Stop Using Sh*tty Data. You CAN'T Achieve Autonomous Revenue Without This - with Elio Narciso

Welcome back to another episode of the Bridge the Gap podcast powered by Revenue Reimagine. Today's guest is Elio Niso, the co-founder and CEO of Scalac, where he's leading a team building what he calls an autonomous revenue engine. It's a system that doesn't just automate GTM work. It actually eliminates the need for most of it altogether. He's run GTM for startups at AWS, sold companies, advised high growth teams, and now he's focused on killing the manual work tax that slows revenue teams down. This isn't going to be a fluff episode. We're going to talk RevOps, AI, what GTM could look like if we actually got out of our own damn way. Elio, thanks for joining the show.

Thank you, Adam. Great intro. Nice to meet you both, and happy to be here.

Awesome. Elio, thanks for joining. I love that the first part of the intro that really sparked my interest was the manual work tax and what that is. So you've called the current state of RevOps a manual work tax. What does that look like in most companies today?

I think it applies not just to RevOps but to many go-to-market teams. The inspiration for this was born out of my time at AWS. Which, you know, I joined after a couple of startups I created. Before joining AWS, I thought, okay, I'm going to join like one of the best companies in the world. They're going to have everything figured out. Every system, every tool is going to work perfectly, and data will flow magically. And then it wasn't the case. We had the same problems. Obviously, there were certain things that they were doing extremely well, but also a lot of the things that I had experienced in my own companies were similar.

For instance, I ran a program to support the go-to-market for middle-stage startups. They were customers of AWS, and we were supporting them in their go-to-market. Now, in order to decide which startups would deserve more support, more resources, more AWS credits, marketing, etc., we basically had to prioritize. To prioritize, we had to look at a huge list of startups and decide, based on a number of criteria, which were worthy of our attention and support.

I tried to do this in Salesforce, which is AWS's CRM. The data in the CRM was terrible and outdated. So basically what I started doing is a spreadsheet. I put together a spreadsheet manually with some Zapier connectors to Crunchbase and to LinkedIn and to ZoomInfo in order to map the market and understand what startups would be worthy of that support and additional marketing and sales co-selling support by AWS.

The list became huge—tens of thousands of companies. Imagine startups created every year, and I had to prioritize them and try to make sense of this data. It became like a huge mess. And that was the point where, for me, you know, the idea that this is not just me at my startup or growing companies but this is also me at AWS, one of the best companies in the world, then this must be true for a lot of other people.

And so it's not just RevOps. It's reps that need to do research on their prospects, or marketing teams that need to make sense of leads and signups and decide which one should be sent to sales. Because if you send them all to sales unenriched or unprioritized, sales is not going to work on them. Or RevOps teams who are tasked with making sense of all of this data and build sort of a foundational model.

And so that's what the manual work tax is—the time that I had to spend building that spreadsheet, the time that the rep needs to spend researching stuff on Sales Navigator, which is not the easiest tool to use. You can spend hours down the rabbit hole.

It's easy.

Yeah, down the rabbit hole. It's a Microsoft product, y'all.

Yeah, it's funny you say this, Elio, because I worked at Oracle and I had that same aha moment where it's like things should be working much more efficiently or effectively, and they own their own CRM. They own it. It was just—it's the same thing. Why haven't all the RevOps tools actually fixed this before? This problem should have already been fixed.

One hundred percent. I think that the CRM industrial complex should have fixed this because, you know, like we call the CRM the source of truth. But if the source of truth is full of outdated or bad data, then you know what source of truth does that represent?

So I do think that the CRM should have solved this. They haven't. The problem is complex, and I think that AI presents both the catalyst for understanding that the old model—where, you know, you laugh when I said the data in the CRM was bad—you know, this is true for every company. And until very recently, there isn't much that you could do about it.

Now with AI, your competitor will start getting better data in their CRM and will start prioritizing their sales and marketing efforts better, and it will become a competitive advantage. So in the age of AI, data will impact outcomes disproportionately. If your data is bad, then even if you feed it to the best AI, the results are going to be bad.

And so AI represents a catalyst for change. But also, the reason why we can now work on that data is because the Scalac platform would not have been possible four years ago. Yes, we could have done automation and workflows and all of that, but without the agentic component. We have agents that go into the CRM, decide what data source is best for any data point, decide when they have enough confidence level about that data point and when they need to release other agents to do research to complement and improve the confidence level about that data point. All of this was not possible until very recently, and now it is.

AI represents, as I said, also the catalyst for change.

I love that. So it is a big catalyst for change. But when you look at specifically why the CRMs haven't fixed this and where AI is or isn't being used, is this more of a tech problem or is this more of a leadership problem?

So I've done a number of roundtables with CROs over the past six months, and I have to say that it is not always clear to the CRO what the role of RevOps should be. RevOps is really important, and it's the core of systems and processes and technologies, but also it's a very strategic component of the go-to-market machinery because it owns the data model. So I've heard CROs talk about their RevOps as handling systems or processes or handling tickets, and that's the wrong attitude.

You want RevOps teams to be strategic. You want to rely on them for the data that will be used to prioritize sales and marketing efforts. So CROs know that they have to distribute an equitable book of business across sales, right? They know that and they do it, but they do it very manually. They do it once a year, and everybody's rushing to say, okay, how do we distribute the accounts?

In order to do it well, you should rely on your RevOps teams that are empowered with the tools, systems, and data to say, okay, this is what this account is worth. And so the way I patch the territories or distribute the book depends on what the value of each account is. And so I can distribute that equitably.

I think there is a leadership problem in that sense—sometimes there is a misconception of what the RevOps role is. But then there is also a technology gap. Until very recently, essentially there was plenty of data out there—ZoomInfo, Crunchbase, LinkedIn—but the key is how do I align that data to my ideal customer profile and to my go-to-market strategy, which will change for each of us.

One hundred percent. And that ties exactly into where I want to take this for a moment. So we talk a lot when it comes to AI and go-to-market about automation versus orchestration. I think a lot of folks just think, oh, we're going to automate it, we're going to automate it. And we say all the time—I think Dale posted about it this week, if not two or three times—automating is just going to get you more of what's wrong, right?

If your ICP is wrong and you automate outreach, you're reaching out to the wrong people. If your buyer persona is wrong, it's the wrong people. If the RevOps data is wrong, you're automating bad data.

Where do you come in when you look at what you're building specifically? Where do you automate versus where do you orchestrate?

So we believe that the orchestration should be automated. Tell me more.

I think that the instinct in the go-to-market space has been, oh, let's automate sending emails. Let's automate, you know, let's say the rush to create AI SDRs, right? So to automate a lot of the last mile elements.

And frankly, a lot of the things that humans should be doing—like a good email or like a good outreach or a good engagement—should be, and it is, like what humans know how to do. And instead, we haven't devoted enough attention on automating all of that like frankly boring work that goes into cleaning your CRM data, removing duplicated accounts or leads, or like building better hierarchies between accounts, or assigning contacts to the right accounts, and then calculating the time in each account or like deanonymizing leads, et cetera, et cetera. All of that frankly is extremely time consuming, repetitive, boring work that I don't really want people to do. And so to me, if we can achieve—and that's why we call Scale Select Scale Stack an autonomous revenue engine—because we want to automate all of that, which is an orchestration. And so we have agents that compose and then run workflows to achieve like that cleaning, that enrichment, that prioritization of the data across the CRM.

Are there tasks that should never ever require human intervention at all? I mean, a lot of this stuff I think that can and should be delegated to AI. There was an artist—I forget her name unfortunately, recently like a few months ago—who became viral because she said "I want AI to do my laundry, not to write my poetry." And that's sort of true. So if I push onto AI all of the stuff that people frankly don't want to do and find boring, I think that's a good place to accelerate a lot of stuff and relieve people. Like, RevOps—if a RevOps team needs to spend like three months to clean and dedup the data, I'm going through this with a client right now. It's driving me nuts, right? And so at the end of that process, you're not going to have like the same energy that you had when you started. And so maybe to the most strategic, last-mile component, you're going to devote less energy.

Instead, what we aim and what our customers achieve is that they input the business logic and the criteria that they want to achieve or the use cases that they need to solve. We absorb all of that complexity with the platform, and then they—meaning the customers—monitor the outcomes. Is the data hygiene increased? Do we have like a good hierarchy? Have we removed all of the accounts that are duplicates, or like have we reassigned contacts to the right account and then cleaned all of the contacts of like bad data and updated and added like new data only when the foundational data model was done right?

So we think that all of this can be done by agents. Right now the platform is heavily run by agents but supported still by like humans—ops people that run the workflow and make sure that everything is running smoothly. I see a future, like you know, around the corner literally, where like our workflows will be composed and run autonomously by agents.

Is there a place to make sure automation doesn't create too much noise? Are we generating too many noise pieces? And along with that, as you were just saying, people still need to like strategically think on the input to put into these systems, because I think that's where people are like "I'm just going to download something and we're just going to run it." They haven't done the strategic thinking part of it. But maybe they combine together, where you may be getting noise if you don't do the upfront work, right?

So I think that until now, the instinct has always been "oh, our CRM data is terrible, and so let me add more data," which adds noise. Let me create a new list. Let me add like new prospects. Let me add like new companies. Let me add more data and signals and intent and all of that stuff, and that creates a lot of noise.

Our stance is that, like, we work with larger go-to-market teams. We work with companies that have at least like 15, 20 reps, and companies that have a few years in market already—way beyond like product-market fit. So they have like already an established presence and go-to-market in the market. At that stage, if you don't figure out like the prior data first, just adding more data will add more noise in our opinion.

And so first, figure out the data model. Figure out like what worked in the past and how, and what are the best customers and how to prioritize the efforts of the sales and marketing teams going forward. But then once, let's say you've cleaned your data and built that data model, I think we believe—you know, like strongly—that in the new world, the spray and play is gone. And so the best teams are those ones that really know how to prioritize.

And so to prioritize for larger companies, you need to rely on data. And so how do I equip the sales and marketing teams with great data to tell them these are the accounts that you should focus on today, this week, this month—which will dynamically change. It's not something that you only do once a year during the annual sales planning, but will dynamically update itself over the course of the year depending also on your changing like marketing strategies or go-to-market strategies and all of that.

And so it's all about prioritizing. For instance, we just deployed like a fantastic, super complex agentic workflow for one of our customers, MongoDB. And they have a free-tier product called Atlas. Atlas is super popular, and so they let anybody without a credit card register because like they've used it historically as a pipeline. But the way they have used it is that like there are hundreds of thousands of people that register on Atlas—developers. They would not really act on that data until like somebody would raise their hand and say "Hey, I want to talk to sales."

But what we have done for them is deploying like a workflow that enables them to first filter and deanonymize profiles at scale, leveraging data from multiple data sources—internal, external, third-party data, agentic research that is done within the workflow. And only a very small subset of these signups are then like fully enriched with a lot of insights and signals around like the company, the person. If we find value in those signups or leads, and then this becomes a subset that sales can work on, or marketing and sales can collaborate on, or they can do like social media campaigns or ABM campaigns and all of that. And so it's not sending an email to the hundreds of thousands of like Atlas signups, but only to those for which like you have a high degree of confidence that could be an interest. And why people buy from people—that's why companies who invest in meaningful connections win.

The best part: gifting doesn't have to be expensive to drive results. Just thoughtful. Sendoso's intelligent gifting platform is designed to boost personalized engagement throughout the entire sales process. Trust me, I led sales for a Sendoso competitor, and I could tell you no one does gifting better than Sendoso. If you're looking for a proven way to win and retain more customers, visit sendoso.com.

Yeah, that's—um, I really like that flow because there are places where people are getting too much data and they don't know how to act on the data. So like, getting really specific, and this goes right back to what you were talking about earlier, is like the structure and the real thinking upfront. Because you have to do the thinking and the strategy upfront before you build the system, or else you're not going to get the data you want out. And people like, "This AI stuff doesn't work or it's not working the way I, you know," I see all this stuff on YouTube, whatever it is.

Um, so it's very interesting. Let's transition a little bit. You've worked at a bunch of startups. Um, and you've worked at AWS, so you've worked in many different places. Um, when GTM isn't working, what's usually broken behind the scenes? I mean, there is a wide range of options, but I do think—I mean, first of all, it always starts with leadership for sure. And that's why I think that if we talk about like B2B startups for instance, um, why do we say like "oh, the founder should be like you know the key seller until at least like a million dollar in ARR"? So I think that's for many reasons, but one reason is that like you know the leadership, which for a startup is like the founder, needs to develop like that knowledge about the market and the customers and like include that knowledge in motions that become more and more established.

So I take this example because like I do think it's always like from leadership. Like you know, so if like a founder doesn't do a lot of sales or like delegates this too quickly, probably things are not going to work out. Or if in a larger company the leadership is not like setting like a goal or like you know like a vision for what they want to accomplish, probably there it's not going to work out.

I mean, at Amazon, what I appreciated for instance is like one of the great things about Amazon is that it has a writing culture. Amazon doesn't move unless for any product launch or new program or initiative, unless you write a doc. That doesn't need to be a PowerPoint. It's actually a long-form, typically six-page, like word document where like people go crazy in a sense that like you know they debate a lot around that document, but that generates a lot of alignment about what to do.

And I think that that's very important, especially for larger teams, because you always have to have a north star. And I think that this changes every year. So leadership needs to lead by example—number one, the example of the founder—and then they need to set a clear vision, mission, and goal, which we will change during the year also, but for sure every year. But it needs to be clear, understood, and hopefully simple. You cannot have a myriad of goals. You need to have probably one or two goals that are the north star for the action of all of the team.

And then I do think it's all about the data. I really believe that, especially at scale—so for sure at scale, but in any go-to-market situation—once you have good leadership, once you have a good vision and goal, it's all about the data and knowledge about who to target in order to make everyone very efficient. So do we have, for marketing teams, a clear understanding of what's the ICP and what are the buyer personas? Many times companies have a wide range of interpretations and lack of clarity, and maybe somebody in the go-to-market team knows, but the product team doesn't. So there is a lot of misalignment that generates problems. And it hasn't been changed for two years, right? So they're working off of old data.

So to follow up on that question—and I seem to be asking this every time—how often should your RevOps, your go-to-market team, whoever it is, be pressure testing those what we're calling GTM foundations, which is ICP, buying persona, value proposition? Like, what's that pressure test timeframe look like?

I think that with AI, for instance, we have implemented now very simple workflows that anybody can do. So you don't need to buy a skill stack for this, but we record every sales call, we record every customer call, and then we have simple workflows to get the transcript, analyze them, and know what are the pain points, what are the things that resonated, what are the things that didn't. And so this constant learning is so important. And so for sure for a company like us—let's say Seed to Series A stage—it's essential because every three to six months we'll need to rethink this and make sure that we are pressure testing ourselves, that we're focusing on the right companies and people, etc. But I would say for anybody, you have to listen to your customers. And we have so much data now that it's easy to be processed with AI that not to use that data is a sin.

I'm learning. I mean, we just relaunched our website recently. A lot of the content of the website is the byproduct of spending a lot of time on customer calls. We have recorded hundreds of customer calls, distilled them into what they said, what is important, what are the pain points, and then taken them and transformed them into some of the copy of the website. And so I think that's basically something that you have to do all the time, but maybe in a more structured way—like, you know, every quarter.

I love it. I agree. So many people don't. We talk to so many people, and we have a client now—and I joke—they don't record any of their calls. And to me, whether you're using that to create your website or it is a sin in go-to-market to not be recording your calls and understanding what works, what doesn't, what are your customers asking for, what's the product feedback. I literally—I don't want to say nothing shocks me, but this one shot. Like, wait a minute. You've been in business this long. You have ten reps, and you don't record your sales calls?

Well, there are industries like finance or healthcare where that is very difficult.

Yeah. But if it's not, then it is problematic. I mean, but I find this even in consumer. I mean, like, you know, when they send you an email and you reply and it's "do not reply." I find that so crazy. So those same companies that send those emails that have "do not reply" spend millions and millions of dollars in marketing campaigns to target, and then when a customer wants to reply to them, oh, this inbox doesn't get analyzed. It's crazy to me. You want to gather every customer input, and so if they want to reply to the email, let them. Don't say "do not reply," which is crazy. Common sense is only common to those who have it.

So AI, in your view and in our view as well, isn't a feature. I think a lot of people right now are thinking AI is a feature, and I think we all agree that AI is much more of a foundation. As we're halfway through 2025, going into 2026, it's not something that you're bolting on top of your workflow. You're actually building from AI as a core. What's different in the building approach in that sense? And as you all are building, where are you finding AI breaks most in that GTM stack that it's requiring more effort to really get it right versus just like plug-and-play?

I don't know. I haven't heard, like, "oh, it's a feature." I've heard that I think it's a systemic change to the way we work, and I believe that 100%. I think that it will completely transform the way people work, and it is already. I started my career in mobile technology in the early 2000s. I'm Italian originally, so I started in Europe. In Europe, we had mobile phones before the US. This is the only recent technology that Europeans got earlier. Then the iPhone came in, and basically they own—Nokia and Ericsson, all of those European companies that had been formed before then. But it reminds me of that. So the change that there was when mobile phones were introduced, but at a much faster clip. And instead of replacing communication methodologies that were used before—like landlines or fax or other stuff—it's replacing and completely reimagining the way people work and process data and so many more things. So I think it's an incredible opportunity. It has a lot of risks, obviously, but the speed is what's impressing me.

I mean, our platform, you know, let's talk about my little world that I know well—our platform has made incredible progress over the past six months, and every week there's something else that we couldn't do a few weeks before. Oh, this is now finally we can do this, or we can automate that, or we can increase the speed, or even reduce the cost because there's so much competition between all of these big platforms now, much earlier than the competition that started to come from AWS versus Microsoft versus Google. Now all of these platforms are competing head-to-head. Prices are going down, speed is increasing, capabilities are advancing enormously. So I think it's going to radically change everything, and very, very fast.

So the recommendation is that any go-to-market teams embrace this change because otherwise your competitors will. And I do think that RevOps is an interesting area. I see many companies coming to us and say RevOps is going to be the playground for a lot of AI initiatives because it's central to the foundational data model, it's central to go-to-market, it's central to revenues. And so we want to start from there rather than, say, customer service, or rather than say, I don't know, finance, and all of that.

Yeah. I was—um, when I was at the gym this morning, Dale will be shocked that I said that—but when I was at the gym this morning on Squawkbox, the CEO of Lattice was talking, and it was all about the effect that AI is going to have on jobs—not so much about eliminating, but like, if you are not embracing AI, if you are not using AI, if you are not deeply learning AI—for everyone who sits at night and just scrolls through Instagram—take that time and upskill yourself. You are going to be at such a competitive disadvantage. Forget six to twelve months from now. I would argue three months from now. You're going to have major, major problems.

Yeah. Sorry, Dale, I cut you off.

No, no. Um, and I'm curious as we progress, like, where's GTM headed next? So fast forward a couple of years. What's one part of GTM that's going to be completely different—um, than it is today, besides, you know, besides what you're doing within Scale Stack and RevOps—like, where's another place inside of the GTM playbook that's going to be completely different?

Well, I think that we are seeing it at the startup level. Like, the old playbook of go-to-market—I don't know, like, white papers or marketing campaigns or search, you know, paid search, and all of that—is being quickly changed. The CEO is the chief marketer, and some examples came from the larger companies. Think about Mark Zuckerberg. He is the chief marketer of Meta, and, you know, it's not that obvious. It wasn't like that, you know, ten, twenty years ago, right?

# Cleaned Transcript

We have CEOs that are chief marketers, and so they need to embody and represent the company, the brand, and the message. In-person events are super important. Humans keep buying from humans, and I think that will stay the same. We just need to become better at optimizing those experiences.

There's going to be lots of interesting stuff around events and the events industry over the next couple of years because people are craving true connection, better understanding, and understanding the landscape—what's noise, what's important, and who are the people from whom I want to buy.

We're seeing, and this is already probably very advanced, that a lot of B2B SaaS companies are content machines now. Content is king. That's why we're doing a podcast. We have our own podcast. We do clips, interviews, and roundtables. If humans keep exchanging knowledge, we continue to represent what's great about humans, which is connection with other humans that generates ideas and then can get implemented maybe by AI.

For us, I wasn't the chief marketer of my last startup. I was actually pretty quiet. Now we have CEOs building in public, and that's interesting.

I don't think it's appropriate for a company like Scalstack that is more enterprise-focused. I don't want to air my dirty laundry to my enterprise customers. But for other companies, it's very appropriate and generates a lot of attention and interest because there is a lot of noise. How do you raise yourself above the noise? That's the goal of go-to-market. How do you make sure you identify people who will understand your message and feel that it resonates with them and are interested in knowing more?

I agree with you 100%. I don't know how I feel about the whole building in public thing. I get it, but regardless of whether it's enterprise or SMB, I'm not a big fan of airing my dirty laundry to anyone. I'm pretty open on things like LinkedIn, but my customers don't need to hear when Dale and I have a massive disagreement about something, or when one of the AI agents we're building broke for one of our customers and how we had to scramble to fix it. No one needs to know that. All they need to know is it works and we're doing great.

All right, let's go into some rapid fire as we wrap this up. We have some questions that hopefully will spur the brain. Ten words or less. The goal is to get through as many as we can.

What is one GTM tool that you think is massively overrated?

Clay. You're not the first, second, or even tenth person who's told me that. Let me say that we compete with Clay, and I need to thank Clay because without Clay, I think the need for Scalstack would not have been understood. Last year we were fundraising. Clay announced their huge round, and some investors came to us and said, "But then Clay..." Actually, the Clay raise generated more curiosity around what we were doing because we often say we are Clay for the enterprise. I actually like them and thank them for what they've done. They're actually based in New York like me. But on the other hand, I think you cannot be everything for everyone. It's a great tool for startups and smaller companies. I think it's not the best tool for many other companies. Their strategy of increasing awareness through agencies, promoters, and influencers—paid or not—has been a little bit too much. The advantages of the product are clear, but they're not everywhere. I think it is a little bit overhyped right now.

What's one GTM task no human should be doing in 2025?

Cleaning the duplicates in your CRM.

What's one piece of SaaS sales advice you just wish would die?

I'm seeing a lot of stuff about pricing. For a long time, SaaS pricing was very stable—per seat with additional features and discounted pricing if you commit for a certain amount of time. I think that's all being thrown out of the window right now, and I see a lot of people suggesting lots of different pricing methodologies. The reality is we don't know yet. It's very hard to understand where we will land, if we will land in the same place as per-seat pricing. I don't think we will. We're pricing based on workflow plus usage but evolving quickly around outcome. So we'll see where we land. A lot of people are sending advice around AI pricing, but I don't think we know yet.

Let's wrap up with a lighter question. What's your dream vacation destination?

I love traveling. I'm from Italy, so I'm lucky that I get to go to Italy a couple of times a year. I've been in many countries, but one of the best places I've been recently is St. Lucia. I think I'll be back soon. It was an incredible, magical place with beautiful sea, water, and climate. So I'll be back.

It is a gorgeous place. I'm sensing a Scalstack meeting in St. Lucia.

Thank you so much for joining the show. Where can people find you? Where can people go to learn more about Scalstack?

It's scalstack.ai. I'm super proud we just released a new website, which I spent three months obsessing about. Please go see it and check it out. My name is Elio Narsio. Thank you for having me.

Thanks for being here. Cheers.