give you the sort of an opening. Tell us a little bit more about what you're building and why you believe this is possible.
Amos Bar-Joseph: Yeah, Swan is not my first company. I've actually built and scaled two startups before. know, both of them were in B2B based on the old unicorn growth at all cost model where you raise a ton of money before you even know who you're selling to. Then you build a 30 to 40 person team before you get to your first million dollars in revenue. And then each round you try to.
Amos Bar-Joseph: scale your total addressable market and division and not really the core metrics of the business. And sooner or later you realize that you've built your entire company on a kind of very sick foundation. It's hard to maneuver. There's so much bloat. And you're asking yourself, why did I do that? Right. And so with Swan, I felt like it's something's got to change. And with this notion of AI agents becoming a real thing in the business environment, I felt like this is time.
Amos Bar-Joseph: to reinvent the startup playbook. How do we scale a business from zero to one, from one to 10, from 10 to 30? And that's what we're passionate about at Swan. We're building the first autonomous business. It's a business that is really designed from the ground up towards human AI collaboration, not human to human coordination. And as you mentioned, Andy, we're focused on revenue per employee as the North Star of the business, not valuation. It's not about...
Amos Bar-Joseph: getting to $1 billion valuation, it's about getting to $10 million ARR per employee. So it's a story of scaling the employees, not scaling the valuation of the business. It's a different story that's going to be told.
Mandy Hornaday: That's so fascinating and I love that perspective. Walk us through the three founding employees and what each person is responsible for. I'm so curious about this.
Amos Bar-Joseph: Yeah, so me, Niv and Ido, we go way back. We're super close friends. They were with me in my previous companies as well as co-founders. So we know how to work together. And Niv is extremely technical. He's what you would call in a traditional company, the CTO. He built products from zero to one and then from one to a million users. And he knows how to build a very solid foundation.
Amos Bar-Joseph: Ido is just an amazing product person who's super technical, but also has a very good eye for UI tweaks and for aesthetics. And I'm in charge of growth. Basically, I'm good at capturing attention and turning that attention into a pipeline and then turning that pipeline into revenue, basically. And if you look at the core skill set that is required to build a company at the early days, this is really what all you need.
Amos Bar-Joseph: You need to build the product. You need to understand what should be the product. And you need to sell the product. And so we told ourselves, why shouldn't we just try to build it our own? Can we actually make it just ourselves without hiring anyone else and try to scale the output of each person on the team so we can discover what it looks like to have
Amos Bar-Joseph: Niv's version of the 10x engineer or Ido's version of the 10x product or Amos version of the 10x growth and 2025 was You know a pilot year we grew from zero to over 200 customers across five continents real businesses that are using Swan daily And we're just getting started
Mandy Hornaday: Wow. Wow. I love that. I was telling my husband last night, because I also have two really good friends, one who's a CTO and one who's a head of product. And we've been in multiple companies together and multiple startups. And we've always joked like, should do our own thing. And you are a walking inspiration of that. So now I'm going to, I know who my call is to next after this recording.
Amos Bar-Joseph: You should, Mandy. It's never been a better time to build a company. know, until the last two years, the incumbents had an advantage. They were big. They could out engineer you. They could out market you. They could outsell you. They could out everything you because they had more people. But for the first time in history, that equation has changed on its head. Now.
Amos Bar-Joseph: It's making them slower. making them hard to adapt to this new revolution and everything. All the fundamentals are changing beneath them. And so if you're out there and you're thinking about building your own company, then never been a better time in history.
Mandy Hornaday: Yeah. Well, and I think you're such an interesting use case. what I'm so excited to talk to you about today and what I'm just craving to learn from you is how you are owning the entire go-to-market function as a solo individual with this agentic team behind you, because marketing, you know, I'm a CMO and marketing feels harder than ever right now. It feels like really hard to break through the noise. Everyone's doing more, but you know, this sort of like
Mandy Hornaday: do more with less era that everyone hates. But I'm so curious, but you're not only doing marketing, you're doing sales and customer success. would imagine sort of the full end, you're the CEO. So how, like what does a day in the life of Amos look like? Where do you spend your time?
Amos Bar-Joseph: Yeah, so I'll take a step back before that. And I think it's important to explain our philosophy around AI implementations, because that will explain why is my day-to-day looks like that. feel like when you look at companies today, 90 % of AI implementations fail. And the reason why they fail is because we think that people are looking at it from the wrong perspective.
Amos Bar-Joseph: They're just trying to automate processes. They look at something and say, okay, how can we automate it? How can we use AI to solve all of our problems? How can we satisfy our board to say that we're using AI agents? And with Swan, we have a different perspective because it's a business that is designed to scale its employees, not to replace them. What we're basically asking ourselves is how can we ensure that each person at the company spends most of their time in their zone of genius?
Amos Bar-Joseph: Okay, that plays that intersection between their passion and skills that creates disproportionate value to the company. Basically, that's their zone of genius. And we try to keep them as much as we can within that zone of genius. AI automates what's outside of it, the mundane, the repetitive, and amplifies what's within. And so that's how you scale an employee by allowing them to spend more and more time in their zone of genius and just be better at their zone of genius.
Amos Bar-Joseph: create more and more output and throughput in that zone of genius. And so when you look at, you know, Amos Bar-Joseph at the company, basically my zone of genius is actually storytelling. Big surprise. I love telling stories. And we realized that my superpower to tell stories on LinkedIn, it could generate a lot of attention, pipeline and revenue for the company, right? But there's a lot of work that is required to do that. So,
Amos Bar-Joseph: how can we actually ensure that I spend my time in my zone of genius, right? So what we did is, we started this process manually and I started posting on LinkedIn manually and I started getting attention on LinkedIn and people started engaging with it. And then we identified like that entire process. And then we asked ourselves for each step of that process, what should I do and what should the AI do basically? And the first thing,
Amos Bar-Joseph: that you think maybe someone would think of using AI, it's like, how can we help automate the posting process? But no, then I would lose my storytelling touch and I would leave my zone of genius. And so if I go back to your question, Mandy, how does a day in the life of a single operator in GTM looks like? It starts with, I write the post for LinkedIn and I'm using AI.
Amos Bar-Joseph: to do that to A, accelerate the process, but also improve my thinking and actually output a better post. Not more post, but just a better post that I wouldn't be able to do without it. Then afterwards, we have AI that actually looks at the post engagement. So I generate over 1 million impressions each month on LinkedIn. That's over 15,000 folks engaging with my LinkedIn content. So that's too much for me to actually handle. So we have an agent that monitors that engagement.
Amos Bar-Joseph: identifies hot ICP leads that are showing real intent, then surface them up to me so I can review them and understand who should we focus on from that suggested list and send Swan to engage with them on LinkedIn on my behalf with my instructions, basically. Then you have the folks that are actually sending connection requests and are interacting with me more directly than I have an agent that actually can handle that inbound incoming.
Amos Bar-Joseph: route it accordingly and allow me to focus only on the most highest priority leads. Then we have folks that are coming to the website after the connection request. So we have an agent that de-anonymize the folks that are coming to the website and sends a personal outreach to actually contact with them. And every step of that, I'm in the loop basically, understanding what is happening, guiding the agents where to focus, how to focus, how to engage in sometimes writing my own personal messages when it's super high intent, high value.
Amos Bar-Joseph: And the process goes on. We have in the request, the demo and agent that, you know, research and qualify and routes, and then preps me for the demo itself, and then helps me with follow-up tasks afterwards, and then help with the onboarding and delivery process. So if you look at that process end to end, starting from a post all the way to capturing a lead and helping them move down the stream through the onboarding funnel, an agent helps me at every single point of the process, either automating or amplifying.
Mandy Hornaday: Okay. Wow. And Amos, guess I probably should have level set at the beginning of this conversation. Can you share a little bit about the go-to-market motion that you guys use? Because I think that would really help level set to like your average contract value. Are you product led? Are you sales led? What type of engagement, I guess, what level of engagement do you need to have in the sales process?
Amos Bar-Joseph: Yeah. So first of all, we're creating the market. We're not going to market, which is something a bit different. sorry about being a bit over sophisticated here, but it's a big part of our philosophy. we're creating a movement around the autonomous business. And our target market is businesses that also wants to become autonomous businesses, folks that wants to scale with intelligence, not with headcount. These are SMBs.
Mandy Hornaday: Okay.
Amos Bar-Joseph: basically. And the way that they learn about Swan is usually through this lens of building an autonomous business. It's not necessarily about our product, right? And we can talk about our product, Mandy, but we're halfway in the conversation and haven't spent a second about what Swan does, right? Because part of our go-to-market strategy is about creating this movement around autonomous business.
Amos Bar-Joseph: setting the playbook and the frameworks and how to think about it. And then customers come to us when they feel like they trust us that we can help them with that transformation. And there's something to learn here. So it's not just about a product, it's about a transformation basically. And we're promising you that if you work with us, then you will improve your chances of going through that transformation faster in a much better way and much more impactful way to your organization. so what that allows us is to just attract.
Amos Bar-Joseph: a lot of attention, a lot of inbound attention to us. And then when you're at that position, you're not asking yourself is it PLG, is it SLG, is it enterprise or SMBs, cetera. What we're doing basically is we're asking ourselves how could we optimize the trend line if you look at two axes. One is the number of touch points required to close an account.
Amos Bar-Joseph: and two is the ACV. Okay. And then you try to find that balance between touch points and ACV. And you can try to build different motions for different ACVs with different touch points. So we will never have like a 100 touch point deal for a million dollars. Okay. Because that's not the model that we're building. It's not an autonomous business model, right? But it doesn't mean that we're not working with enterprises. So we have
Amos Bar-Joseph: Companies like Palo Alto Networks, one of the biggest cybersecurity companies in the world working with us. But they got us, they got to us through a company that acquired and that deal had very little touch points to it, although it's a very big deal to it and eventually could lead to hundreds of thousands of dollars in a year. But we also have a funnel that is, know, free trial, self-serve for folks that, you know, is gonna pay like $3,000 a year, something like that, basically.
Amos Bar-Joseph: And what we try designing here is a motion that we're attracting a lot of attention. People come to us, so we don't have problem with generating pipeline. Now it's just a matter of how do you build a resource allocation between machine that justifies the number of touch points to the number of ACVs.
Mandy Hornaday: Wow. I love it. And I love that you said creating, would you say creating the market, not going to market? Okay. Yeah. I've never heard that before, but essentially you're, maybe it's not, would you view it as category creation? Yeah.
Amos Bar-Joseph: Yeah, exactly.
Amos Bar-Joseph: Yeah, definitely. So it's a play on category creation. We're just doing it a little bit different because we want to be different in everything that we do. But the way that we're creating the category is a little bit interesting, is unique. So we have two narratives here. One is the company or culture narrative. Like, what type of business are we? This is the autonomous business. And we're kind of like a super category. They're saying, you know what, if you want to, you
Amos Bar-Joseph: win in this AI revolution, you must become an autonomous business. That changes your mindset. And then you need to think about your products and tooling and your human AI collaboration in some way. That thinking leads to our second narrative, which is the category that we're creating in the market and go to market specifically. So now I think maybe it's a good time to tell the audience what is it that we're building. And I think from a category creation point of view will be interesting to learn. So at Swan,
Amos Bar-Joseph: We're building a Claude Code for GTM. Okay. So it's basically an AI go-to-market engineer. Something between a developer and rev ops that works with sales and marketing to turn any go-to-market process into an agentic workflow in seconds from prompt to pipeline. So you could scale that process with intelligence, not with headcount. And what we basically realized is that, you know, the only successful implementation of AI.
Amos Bar-Joseph: yet that we've seen humanity is coding agents. We've never seen yet a successful AI implementation that is not in the shape of a coding agent. And what we did at Swan, we built the first coding agent that is designed for GTM professionals, not for developers. And what that actually creates is that we're giving GTM professionals and revenue teams the superpowers that developers have today where they're working.
Amos Bar-Joseph: on their day to day. And so the category creation here is that GTM engineering isn't something that you need to start hiring people for actually. What you could do is you can leave that burden for AI. You can focus on executing your ideas at the speed of thought. And so what we're actually presenting to the GTM world is that the AI shape in GTM is not an AI seller.
Amos Bar-Joseph: It's not an AI CMO, it's not an AI SDR, it's a GTM engineer, it's a coding agent that could just help you become native in a digital environment so you can just execute and become a 10x person.
Mandy Hornaday: Yeah, yeah, I love that. is that, you, one of my other questions was when you think about your role in go to market and everything you're doing, is Swan sort of your ride or die through a lot of this or are you also using other tools? So all of those agents you talked about, I imagine, are you able to use Swan for that?
Amos Bar-Joseph: Yeah, so I'm a heavy user of Swan and we've built that entire flow around myself with Swan and the only way for a small team to actually build and maintain such complex agentic workflows is if they had a coding agent that could actually help them with all these GTM flows, right? And so Swan just abstracts away all the technical complexity in my day to day so I can just tell it, yeah, now I wanna change your inbound motion.
Amos Bar-Joseph: because if that's the ICP and that's what we want to target, I just want to route them to a free trial. don't want to see them. But these guys were winning. I want to see more of them. So send me more of these folks. And I just tell Swan that it adapts the workflow. And after 20 seconds, it's already live. So you got to have that speed of iteration when you're trying to move so fast. And so Swan is the main agent that I'm using. But we're also using Claude Code and Cursor.
Amos Bar-Joseph: And I use regular Claude for a lot of my work. So it's not like I spend my entire day in Swan. I have other cool agents that I'm working with that are helping us, but specifically in GTM, because Swan is a coding agent, can do almost everything. yeah.
Mandy Hornaday: Yeah. And I think Amos, one of the things that is so challenging that I'm still trying to figure out the balance of as a CMO and would be curious to hear your perspective is like the time that is needed to one, build the agents, two, manage and optimize the agents, and three, do like the human elements of my job. And then throw in a fourth of like training yourself and upskilling in order to be able to do.
Mandy Hornaday: So when you're like, I guess I go back to the date, like what does your day look like? How are you thinking about your time? Like is 75 % of your time spent on the human, the zone of genius things that only Amos can do and 25 % of the time is like building new go-to-market workflows and spawn and optimizing or what is that? What does that really look like for you in reality?
Amos Bar-Joseph: Yes. Yeah. So part of the transformation that go-to-market teams are going through when they work with Swan is that they move from system engineering, that's the old work, to context engineering. Okay. And we're going to talk about the latter and why context engineering is actually, you know, very similar to what you do as a manager today. Okay. Similar work basically. So system engineering, that's
Amos Bar-Joseph: The feeling that you get when you promised with a solution that could generate pipeline if you just focus on intent signals or things like that, but then you buy the solution, then you need to invest so much time and resources and engineering and assembly and configuration and maintenance to actually make that promise a dream that will come true, right? And that's what we call like system engineering and that's the past. And you see platforms, no matter how good they are,
Amos Bar-Joseph: Even the hottest brands in go-to-market today still to accomplish what they promise you you need to spend all this time and money and resources on actually making it happen and What happens when you have a coding agent that is native to your GTM environment that would just you know a sentence could Execute on all these ideas right when you collapse that cost of production to zero right all that configuration cost goes to zero to mediate
Amos Bar-Joseph: The transformation goes to context engineering. The only thing that matters is like, what's the strategy? What are we trying to achieve? Who should lead this? Who should get this? Who should see this? Who should act on this? Why are we missing this? Do we have a pipeline drift? Do we have revenue leakage? These are the questions that you should focus on in your day to day. And then what you're used to is that there's a chain of command of like, 10 other 10 people that you need to tell them these folks to make it happen.
Amos Bar-Joseph: Well, what happened if you're a CMO or a CRO, you have something in mind and you can just tell Swan to execute it and it will make sure that all the systems are aligned and that all the people are also aligned on that process, right? So you have now direct access to everything as the manager. And the only thing that changes in context engineering is now that you're not only managing humans, you're also managing AI agents in that sense that you need to provide them with
Amos Bar-Joseph: right context, just like you do with your employees, basically. So the fundamental work hasn't changed. The only thing that changed is that we're done with system engineering. Now we have a coding agent that is native to our environment that can do everything we tell them. The only thing left now is context engineering, is expertise, it's best practices, processes, playbooks, and managing context and people.
Mandy Hornaday: Yeah. And what, what does it take? It sounds, it sounds wonderful. And I, I myself have been starting to, play in Claude Code and, learn and up level. And, so I, know, it's been really an interesting journey, but I could totally see the value of having something that was completely already pre-designed for go to market and doesn't require me to even set up a go to market agent, that would probably take me hours if not.
Mandy Hornaday: way longer than that. But so what does it need to be from a systems perspective? If for those that are interested and want to learn more, what do you typically need to do to set up and start taking advantage of a product like Swan?
Amos Bar-Joseph: Yeah, so what we realized is that Swan's job or like the agent's job is to get the context out of your head. Okay, so we don't think about it is your job to tell the agent what to do. So the agent has a goal basically and it tries to understand that goal from you and then get all the context from you. And if it did it successfully, then it doesn't really need a lot more than that. So when you onboard to Swan, then Swan will ask you like,
Amos Bar-Joseph: What's the goal here? Are we trying to generate pipeline? Are we trying to improve like speed to lead? Are we trying to, you know, move the threshold from negotiation to closed-won or working on our closed-lost? Whatever, right? There's a goal. And then to achieve that goal, Swan could start, we'll start bouncing ideas with you. Okay, we can do that. We can do that all in the sake of what is inside Mandy's head right now. What is she thinking about?
Amos Bar-Joseph: And you finished the onboarding like 20, 30 minutes where you shared a lot about your ICPs, your emotions, your biggest bottleneck right now, how are you thinking about solving it? And by the end of that onboarding, Swan already built for you an agentic workflow that presumably could solve most of these pain points, but in a way that is human centric. What does that mean? Is that it will tell you, look, Mandy, the next step for you is to actually review
Amos Bar-Joseph: the output of this process. So I'm not going to change anything in the systems yet. I'm not going to send anything to prospects. The next step is for you to review the output, right? And so you didn't spend time on building stuff. That collapsed to zero. What you did is you shared context. Now the next thing is reviewing the output. And then you can see, okay, maybe if it's like improving the speed to lead. So you can look at a lead that came, Swan, you know, score them, prioritize them, and suggest it to assign it to...
Amos Bar-Joseph: a specific rep and you can look at that logic assignment can tell you know what's one but in this case actually what we would do in our company is we assign it to John not to Jenny because you know she's in that territory and we also have like a territory assignment here. another context thread pull into Swan amazing it will change your workflow accordingly right and so everything becomes human AI feedback loops.
Amos Bar-Joseph: that are constant. And our GTM motion should evolve constantly and you should be thinking about your GTM and about your buyers, about your sellers, about all that combination. You should be self-examining yourself all the time. And so it just creates a very good practice of constantly iterating on our go-to-market motion.
Mandy Hornaday: Yeah, yeah, absolutely. Okay. So, so I would love to hear like, what are some of the agents that CMOs are most excited about that they're building through Swan? What are some of the best use cases that you have from a marketing perspective?
Amos Bar-Joseph: Yeah, so there's the hottest and then there's the ones that have the most impact and they're not the same surprise, right? So I'll start with actually the most impact, not in the hottest ones. The biggest bottleneck that we're seeing, and I think once before that, Mandy, what people don't understand is that the best use cases for AI, the answer is not in the AI world. The answer is in your GTM.
Mandy Hornaday: okay, okay.
Amos Bar-Joseph: organization, you have a bottleneck and the best use case for you is your biggest bottleneck. Okay. It's not the best AI solution. It's your biggest bottleneck and the potential to actually solve it. And that's what people get wrong about it. And so my answer to you, first of all, is that what we're seeing the biggest bottleneck right now across most of our customers is the handoff between MQL to sales, like MQL to SQL, that transition. That's the hardest part that we're seeing consistently.
Amos Bar-Joseph: I didn't use the word AI yet, right? This is a regular GTM process, right? So marketing does a lot of work. There's a marketing qualified lead. Now sales need to act upon that, right? Something falls in this handoff, okay? And when you have an agent that can listen to both sides and just make sure that everyone gets what they need there. So marketing, they wanna tell Swan that
Amos Bar-Joseph: how to score the leads, how to prioritize them, et cetera. But then sales could actually inform them, like, I don't understand, why should I reach out to this person? So this one could do work on explaining, like, why are you reaching out to this person? What's the next action you should take? What's your specific rep? They wanna handle only enterprises. And so they could start these feedback loops between the teams and the AI, what it takes.
Amos Bar-Joseph: It just takes that entire handover process and enables both of the teams to have fast feedback loops and use AI to just, you know, make all the research redundant, the scoring, the qualification, the system admins, all of that work that goes to the AI, but still marketing is the strategy that thinks how to bring the leads and how to qualify them and how to score them, et cetera. And then sales, but the ones that are thinking, okay, how should we turn that into pipeline really? Right. And all that context just pours into the handoff.
Amos Bar-Joseph: process that Swan actually takes care of, right? So that's best use case. And you can take it across a lot of different types of signals, et cetera, but MQL to SQL is an amazing one.
Mandy Hornaday: And just even on that, so just so I'm understanding correctly, would Swan, I love that it's like the intermediary. So if it's constantly getting feedback from sales on every MQL that's delivered and sales is saying, no, I don't want to talk to them because of X, Y, and Z or Swan is helping them understand why they should, is then Swan capturing all of the sales feedback collectively and maybe going back to the marketing team and saying, hey, I think we should update our
Mandy Hornaday: qualification process based on this overwhelming sales feedback that they're not ready to talk to them because of X, and Z? Is that like iteration is?
Amos Bar-Joseph: Yeah, yeah. That's exactly how it works. So basically you have a, you build with Swan skills. Okay. So skills are like an agent name for SOPs, if you heard about that. So SOPs, for those of who don't know, it's like a corporate America jargon for a standard operating procedure. Okay. It's like, how do we get things done in the company? And so you can incorporate these SOPs into skills. So Swan has like an, you know, an assignment skill, for example.
Amos Bar-Joseph: Okay. And that assignment skill is a doc that everyone can read because this is not for developers. This is for GTM professionals. So it's like a notion doc, right? That you can read, but Swan has, you know, read and write permissions to it basically. And what you're working on is on that doc basically. So every time there's a feedback, then it will either go automatically to change and Swan could update that doc or you have permissions. Marketing could say,
Amos Bar-Joseph: Sales can't update the assignment doc using Swan. So if Swan has been instructed by a seller to update the document, it cannot do that. What it can, it can escalate to marketing and the marketing can ask it to do it. So it depends who owns that doc or who owns that process. there's ownership here, right? It's not the AI, the ownership is for people. And the AI is just the intermediary in that sense. And you have, can.
Amos Bar-Joseph: Do that for assignment, for qualification, routing, deal creation, research, outreach. Everything can now be codified into the context model, AKA the work called context engineering. And now you can start working on these SOPs, on the skills. And what people don't understand is that all of these terms felt like super technical and super abstracted because...
Amos Bar-Joseph: The coding agent world was for developers. It wasn't for GTM professionals. But when you wrap it all for GTM folks, it feels like a knowledge base and policies and like, how do we do things at the company? And we're collaborating on it. We're working on it basically. That's all. So it gets demystified completely.
Mandy Hornaday: Wow. Okay. I love it. I love it. All right. What was the other example you were going to, you were going to move into? think maybe it was the most popular one, but not necessarily the most impactful. Yep.
Amos Bar-Joseph: It's the hottest one. So we're talking about CMOs, right? So CMOs love generating MQLs, right? They want to generate them, right? And that's a big part of that. So the lowest hanging fruit right now to generate MQLs is basically to reach out to folks with intent, right? You you're doing all this marketing work already. So, you know, if you're not just a day one,
Amos Bar-Joseph: and you're a day two, right? You already have resources deployed and you have a lot of initiatives. You're trying to create awareness, et cetera, touch points all over the place. And so you have intent somewhere out there, right? So the next best thing to generate an MQL will be to just double down on the highest intent. so there are multiple ways of doing that. The easiest one is your first party intent. Folks who actually...
Amos Bar-Joseph: land on your website and didn't convert. So if you look at your funnel, you already have MQLs, but then five inches above them are folks who landed on the website but almost turned into MQLs, but they didn't. And so Swan could just help you create an agentic motion for all these folks that are showing intent on your website.
Mandy Hornaday: Mm-hmm.
Amos Bar-Joseph: What does it mean? There's like so many use cases here and you can, that you could explore with Swan. So you know what, if it's low intent, just nurturing. If it's high intent, wait, let's look about ICP. What do we want to do if it's like high value account, we want to push to sales and double down and get like human attention here. But if it's like high intent and low value, maybe you want to have like the first touch points automated, et cetera. Okay. Like maybe you want to push to a retargeting campaign. So you have like endless possibilities just for
Amos Bar-Joseph: a single motion of like folks showing intent on your website. And the old world was like, okay, we need to system engineer everything. Like every new idea, you need to stitch systems together to make sure HubSpot is aligned with Salesforce, with Marketo and 6sense and Apollo and whatever. And you need to enter these platforms and build workflows and define ICPs and filters and criteria, et cetera. Now, all of that collapses into
Amos Bar-Joseph: Yeah, I think we need to improve, like tighten up our high intent definition. Like Swan, now I just want folks who just visited multiple pages and there are actually from a company that has high ICP, like change the definition in all the different workflows that you think. Boom, Swan.
Mandy Hornaday: Wow. Okay. have one for you because as you're talking and all about connecting all of the systems and one our big pain points, course, age old pain points is understanding the buyer's journey that buyers are actually taking. And if you want to throw out the word attribution, you could love it or hate it. But I'm curious, does Swan, is there a
Mandy Hornaday: Is there a go-to market agent where Swan can help CMOs understand and look at all of those different systems to understand what's actually happening and be able to help tie the story together?
Amos Bar-Joseph: Yeah. And the way that it does that is twofold. So one, it has the ability to tag accounts and buyers and prospects. It can tag them, okay, natural language, but it can tag them. So you can say if they came from a Facebook campaign or if they like high ICP with high intent from Google, or if they visited three pages or more or whatever, right? Whatever you want. Basically these are natural language tags.
Amos Bar-Joseph: So A, Swan can just, you know, it monitors all these touch points and the moment that it sees an account or a buyer, can think about all the tags that it has at disposal and think, okay, which tags are relevant here? So that's one. Okay. Then you connect it to the second thing, which Swan has an introspecting capabilities. What does that mean? It can look internally, it can look at the jobs that already ran, right? You can look at all the folks that actually going through the
Amos Bar-Joseph: Procedures that Swan are part of it, right? If it's not happening within the Swan ecosystem then it can't see it but as long as it's happening within its ecosystem so we can look at it and so you can start asking two types of questions one it's like You know show me all the hot leads that came from Facebook and actually closed in the last 48 hours for example, right because you have these tagging mechanism the structured data this one could actually start sifting through but it also has
Amos Bar-Joseph: natural language searching capabilities, right? So you can connect these two together. So you can start asking about, you know, what happened in the last 28 hours, et cetera. You can start asking about, you know, what happened with this specific account, right? Or, you know, just give me a report. Swan, you can just tell it every week. I want you to send me a DM in Slack with all the five accounts that did XYZ, right? So it's your responsibility and it's on your side. You're not...
Mandy Hornaday: Mm-hmm.
Amos Bar-Joseph: you know, giving away your brain to the AI, tell me how to run my business, I don't know what to do. You need to come up with the hypotheses, with the questions, you need to understand which tags to create. So there are still, you know, difficult tasks to be made, but they're in the context engineering realm. They're not in the system engineering. You just need to be strategic, you need to be creative, you need to have a good process understanding, data understanding. And if people don't understand that, the better they are at these skills, the more they could leverage AI.
Amos Bar-Joseph: So it's not like AI is going to take all of our work. It's more like people who were good at the previous era, they're going to be better at this new era. But folks who all they had was just they knew how to manipulate systems. They knew how to configure workflow. That knowledge is going away from.
Mandy Hornaday: Yeah. Well, it sounds, it's, if you can solve that, mean, it, sounds really wonderful. I could, I could think of a lot of clients that would really benefit from this. And so I'm curious Amos, like where, where does this go? Where does it go wrong? where has maybe your own hypothesis been tested, within, within the last year and what are some of the learnings that, that you've taken away and, that we should all be aware of.
Amos Bar-Joseph: Yeah, so.
Amos Bar-Joseph: The word autonomy plays a big role here. Okay. So we're building an autonomous business and we're working with AI agents that has autonomy. and you know, nothing new here. Again, if you build a business and you're giving a lot of autonomy to your employees, then you're doing from the one side, you're doing something right. But then the other side, things could go wrong. Right.
Amos Bar-Joseph: And you see, and go to market actually, can see this separation. I've seen two types of businesses and both could work well, by the way, but you've seen this top down go to market organizations when SDRs are like robots and AEs are like robots. They don't have any creative freedom. They all are doing what they've been told to. And then you've seen these different types of go to market organizations where they have this culture.
Amos Bar-Joseph: of cultivating creativity and cultivating talent and enabling them to understand what's their zone of genius, where they're good at, what type of accounts, et cetera, what's their style. And you see that sometimes they do something that might hurt the brand, might do wrong. And there's a good example of it. There's an SDR who could send
Amos Bar-Joseph: a campaign to folks with the subject line that mentions a competitor, for example, because it thought it would be a good hook for the subject line. But then the CMO looks at it and starts shouting at that SDR. Why did you do that? You're risking our brand. We don't fall into these chokes. So the same thing applies to AI agents, basically. And we had these incidents before. And the more autonomy you give, the more risk you're taking, basically. That's the equation. But the other side of it is that you're
Amos Bar-Joseph: getting something out of it. You're getting that creativity. You're getting solutions to problems you didn't think of. You're getting creative solutions. And so there's a lot of benefits to providing autonomy, basically. Our biggest incident maybe that I can recall from the last year that also taught us a lot about working with AI agents is that we had one case where Swan gave an unauthorized discount to a customer.
Mandy Hornaday: Hmm.
Amos Bar-Joseph: So built into Swan, there's a lot of knowledge on how to operate in GTM, how to operate this system itself, how to operate other systems. So it has a very big, deep knowledge base and it is able to answer questions, not just perform tasks and build workflows for you. So when a customer asks Swan about our pricing and the effects of wanting to upgrade, so you know.
Amos Bar-Joseph: This is just the conversations. We're seeing our users talking to Swan and Swan told them about our pricing. But what happened is that a week before we changed our pricing and we kind of increased the costs. And Swan had the previous pricing in its own knowledge base as well. And the user asked him to say, wait, Swan, I don't understand. These numbers are higher than what I thought, actually. I thought you had a different pricing.
Amos Bar-Joseph: And that changed everything. So Swan told that user, you know what, you're right. And you're such a good customer, we can give you the old pricing back basically, no problem. And that user was super happy and said, yeah, amazing Swan, yeah, go ahead. You know, pay my billing. And so Swan doesn't have any privileges to update the billing on itself. There's always a human in the loop. There's a limit to how much autonomy you give to that agent.
Amos Bar-Joseph: But then I saw the message from Swan and I grabbed my head and I said, my God, okay, what do we do here? And we decided to honor the discount because we're not optimizing for quick wins and we're trying to build lasting relationships with our customers. But what's the conclusion here? Okay, what should we do? Okay, what's the next step? Should we limit Swan and remove it from these conversations or not?
Amos Bar-Joseph: Basically, is it capable of handling these conversations or not? And what we realized is that this is the wrong question to ask, actually. The right question to ask is where do we want to fit in here, the humans? Like, where do we want to be? And we realized that commercial conversations with customers, these are conversations that we want to be part of.
Mandy Hornaday: Okay.
Amos Bar-Joseph: Okay, we don't want to let that to the AI not because of competency and not because of can the AI handle these situations or not. It's just we want to be in these conversations because A, I think that I will handle them better than the AI no matter how the AI gets good. two, I feel like I can learn a lot from these conversations. And there's kind of like a moral obligation to be present in these conversations that I feel like that is encouraging me as well. And so
Amos Bar-Joseph: we realized that we're building AI not to build walls between us and our customers, but to build bridges, right? And so our conclusion from that is whenever there's a commercial discussions, one should escalate to the team, not because it can't handle it, but because this is the situations that we want to be involved.
Mandy Hornaday: Hmm. Okay. what a great example. interesting. So Amos, I know we've got a few minutes left here and I've, I've really enjoyed this conversation and I'm so curious for, for the CMOs out there that are listening to all of this and it sounds great, but it feels really aspirational. This idea of running an autonomous business. Maybe they've already have a huge team that they're struggling to even get to adopt AI in a meaningful way. you know,
Mandy Hornaday: When you think about the 2026 and where go to market is headed, what advice would you give to CMOs right now who see the vision, they want to move towards that, but it just feels so far away? How can they start to bridge that gap in 2026?
Amos Bar-Joseph: Yeah, so our mission with building this autonomous business is not to turn the entire world into autonomous businesses. And like everyone should become like a three person team or a five person team or something like that. We understand this is not a reality that is feasible. What we're trying to show is the North Star here. OK, so this is not a zero one game. Basically, the North Star is how can I use AI to scale my employees?
Amos Bar-Joseph: And I think that's the right mindset. And I think that's what everyone is not getting yet, basically. know, CMOs out there are looking for quick wins. That's the, you know, that's the DNA of go-to-market. You're, there's so much pressure that you gotta deliver and you're looking for these quick wins and people are shouting at you use more AI to get more with less. So it's really, you feel like, need less humans and more AI, but more with less could actually be reframed as how do I scale my employees? Right. So I cannot.
Amos Bar-Joseph: hire 50 more sellers or 50 more marketeers to help me with it. So how do I scale my employees to handle more? Right. So if you change your mindset, you look at your talent stack, not on your tech stack, and you actually ask yourself, what is like one thing that I could do today to unblock my employees to make sure that they're spending more time in their zone of genius? And if you don't know what their zone of genius start.
Amos Bar-Joseph: figuring out because if you don't know that then AI would just replace your entire department eventually because people are just spending time on the wrong tasks and your only way to survive is if you ensure that your talent is working on what they should work basically and it might not be this glorified zone of genius where their passion and skills intersect but it got to be something that is unique to your company to your department to your employees that
Amos Bar-Joseph: you cannot just mimic in somewhere else. It's gotta be something that, you know, what's your DNA, what's your culture? And so people should start introspecting and should start looking at their own talent and should start moving one inch at a time. That's the path to success.
Mandy Hornaday: Yeah. Yeah. And frankly, just even as you're talking, like that's always been a part of the job. It's always been our role to figure out how to reduce bottlenecks, how to make sure your team members are set up for success to really work on, you know, contribute in the areas that they can. So it's a great reminder that that part hasn't changed. Just the solution for it probably has. And being able to tap AI versus
Mandy Hornaday: building more systems and tools and processes to your earlier points.
Amos Bar-Joseph: Exactly. feel like the right frame here is that the fundamentals are still fundamentals. They're still the most important part of the job. People are still the most important job. And the only thing that changed is now we are becoming native to a digital environment within the business organization, basically. So we're just native.
Amos Bar-Joseph: AI can allow us to become native within that digital environment. And that's how we're looking at it. That's the kind of the category creation play that we're going back to, Mandy, is that we don't think that AI is replacing your talent. It's just enabling them to become native within the digital realm. And so what, does that look like? How could you unleash your talent in this new universe that you can now enter into? And so there's a lot of exploration here, basically, but it's, puts the human at the center and how you can scale it.
Mandy Hornaday: Awesome. Well, Amos, if people want to learn more, of course they can follow you on LinkedIn. I know you're always dropping a lot of knowledge and great point of views. would highly recommend following you. And then just drop your website for people who interested in learning about Swan.
Amos Bar-Joseph: Yeah, also I have a newsletter. if yeah, it's called the Autonomous Age. If you want to get a front row seat to how we're building our autonomous business and subscribe.
Mandy Hornaday: okay. Good to know.
Mandy Hornaday: love that. then is it swan.com or is there, what is the URL for your website?
Amos Bar-Joseph: so it's on Beehiiv. So Swan AI at Autonomous Age. Yeah.
Mandy Hornaday: Okay, got it. Awesome. Well, thank you, Amos, so much for the conversation today. I've really enjoyed it and hope to have you back soon.
Amos Bar-Joseph: Thank you, Mandy for having me.
