S2E5: "Foundations Before Automation" | Mario Moscatiello

Welcome back to the GTM Engineer Podcast, Season 2, From Hype to Operating System. Because now that GTM engineering is here to stay, we want to know where it's heading. And today we're going one layer deeper into the data infrastructure that makes every GTM workflow actually work. And because every clay table, every enrichment waterfall, every signal you're tracking is only as good as the data moving through it. That's why we brought on, uh, Mario Moscatiello, VP Marketing at Airbyte. It's the open source data integration platform that connects, uh, 500+ sources to your data warehouse, CRM, or activation tools. If you're syncing product signals, stitching together enrichment sources, or moving data between systems at scale, Airbud is probably already somewhere in your stack. And what makes Matteo's background unique is that he spent his entire career in developer tools and infrastructure. He ran growth at Pusher, the real-time messaging APIs, back when he was in London and we were living in the same city. Then moved to GitBook, where he scaled the developer documentation platform. After that, he went on to the investor side as a principal at Flex Capital, so switching, uh, the point of view from operator to investor. And he's also been board observer at Strapi before leading growth at FluidStack, one of the main AI cloud platforms in Europe. And here's the thing, he'd been advising Airbyte since 2020, 4 years before he actually joined full-time. So when he took VP marketing role, he already knew the product, the community, and the marketing inside out. So since joining, Matteo built what he calls a warm-up outbound motion. He uses GitHub Signals, product usage data, and pressing page intent through Commonroom, Octive, and Clay to triangulate exactly when an account is ready to buy. And that's exactly that approach that doubled Airbyte pipeline growth rate since he actually joined. So today we'll dig into that, how to build GTM motions when your buyers are developers, which is not as simple as other personas, the warm outbound playbook that doubled his pipeline, as well as a bunch of other things like GTM engineering, hiring, and team structure. So Mario, stoked to have you on the show.

That's awesome. Thank you guys. I'm excited to be here.

Mario, it is really great to have you. You are one of the most respected folks in this area. Your name gets mentioned all the time. It's going to be great to chat with you today. Go-to-market engineering, you know, it's not, it's not new anymore. And I know that you were evaluating go-to-market motions at Flex Capital before you came back on the operating side. What looks different now that you're back in the seat versus what you were seeing from the investor side?

Actually, that's, that's actually why I came back. So I, I was super excited of being an investor, you know, spending a few years I was still doing a lot of go-to-market work advising portfolio companies and whatnot. But when I saw everything that started to become possible with AI and marketing, I got super excited. Especially I think when it comes to developer motion products where you have so much signal and that's the beauty of those products. Like if the product works, if you have organic traction, you have so many signals and you can score and you can do a ton of things. I think what changed, like in a nutshell, is just the expectations are way higher. You know, like before you were running maybe, I want to say like a couple of experiments a week, or, you know, a few experiments a month, and then looking at results and so on and so forth. We could now be running like 5 experiments a day, you know, and like really, really understand what's working and what's not working. I think another thing that changed is that every company is now a go-to-market company, like in a way, like every some companies were always, you hear about stuff like Clay writing about GTM alpha and finding what works for you. And I think companies before AI always thought about go-to-market as a cost center. And now it's becoming more of like, no, actually we need to be the best at go-to-market at distribution because distribution is ultimately the only thing that matters. When the cost of writing software goes to zero, your distribution and the way you The faster you can acquire customers, the better in that sense.

So you mentioned, you know, this acceleration and how many experiments you can run. What specifically caused that? Is that everybody has access to create now with these tools or the data is stitched together better or what is it?

I think it's a combination of both. I think we get to it when we talk about hiring and the differences, but The way I think about every person in my team can now own their workflows end to end in a way like if I had my paid SEO person, they maybe had to wait for design to come up with a blog post header and then they have to wait for copy review and so on and so forth. With AI, all of that time gets compressed where they can go from, we can go from idea to keyword research to writing a blog post in an hour and publishing the blog post with the graphic assets and so on and so forth. And I think the same goes for outbound where my SDR manager can go from creating a list to working on creating a prospect list to working on an email copy to launching the campaign within the hour. And that's just, again, like part of what I think of owning the workflow end to end. And if you find operators that are able to go at that speed, technology is not a blocker anymore. So.

Now I know you have been in developer tools pretty much your whole career. Pusher, GitBook, Airbyte now. Is go-to-market engineering, is that hitting differently within this sector in the dev tool sector? Does it land differently there versus other categories or is it really like category agnostic at this point?

I think it's category agnostic. I think go-to-market engineering has always been a thing in developer tools. The interesting thing, if you look at, you know, what companies like Auth0 were doing back in the days, they already had engineers. You guys are good friends with Gonto, like they already had engineers in their go-to-market teams. So that's not new for developer tools. If anything, what we used to do 10 years ago was, oh, we had Segment, we had a few APIs and we were already stitching together tools. And the reason being is that out of necessity, when you have thousands and thousands of signups or interactions with your GitHub repo on the open source side, you have to sift through the noise and you don't have like hundreds of people that can reach out and call these people and so on and so forth. And so I think it's now vertical agnostic more than ever. I think if anything, a lot of things that were only doable in developer tools playbook are now doable for everybody. And I think it opens up a lot of possibilities. But yeah, engineers as part of the go-to-market teams were always a thing for us in developer tools. Yeah. Yeah. Yeah.

Back when I was at Auth0, I wish I had Common Room or one of those tools back then.

Yeah, we were kind of trying to build that, you know, and then you had this very ugly iframes within Salesforce that were trying to tell people, like, you know, in Marketo you had interesting moments and like the fire emojis that were telling you, oh, this person is showing intent. So I think it's so cool now with everything that's available. You don't have to spend time building.

Absolutely. All right, we're gonna get into trying to separate the hype versus reality now.

Yeah, Mario. Like, again, you've been at the— you're at the front row at Airbyte. You've been advising also like a bunch of other companies on the side. So you're super deep in the space. Where do you think from your point of view, go-to-market engineering feels like an actual step change and where is actually being oversold?

That's an interesting question. I think it's a step change when you have a couple of things. It's a step change when you have your sort of like PMM hat on done right. So you have product market fit, you know your personas, you know your playbooks, you know what works, what doesn't work. That can really be a catalyst. I think what you can do more damage than good is when you're starting to send thousands of emails to the wrong persona with the wrong messaging and then you sort of like don't realize, sorry, you sort of like think, okay, like why is this not working? Like everybody's doing it. You go on LinkedIn today and the average CEO gets probably like 10, 20, 30 messages a week from clay agencies promising these insane outbound workflows. And I'm like, yes, but you have to do the foundation. You know, you have to have the foundation right. There are two foundations that I think are very important for, you know, go-to-market engineers to be able to do their jobs. As I said before, one of them is the product marketing foundation. You need to know your persona, your competitor, your market, and so on and so forth. And the second is your RevOps. I think as companies grow, you know, bad data equals bad signal equals bad results. Like, you know, at Airbyte Data Integration Platform, we always say garbage in, garbage out. And I think when AI is in the mix, that's a multiplier effect. And so like, if your data is bad, if your assumptions on your market are bad, you're just going to do more damages. And I think that that's where it's oversold. These companies that are pre-product market fit, the things that they can just hack their go-to-market by using AI. And I think it's not always the case. I think When it comes to these earlier stage companies, yes, founders can do way more now. It's great. Like founders don't have to worry about deliverability. They can just do outbound and so on and so forth. It's amazing. But at the same time, I'm still a big fan of, you know, going in the concentric circle motion for those companies. What I mean is like you start with who do you know, who can get you an intro and so on and so forth. And I think a warm path can be way more effective in those cases. The other way that I think it's oversold is when you just think you hire like an agency and you're solving your problems. Again, an agency can be amazing. They can help you. Like we work with outbound agencies all the time and there are some of them that are great, but at the same time, you need to know what you want. You are the customer, but you're also, you need to put yourself in the builder seat and instead of, hey, what did I wish I had the time to build? You are driving the experiments and they're there to assist you and give you the expertise on like, hey, if you want to run this, this is how you do it. So.

Love it.

Yeah.

Yesterday we actually kicked off the cohort 3 of GTM Engineer School and we were chatting about in our intro session about two main components of the new modern operator. It's not just about being AI-enabled, but it's also about having subject matter expertise and taste. Like, you need to not just know what— like, how to use AI or harness it to achieve your outcomes, but you need to tell it what to do. So you need to give it the right foundations of context, like, you know, what you mentioned about having the PMM hat on, but also having the RevOps hat on, you know, like even if you need to, like you need to know how to structure a CRM, you need to know how to, you know, which enrichments to, which fields to enrich and so on and so forth, right?

So yeah, and I think, I think like if you look back at the last 10 years of marketing, like marketing was more of an art and people were trying to make it a science. Now then making it a science is basically free. You know, you have way more, you know, scientists that are like don't have have not gone to art school, like the way I think about it, you know. And so I think brand and messaging and again, like, just like the product you build are essential, even more so in the AI slop era, you know.

So there's a comeback of brand and product marketing. Awesome. And you mentioned before that, you know, like now you like what you're hyped about GTM engineering is that now all your team members can own end-to-end workflows, right? So does that mean that, you know, GTM engineering has literally unlocked the capabilities for one person to do the work of 10? Or do you reckon that's the wrong framing?

I think it's the wrong framing. I think that can definitely be, again, like depending on the size of your company, like absolutely, like you can have a person running an army of agents that can do multiple things. But I think if you look at like the fastest growing companies in the world, like, and I think of two, like, you know, Anthropic and OpenAI, They're hiring for two profiles. They're basically telling you, you don't need engineers. And yet they're hiring like for a ton of engineers. And they tell you that you don't need go-to-market people because AI can do it. And they're hiring a ton of account executives. And so I think the way I think about GTM engineering and sort of like this workflow is that they're a force multiplier. It's more like, hey, can we double the output of a person? Can we 3x what that person's output expected was, expected output was? And I think if you think about it that way, you're building in a scalable way. Because what is the key man risk then? Like if you have one person really doing the job of 10, sure, that person might be making half a million dollars, a million dollars a year, and that's great. But like if that person leaves, you basically have, like it's hard for you to just repeat that motion. And so the way I think about it is like you still need to build a team. You still need to have different expertise and diversity of thought within your go-to-market team. But I think again, like everybody can just do more. And so the expectations are higher. Sure, you might need 5 people instead of 15, but I think of it as like, we should be, it's something that we should be democratizing. Having been in the developer tools and data space, it's something that I've seen like, sort of like maybe 5 years ago when there was a problem, like everybody was hiring data scientists, but data scientists like live on their notebook. And so like, if you know anything about data science, they have these notebooks that are run locally, like Jupyter and whatnot. And those are a ton of just like, files that are sitting on their laptops. And so when data scientists used to leave a company, it was a big mess because they were leaving nothing behind. I think the same thing is true for go-to-market engineers because a lot of what they do is in their Claude chat or in the ChatGPT chat. And so now what we're doing is actually going back to pure developer workflows where my team is going to hate me for it, but we're going to write a ton of markdown files in a GitHub repo and everybody's just going to work from a central place where Everything that you do as knowledge should be stored and accessible for anybody else that's going to join or leave the company.

So yeah, I think that's the way, having a GitHub repo is the way of actually sharing, making everything in your cloud code chats shareable and enforceable. Right.

But I also think it's because if I'm like, what I notice is that when I'm giving feedback to my team, for instance, And it's not because they're amazing, but it's not because like, you know, like they're not great people or anything. They're insanely awesome. Shout out to my team. They're awesome. But I end up giving similar feedback a lot. Maybe it's because I'm lazy or maybe because it's just like things that I see and I had experience on. And so like if I'm reviewing a landing page at a certain point after like 10 rounds of revision, I'm just gonna create a skill for, you know, for reviewing that landing page and then I'm putting it somewhere where my team can access it. And the next time they ask me to review a landing page, I guarantee that there's gonna be like maybe 50, 60, 70% less for me to look at because a lot of the questions that I would already have asked were already addressed.

Big time. And you touched upon, I mean, we've seen actually bigger and bigger companies now adopting GTM engineering job description or hiring for these type of roles. In your experience, I mean, like, you're probably in the mid-market, you know, like segment with a pretty lean team as far as I remember. But which kind of companies do you think benefit the most from having a GTM engineer on? And which actually are the ones that probably shouldn't, you know, like touching it?

I think I kind of mentioned it before. I think like companies with that are post product market fit are probably going to benefit the most. And the reason being is that I think there is two schools of GTM engineering, right? There is like, kind of like you're creating signal, which is like you're in clay, you're trying to figure out like, hey, like, let's do deep research on these companies and figure out like what, why, like what are some compelling things that they're doing, they're hiring people, they're doing all these things, and so on and so forth. But when you are post product market fit, you kind of know why people are buying, like, or at least you have some ideas. And so your experimentation is just much faster, you know, because yes, you're still gonna go and create signal and try to figure out Hey, like if I'm selling into legal companies, maybe like every time they update their privacy policy or something on their— I'm just like random, right? Like every time they update a page on their website, like I can infer that they're now doing XYZ. So that's amazing. And you can run a lot of hypotheses and I think GTM engineers can really help, you know, like there, but there's nothing like a company with product market fit and every existing team and existing customers where you're just going to have a 10x impact. So. And I also think good market engineers is just like not something you do in isolation. So for me, it's like, yes, you can do outbound, you can do outbound at scale. But like if you were doing outbound and at the same time, the audience you're targeting is also receiving your CEO's thought leadership posts as ads and so on and so forth, everything just works better. So like I think the other thing is good market— having a GTM engineer doesn't mean that you can drive your budget to zero.

Not at all, if anything. Right. We also like mentioned already skills, you know, we mentioned GitHub repos, we have to talk about Claude Code or OpenClaw or everything in between. And we'd love to hear your take about whether, you know, is it actually worth moving into general purpose IDs like Cloud Code Cursor or from other go-to-market engineering tools, you know, like we're seeing every day, you know, like I rebuilt Clay into Cloud Code, I rebuilt Instant Lean Cloud Code, blah, blah, blah, you know, but what's your take on this?

I think it's exciting. And I think like, I'm also, you know, guilty of having done some of this stuff. But whenever, you know, after a while, you look at it, and you're like, okay, like, what can I really— what do I really need to build versus what What should I just buy? And you know, build, again, developer tools, build versus buy has always been the dilemma. It was like when you ask an engineer to buy an API product, they're like, ah, I can probably build this myself. What you don't realize though is like, what is your core mission? And I think yes, with AI, your core mission, it's always bigger and bigger. But at the same time is, you know, working with 500, 600+ connectors at Airbyte, we know that APIs are changing all the time. Okay, you're building something, you're super excited. The vendor you're using is gonna update their API. Who's gonna update that? Who's gonna fix that? Whenever you're building an SEP for AISE and all of a sudden your emails don't send in the middle of the night, like what are you gonna do? Are you gonna run the campaign or are you gonna go and fix the tool? So I think like companies that are like slightly larger, like I think about companies like Vercel, for instance, like they're building their own tooling or RAMP. They basically build their own, you know, common room or their own like marketing brain as I call it. And that's great. They have a team of engineers, they have thousands of people, they have a team of dedicated people that is building and maintaining that. Where I get more excited is actually about stitching together solutions that like couldn't work together before. So for instance, now if I wanna connect instantly to, I don't know, Commonroom, I can probably just use an MCP server and just like have them communicate. So before we were always at the mercy of vendors and their integration roadmap, and now you can just make anything work with anything. And I think that that's kind of like the exciting part is just like going faster at experimenting and saying, hey, I need these 3 tools to talk to each other. Great, you can just do it. Or, you know, like stuff like Clay is now selling their enterprise plan, which I think is great. But at the same time, you can achieve pretty much the same thing with a Supabase backend and stitching together a few tables. And I think that that's, it's more sort of like what I'm excited about, which is like augmenting the capabilities of existing tools and yes, consolidating a bunch. Versus building that new. This doesn't mean that we're not building that new. I think when it comes to stuff like just purpose-built agents that run on your company context, I think that that's something where we're experimenting more with building. But again, like we're building an agent that is powered by our context, but there's always going to be something like Octave running in the back and so on and so forth. So it's not about replacing technology, but again, like it's making technology work for your workflows. And that's what I think I'm very excited about, like building. When it comes to cloud code and these IDEs, I think again, it's just what they bring to the table is like the discipline of documenting everything. I think there is this, you know, in San Francisco especially, we're going back to the office, everybody's excited. You know, some people hate it, some people love it. We're back in the office 5 days a week. What you lose, what you gain is like speed. You go way faster. Like there is no doubt. You gain like social interaction. I think, you know, we've all been missing it, but what you lose is a written culture. When you're remote, everybody's written down. And I think that working out of IDEs and working out of this sort of software engineering type workflows brings you back to a written culture. What I was saying before, everything is written down. You have Markdown files, you build your repos. And so again, you're just better off because everything is encoded and you have Git, which means that you can branch out and go back and forth. If you make changes that don't work, it's easy to revert back.

That's a fascinating take. I hadn't heard that about going back and losing the written culture.

Yeah, Matteo knows about this. Like, he's, uh, he writes everything, you know. And I think, like, between me— between, you know, we've been friends for a number of years, and I think that I've always envied this ability of Matteo being super organized and super, you know, writing everything down. And I'm like, ah, like, I have notes everywhere, you know. And so now it's, it's amazing. Like, one of the things that for instance, that I did is because I was a GitBook, I'm getting back at using that tool. It's basically a documentation tool and it syncs with GitHub. And now I basically, my workflow is that I write notes every day and then I have a cron job that in the evening just cleans the notes from my day before. And in the morning I wake up and my notes are clean. And by the way, like I have a CloudMD file, knows my priorities, knows what I'm working on. And so you say, hey, like you've heard these 10 notes yesterday, I think these are the 3 things you should be focusing on today. And so I think those are things that were just not possible before. And it's amazing, you know, just having your own assistant like 24/7. Wow.

Yeah, we really are becoming cyborgs. Mario, your stack includes what? Commonroom and Octave and Clay?

Yeah, Commonroom, Octave, Clay, Instantly. We use Vector as another amazing tool for pumping audiences. I'm always a big fan of like, hey, if my BDR or my AE are reaching out to a person, like I think they've probably done their research job. And so like I wanna show them ads. And so like that's pretty much what's in our stack. And then of course we have Gong and yeah, like we use Cursor or Claude and so on and so forth.

It may be obvious to some of our listeners, but for others this may be useful. What can you, could you just kind of explain the job of each tool and why you select those and there's sort of, There's tool sprawl. So like, how did you get it down to a core set? And then also like, where does Airbyte come into play? Do you use your own product in your GTM infrastructure? I imagine so.

Yeah, for sure. So in terms of, you know, I started with Common Room is our signal hub. We have signals everywhere from our Slack community to GitHub to, you know, product usage data to like website visits and so on and so forth. And so like everything gets organized in Common Room in terms of, you know, our signals. I think where Clay comes in, and I think people think that the tools are interchangeable, I actually love using them together, is that you can actually run custom research and deep research. So the idea is that like, okay, if a contact or a company makes a segment, it's just a way of coming around to say like, hey, these contacts or company has hit a signal that we care about, then we can go and enrich those like in Clay even further and unlock custom workflows. And so, and then you have, you know, those are are the main two things for orchestration and signals. And then we have Octave, which is basically the center of our stack when it comes to outbound specifically. And the reason being is that I think AI is way better at writing emails than humans are. And we've just been, you know, we've just been scared, scarred with all of these AI SDR companies that are just looking at your LinkedIn and saying, hey, I saw you went to a funeral last week. We should talk about B2B SaaS. And I'm like, that doesn't work, you know? And so for me, it's having something like Octave. And for the people who don't know Octave, it's basically, I call it like your marketing brain, you in that sense. And it's like, that's all the— it's a platform that has all the context, your competitors, your playbook, your personas. And then there's a couple of things like qualifying agents that can go and qualify and score a company based on whether it's ICP and then go and write these super personalized and amazing sequences. And so we orchestrate that into Clay and then we use Instantly primarily for all of our outbound. We still use Outreach. I think Outreach still has a place, especially with AEs and more manual type workflows. But that's kind of like the main outbound stack. And then of course, as I said, we use Vector to orchestrate and sync all of the audiences to our ads platforms. In terms of Airbyte, we also use Airbyte heavily. It's basically like, that's how we get all of our product usage data into Salesforce and we can just dump any source dump any source into any destination. And it's super, it's nice, you know, to be able to rely on that. You know, a ton of companies have to spend time figuring out their data strategy. For us, it's what we do for, you know, for a living. So like, it's having clean data, it's super important.

So awesome, great explanation, Mario. Can you think about or walk us through one warm outbound motion that comes to mind during the time which, you know, you've had such a big impact at Airbyte and doubling pipeline? Like, what signals were unique that you're actually triangulating around? And how do you go from data that everybody can get to the really granular Airbyte-specific data and gain confidence that it's going to land in your outbound sequences?

Yeah, I think I can think of two. Like the first one is we still spend a ton of money on events, and I think events are actually a great way, you know, of generating pipeline. But then you get home from the event and you have thousands of leads, which is amazing, but like not all of them are qualified. And so we can just upload those leads into something like Commonroom and then rank them by, hey, have they ever signed up for Airbyte? Have they ever used our open source product? And so on and so forth. And so like then you combine the amazing works that our marketing and sales team do on the field with like our intelligence on like which contacts and companies we should prioritize. And so when we then like, you know, give that back to SDRs, What we see is a higher conversion rate to meetings because the people who are calling, like, already know about Airbyte somehow. And of course, like, they came to the booth and they remember that interaction. But it's one thing to call a company that has been actively trying your product or interacting with your open source repo and somebody that just says like, hey, yeah, this sounds cool. Like, I would love to have a call. And so that's one workflow. And I think, again, the other workflow is our PLG motion. You know, we have thousands of signups every month, usually It's, you know, if you believe that only 5% of your ICP is in market, I think that people signing up and trying your product is probably the highest signal of that. When it comes to engineering and, you know, developer tools, usually the engineer signing up is not the person that is going to be ultimately the decision maker. It might be in smaller companies. When it comes to larger companies, they're usually the people testing a bunch of solutions and reporting back in that sense. And so what we're able to do with that is basically say, okay, we know that this person signed up, we're going to enrich them, we're going to outreach to them and see if they need anything, if we can be helpful. But at the same time, we can go and prospect for decision makers in the organizations. And maybe instead of outreaching to them straight away, we can decide to show them ads to warm them and so on and so forth. I think the Twilio Ask Your Developer campaign, like that's what you can now do at scale and it's awesome.

So you're really leaning into where there's already signal using these workflows. And that's sort of how you avoid the VP feeling sort of just cold called out of the blue is it sounds like you're not really doing that that much. There's a signal there.

Sometimes we have to. We launched a new product in September last year and it was a completely new audience for us. It's the same product like Airbyte, but it was a completely different target market and we had like little to no signal. And so we had to go and build a cold motion. I think that's unavoidable. I think though, what you learn is that you can make it warmer. It's like, hey, can we, even though it's cold, if you believe in warm outbound or warm go-to-market motion, it's like, how do we take this cold motion and warm it up a bit? And so I think it's learning and having those skills is super useful for that. People call me crazy for wanting to do ABM at a Series B, like early Series B company with a small team. But I think that that's what's easier than ever today. And if, again, if you know your audience, you know your personas and you do your job right, that can be really powerful. So.

Very cool.

Let's switch gears and talk a bit about hiring and teams, which I know you have a lot of experience on and you've, again, I guess like the results also like speak for themselves, like in the last, in the last year. But we were also like witnessing like a job crisis or, you know, like massive changes with obviously AI and everyone's kind of like trying to expect it to do everything. You also mentioned it. The bar has never been higher. Technology is not a blocker now anymore. So if we take a step back and take all of this into account, all of these new changes, How has that influenced your team structure at Airbyte when it comes to your GTM org?

I think as a, like, that's what I've learned as a VC, you know, like I probably talked to thousands of founders and your job is basically, when you're a VC, your job is basically between being an AE and a recruiter essentially. So like you're looking for talent because like that's a number, especially in early stage investing, that's the highest signal. Again, going back to signals is like, Who are these people and why should I bet on them? And then at the same time, you're an A because you're trying to sell your product, which is capital in that sense. And so if I go back to that, one of the frameworks that has been super helpful for me is something that Keith Rabois, investor at Founders Fund and now at Khosla, investor in some of the best companies in the world, I think he's actually one of the people that has the best eye for talent, is this concept of barrels versus ammunition. And a barrel is essentially a person that can help you take something and start a project and just finish it. You know, they were gonna figure it out. They're not experts, or they're not like, they're essentially generalists that are gonna just figure stuff out. And then Ammunitions are more like the people that have some expertise. And, you know, they're really good at one specific thing. And I've always been in the camp of hiring barrels or hiring generalists, because I am one of them. And so for me, it's really about drive over experience, agency over experience. Like, I don't care if you've never done this job, like, The best SDRs I hired was a person that was selling pest control products like door-to-door. They had never opened Outreach. They never opened any tools in their life. Last year they were doing like 2, 3x the pipeline of anybody else in the field. And so for me, it's always about, especially with AI, like I think it's finding people that can find and come up with creative solutions to problems. And those are the type of people that I look at bringing on board. Do you still need expertise? Yes, 100%. I'm not saying no. Like my go-to-market engineers should have an idea of like how data models work. And at the same time, my paid person should know that $150 CPM on LinkedIn is too much. But there's never been a better place for generalists within their own fields. You can be a paid media expert and know everything about bidding, or you can be a paid and SEO marketing person that knows everything about paid and SEO, and you can just build a lot of workflows. And so that's kind of like how I think about hiring and bringing people on board. So, so yeah.

Love it. And how is actually your GTM org has reflected that? Like which kind of like roles are there and how was that before you joined? Actually, I'm super curious about that.

Yeah, marketing team was actually pretty small there, but I have 3 orgs reporting into me. So first one is growth and marketing team. The second one is our dev rel organization, which we're rebuilding as we speak. And then third is our BDR team, SDR teams. Like we could start the should SDR report into marketing or to sales debate. I'm in marketing camp and it's been great for us. But like the way I think about, you know, we have a pretty traditional marketing team in a way. I look for what are the channels that we know work really well for us. And so I go and outsource those, like in terms of I go and hire people that can manage and scale those because we feel comfortable. And everything that is like unknown, or, you know, we're trying to figure out is a cross-functional effort of everybody chipping in and figuring things out. So in terms of like the specific roles in marketing team, it's, we have an events person. Again, we do a lot of events and those are great. I think especially after, you know, a few years of being at home, like people are craving that in-person interaction. We have a, as I said, a paid and SEO person. They're, you know, focused on like anything that is like content programmatic and paid ads and channels. We have a GTM engineer. And then we have basically like a PMM that is way more than a PMM because they're not doing just product marketing, but also running some founder-led outbound workflows and stuff like that. And then I have a lifecycle marketer because PLG, so a ton of, you know, again, signal and inbound and so on and so forth. And then on the sort of like BDR side of the house, we have SDRs and we have product advocates, which are basically very technical, junior people that can help prospects, especially like when you talk to engineers, they can help them debug and figure stuff out, which is great. And then as I said, we have that REL team that we're now rebuilding. So those are the— that's kind of what my work looks like.

Awesome. And you mentioned about barrels and ammunition. I think like it's fascinating because like it, in my mind, it linked back to like the previous metaphor that we talked about which is build versus buy, you know, you want to optimize for folks that have agency and, you know, like soft skills, but ultimately have— are generalists at heart. But then on the other hand, you also need for some stuff, some more subject matter experts in a specific field. How does that translate into your approach when it comes to actually leaning into agencies or consultants versus hiring full-time, which again, I think like it's build versus buy in a way, right?

Yeah. And that's kind of where I was going to as well. It's when you started asking the question is like we complement our team with agencies. Like I don't need my team to be an expert in deliverability. We work with an agency that helps us with that and helps us with maintaining like those go-to-market workflows. Like, so we work with an just give an example of outbound. We work with an agency that helps us with saying, okay, we figured this campaign works, let's go and scale it. And then we have hundreds of inboxes, like how do we manage them and so on and so forth. And that's great. In the same way, we work with a paid media agency because I don't want my team to spend time figuring out bidding and all of that stuff. Although now there are agents that are starting to pop around, which is interesting. But I think what that type of agency gives you is that they see what works and what doesn't work across a ton of clients. And so they, you just go much faster. Same thing goes for, you know, SEO and content and so on and so forth. So I think like in a way, like one of the things that, you know, I was reading, you know, if you look at X, like some of the best companies are telling you that their engineers have not written a line of code, you know, since December and so on and so forth. I think for marketing, it's kind of like the same thing. I want people that are going to be managing agents and workflows and other people versus actually like doing just like execution for the sake of it. And so for me, it's like, how do I get the most leverage from my team, enabling them to focus on what matters versus focusing on what is not going to drive your organization forward? What I'm saying is those partners are amazing. And so like, that's why we rely on them. We don't need that expertise in-house. So very true.

And when you've been hiring your GTM engineers, or when you approach like this type of like hiring, problems. Like, uh, what's been in your experience, like, uh, the, like, the stuff that separates the good from the great?

Um, I think for me is that the why, you know, is why you're doing something, is again in the same way, like, if you think about a few years ago, like, everybody wanted to work at startups until they saw that startups were super hard. And so, like, you had a ton of people that started working in startups, maybe, like, coming from all of, like, consulting or investment banking and so on and so forth. And then they realize, oh my God, this is the hardest thing I've ever done in my life. And then they quit. And it's the same thing with go-to-market engineering is like, do you want to do the job because it's the coolest job title? And I'm, you know, if I wasn't doing what I'm doing now, I would probably, and I told my go-to-market engineer this, it's like, if I didn't have to manage all of you guys and I have a ton of fun doing it and, you know, building the go-to-market motion, I would go and become a go-to-market engineer. That's, it's awesome. But again, like at the same time, it's like, why are you doing it? And I think that In a way, like GTM engineers are this sort of like misfits that have always been seen as like, you're not really a marketer, but you're not really an engineer. And finally, AI is giving them that power to shine because they understand both motions. And so for me, what's great is, you know, good is somebody that wants to do the job, they're excited, they understand the tools, that's okay. But somebody that understands engineering workflows and how to run them and data and also understands like, why are we trying to do this from a go-to-market perspective and understanding business, I think it's just a great combo.

You know, when you mentioned about the misfits part, like it brought me back to when I used to, we both used to wear our growth hat on. Do you think like it's just another, you know, like version of that, you know, like sitting in between because it's just like cross-functional operator and stuff?

Totally. And I think like, especially, you know, this is true for a lot of PLG companies, like you always have this growth PM where they're not really a PM, but they're not really a marketer at the same time. And if they're lucky, they get a few engineers, you know, full stack or front end to just play with them and change, you know, onboardings and do A/B testing and so on and so forth. And so I think it's been, as I said, it's been true forever. It's just that the possibilities of what you can do with AI have just elevated these type of jobs like 10x. I think like even our PMs at Airbyte are now just like spending time in Lovable, you know, you go from, PM that is looking at a roadmap, talking to the customers, you know, asking the design team to put together front-end designs, and then you go and review those, and then you do the customer sessions and so on and so forth. They can now be live on a call with a customer and just like prompt Lovable to change an onboarding flow and ask the customer, is this better? You know, and get that feedback live. So as I said, I think everything is compressed to a point where you're looking at that real-time interaction that it's only possible with AI.

Love it. So now we've seen that again, it's not just about GTM engineering. There's context engineering, there's vibe coding, there's content engineering, you know, and we've seen, you know, like there's a, it's more of like, it's not just like micro trends, but they're all part of like a bigger macro trends. Obviously, I think like AI enablement is this new Late Motif. Like, what do you think, you know, like you'd recommend to a person that would want to work with you in your team? What are the highest leverage skills among these new ones that are coming out that you recommend folks to invest in? In 2023?

Yeah, I think it comes down to 3 things. Like, as I said before, number 1 is agency. It's like, are you going to do the thing? Or are you telling me that you're going to go and do the thing and, you know, build a plan for it? Just do the thing. I think I said the second is curiosity, like, because it's been true more than ever. It's like, yes, you can go and you have an idea of a playbook that works. When I joined Airbyte, I had to reinvent all of the playbooks. Like I was like, oh, I'm going to do this thing that I used to do 5 years ago. Guess what? It doesn't work. So curiosity and really like having the pulse of what's going on. And I would say the third thing is taste, because again, AI, you know, brings the cost of writing copy, writing code and everything to zero. What is going to differentiate is your taste and understanding of a deep understanding of who you're selling to. You're marketing to and why they should buy. So like I would say like those 3 things are the most, you know, and then there is a 4th thing, which is usually like, I like people that have a chip on their shoulder and have something to prove. Like that's the profile that I usually work best with. And it doesn't mean that it needs to be like, you know, people that are coming from different backgrounds and so on and so forth. Like I've seen people whose parents were super successful, you know, CEOs and they just wanted to show them they were as good. And then I seen people, you know, immigrants like me or you, Matteo, that are just like leaving everything behind and following like a dream of going and building something. So I think it can come from everywhere. It's just like, why are you so driven to succeed and why can you not fail? Those are the type of questions.

Love it. So if I have to play back, agency, curiosity, taste, and having a chip on your shoulder.

Yeah. If you find a person that has all of those, just let me know because I'm hiring. I'm always hiring. So bring it on.

That's a good list, Mario. We are coming close to the end of this episode. I got one more question for you. If you were giving advice to a VP of marketing or head of revenue today, what is the one thing that they need to get right? Or what's the one thing they should stop doing? What's the biggest piece of advice you have from your time in the seat?

I think the biggest thing you should stop doing is just doomscrolling social media and like sending like, oh, sending your team everything that other people are doing is like, oh, we should be doing this. Why are we not doing that? And like, that's the number one thing. So that's the easy part. I think something that they should start doing is listening more. You know, it's like in a way, like a few years ago, like again, it was a trend. Like everybody was talking about listening to Gong calls when you were working out and going to the gym and so on and so forth. And that's great. I think today is like, have you built yourself, like people talk about agents, like have you built yourself a workflow that summarizes these like yesterday's gone calls and send you a digest on Slack? You know, it's like just start listening to what your customers are saying because I think like they're the best people at telling you how to sell to them, you know? And so like I think that that's the number one advice.

That's fantastic. I really enjoyed this conversation, Mario. My takeaway, let's see here, I think today You know, Mario explained that, that go-to-market engineering, it's a real step change, but only when the foundations are solid. Like, I think you mentioned clear product market fit, having like sharp personas, a strong product marketing group, and then, well, I mean, clean RevOps data. Yeah, those are some of the foundations. And, and AI and modern tooling, they do let every marketer, SDR, own their end-to-end workflows, which you said could mean that people can run 4 or 5 experiments where they maybe used to just be able to run one.

Yeah.

So this actually acts as a force multiplier and not a replacement for teams or critically for strategy. Mario emphasized also that warm outbound is built on real, real signals. In your case, product usage, GitHub activity, events, you mentioned very important community, all this intent data, rather than if you can help it, just blasting cold lists, which you have to do sometimes. This is how Airbyte really doubled pipeline, digging into real signals. And finally, Mario also argues that the real advantage now is stitching together best-in-class tools like the Common Room, Clay, Octave, etc. Instead of rebuilding everything from scratch, which you could do, and that's the temptation, and some companies will do it, but encoding that go-to-market knowledge in developer-style systems like Git so it's reusable and it's scalable over time. So this is, this is fascinating, very insightful. And Mario, where can our listeners connect with you?

I'm on LinkedIn, you know, like just keep it simple. So that would be like the way to connect with me.

Awesome. Well, thank you so much. This is fantastic.

Thank you, team. Appreciate you so much, Matteo. Had so much fun.

All right, take care.