Validate Quietly, Build in Public: Behind the Scenes of HockeyStack’s Shift into AI for GTM

Show Notes:

In this episode of AI for GTM, HockeyStack's Co-founder and CRO Emir Atli speaks to Steph Bian, Growth Lead at HockeyStack to unpack the company’s shift from a marketing attribution tool to an AI platform for go-to-market teams.

Emir shares why they’re betting on vertical AI agents like Odin and Nova, how they spent a full quarter validating direction through hundreds of customer conversations, and what they learned about the unsolved pain points in sales—like stakeholder maps and account planning.

He talks through why dashboards aren’t enough, how GTM roles are rapidly evolving, and what it takes to build tools that drive real progress, not just more activity.

Timestamps:

00:00 – Introduction and Welcome

00:09 – HockeyStack's Strategic Shift to AI for GTM

01:40 – Finding a Unique Position in the Market

03:49 – Challenges and Disproven Hypotheses

06:28 – Customer Discovery Lessons and Approach

09:57 – Building Trust and Authority Before Launch

15:25 – The Future of AI in Go-to-Market Strategy

18:19 – Why Sales and Marketing Roles Are Merging

24:27 – What’s Coming Next at HockeyStack

[00:00:00] Steph Bian: Hi everyone. Welcome to episode four of AI for GTM. We're doing something a bit different today. We're joined by Emir, CRO of HockeyStack. Welcome. 

[00:00:08] Emir Atli: Thank you. 

[00:00:09] Steph Bian: Before we dive in I'd like to zoom out for a second. So we're at an inflection point for HockeyStack right now. How are you thinking about the strategic shift and where are we at?

[00:00:21] Emir Atli: Yeah. HockeyStack started as a marketing attribution and reporting product. As you also know, we quickly found product market fit and we scaled in the last two years or so. We did our series A with Bessemer in February of the year, and there was always a strong market pull from marketing reporting to the sales side, the SDR side, AE side, and even account management side as well.

[00:00:43] Emir Atli: And we kind of saw this from the very early days when we were in YC in late 2023, in the summer batch. And we always planned for this move, but I didn't really know when it would happen. And yeah, this Q1 of this year, we kind of released our first version of account [00:01:00] intelligence. And the early feedback was really good.

[00:01:02] Emir Atli: Our first 10-20 customers really enjoyed it, enjoyed using it. And this week officially, we changed everything to AI for go-to market, to a platform that is marketing reporting intelligence workflows to say AI platform for the entire go-to-market team. We will continue building in this, in this vision.

[00:01:22] Emir Atli: Marketing reporting will always be a big part of it. Reporting in general is always gonna be a big part of it. It will be a differentiator, but HockeyStack features a horizontal product in the go-to-market space for B2B Tech companies. Yeah that's kind of an overview of where the product is going.

[00:01:39] Steph Bian: Cool. And what did you find that competitors weren't doing, and how did you find a unique position in the space? 

[00:01:47] Emir Atli: Yeah. So to build a reporting product that's flexible and is customizable like HockeyStack. Essentially what you need to do is a really giant data layer, which can ingest any type of data point and turn it [00:02:00] into an act action-based data model, which essentially means if you're using a CRM, you would be used to this model where you cannot report on like different objects.

[00:02:09] Emir Atli: Um, it's very hard to report on like. All the final stages and all that stuff because it's a relational database. So we essentially worked even before HockeyStack, in our previous startups, we were always in the data space. We spent almost five years just building the data layer that can turn any type of data, point into an action on the buyer journey, um, and then allows us to report on it to very, very easily and customizably, so that is a unique approach to solving the go-to-market challenges with AI because essentially you have unstructured data points, you have structured data points, you have web signals you have actions from the sales team, you have actions from the marketing team. How can you turn all of those data points into something actionable?

[00:02:49] Emir Atli: I think most companies in the space are more focused on how can we get the most emails out? How can we get the most calls out? It's more like a breadth thing. We are taking the other [00:03:00] approach, which is depth into every single account. Every single action. Another thing, again, as I said, is we are not focused on as many signals as possible.

[00:03:08] Emir Atli: We are focused on the action piece. So how can we action all of this data in a measurable way, in a way that stake, it can turn into revenue and pipeline. And then the third thing is most companies stop at pipeline generation. So if you look at the space, it's mostly like, how can we get most pipeline?

[00:03:25] Emir Atli: How can we increase our pipeline by x, y, z percent kind of thing. We don't stop at pipeline because if you create pipeline, you also need to close it. So HockeyStack's, especially the Account Intelligence product, takes it from pipeline generation all the way to close one step by step. And it's just, it's not just for like SDR team or AE team, SCR team or marketing team.

[00:03:44] Emir Atli: It's a go-to market platform for every single go-to market member. 

[00:03:49] Steph Bian: And was there anything that you found when you were moving from account intelligence version one to two were there any hypotheses that you found [00:04:00] were proven wrong and how did you pivot from that? 

[00:04:03] Emir Atli: Yeah. I think one hypothesis that was wrong is I, and we in general thought that a lot of the things that we wanted to build were solved problems like even stakeholder maps, we thought it would be a whole problem or like account planning. We thought people would have had a good solution for it internally. Or like a platform that they use. Those were not solved problems. That is something that was, that was a wrong hypothesis. And that I, we, I don't know if he had this hypothesis, but essentially the market wants like an end-to-end AI platform, not like a point solution. So if you look at the market, there's like, I dunno, this tool has this AI feature, this tool has this AI feature. And then if you use them all together, you are not really an AI first go to market company. Or like you don't have AI first go to market team.

[00:04:51] Emir Atli: So that was also a hypothesis that we proved in the market where people are really hungry for a solution like this where it's end-to-end AI for go to [00:05:00] market. And the last one was we thought like a lot of companies are focusing on the individual contributor, but as not just in, it's not just for individual contributors, the managers so even first line and second line managers are also interested in how can we use AI day to day? That was also the hypothesis that we proved in the market. 

[00:05:22] Steph Bian: Super interesting. When you're talking about an end-to-end solution I imagine there's still like priority areas that you wanted to address first.

[00:05:32] Steph Bian: How do you think about prioritizing the like aspects of go to market that are most important first? 

[00:05:40] Emir Atli: Yeah. So we first suggest the marketing reporting problem. So biggest uh, time waste for marketing teams is reporting. So we solved that with, which is our AI analyst. So it was the first, second one for sales teams is the biggest waste of time is other than like updating CRM and everything, tho those are solved problems.

[00:05:59] Emir Atli: [00:06:00] It is essentially how can we progress a deal. So to mo, if you take away like the updating the CRM, um, and keeping all the systems updated, the biggest waste of time or the biggest time the sales teams are spending on is how can we progress the deal? So that's gonna be our main focus with all of the products that we built on the sales side and all of the functionalities that we are building in the product.

[00:06:21] Emir Atli: It is basically, okay, we generate pipeline. Now how do we progress the deal? Okay. 

[00:06:28] Steph Bian: And when we are talking about. Like figuring out what to build. We often talk about the importance of speaking to customers and going through that discovery process. You went through that recently with Account Intelligence 2.0 and Nova, how did you think about structuring that discovery process?

[00:06:47] Emir Atli: Yeah. This is a fundamental thing that we learned during YC is talk to users. We even have hats and t-shirts that says talk to users. So even before getting started, [00:07:00] even before sharing this with the company, I mean, you were already at that all hands, um, where we kind like explained this before, that we spent a whole quarter.

[00:07:08] Emir Atli: Just like me and my co-founders just talking to people, um, before we were talking about it publicly or even in the company how we did it is, I mean, there's a really good book called the Mom Test. Essentially what what it says is if you ask your mom, is this idea a good idea, she would always say. Yes, it's a great idea I would buy it. But that doesn't really happen in the market. So how we position this as like, we schedule a ton of, uh, user interviews from our network. Even like cold DMing people call emailing people. Um, and we talk to dozens of people both on the marketing side and sales side, and then we.

[00:07:46] Emir Atli: Like really thorough discovery, not like about the product. We didn't ask any questions about the product or like, if we built this, would you buy kind of questions. It's mostly what do you doing today? How do you solve this problem? Is this a problem. If this was solved if it's like an individual contributor, if you a [00:08:00] manager came up to you and then ask you like, what are the three things that I can unblock for you?

[00:08:04] Emir Atli: What are those kind of questions? To do lots of discovery in the market. And then once we've done those, we kind of kept them updated on like the product progress to get their feedback. Um, and then some of them turned into pipeline once we actually built the product and went to market.

[00:08:19] Emir Atli: So yeah, essentially it's said discovery at scale. 

[00:08:23] Steph Bian: Was there anything unexpected that you found in the discovery process that changed the road map? 

[00:08:29] Emir Atli: Yeah, um, I think. One of the things that was interesting is if you ask like a manager, they always think pipeline is a problem. If you ask like, Hey, I see it's always the other things, like progressing a deal or like actually understanding what's going on in the deal.

[00:08:45] Emir Atli: That was one interesting thing. And then the other thing is, again, the managers like first line, second line, third line leadership on the sales side of things. They're really, really hungry for a solution that uses AI to make their [00:09:00] jobs easier. I thought it would be mainly like them unblocking the people that they work with or they manage, they, they actually want something too.

[00:09:08] Emir Atli: And on the marketing side, yeah, again, like as you build like a data product, as you have more and more insights, more and more, I know dashboard supports everything. You kind of think like it's, it's easier because you have more data. It's easier because you have more insights, but every quarter, every month you have more data to work with and it just never decreases.

[00:09:30] Emir Atli: So I think one unique hypo, not hypothesis, but one unique insight or learning from that process. Even like talking to our customers when we first release Odin was essentially not increasing the data volume or it's not increasing the insights. How can you make it actionable? Um. Everyone is getting burned out by the amount of data they have every month, every quarter.

[00:09:56] Steph Bian: Super interesting. And thinking about like building in the AI for G [00:10:00] TM space at the moment, it's very crowded. How do you think about standing out and positioning the product in like really crowded space? 

[00:10:09] Emir Atli: Yeah. And we talk about this all the time internally and also externally too. For AI to work really well, it needs a lot of context.

[00:10:16] Emir Atli: It need to categorize data, it needs clean data. So if you have, if you don't have clean inputs, you don't have the good outputs. So the differentiator of HockeyStack is always gonna be our data layer and our data foundation. So that is always gonna be the main, like big moat on the product. Other than that, again, our focus is not increasing the capacity, increasing the um, breadth, increasing the amount of emails, increasing the amount of calls. It's not increasing the amount of signals. So we're gonna be that it is always actionable data and actionable signals, actionable insights. And how do you actually action it? The way that we would build product is essentially we think about like the persona that we are billing for, let's say it's marketers and let's say [00:11:00] AEs and SDRs.

[00:11:01] Emir Atli: And then we basically, we have an internal doc in our product development cycles where we essentially turn their days into different agents and then which agents. Should be in the product and which agents are not really needed. Those should be human components. And have we built towards that? I think that's a unique approach to building in AI versus just automating everything 

[00:11:25] Steph Bian: and when we're like going to market for the first time with the new product how do you think about balancing speed of execution with all of the teams, like sales, marketing, customer success, all saying the same thing? And like being cohesive as a team? 

[00:11:41] Emir Atli: Yeah, I think I think especially, I mean, for startups, you, there are always things that are broken and you just need to be fine with it.

[00:11:48] Emir Atli: I think there's even more true for AI startups. I think a lot of companies are moving from like monthly sprints, monthly cycles to biweekly, even like weekly sprints. [00:12:00] Um, that is because it's like changing all the time. I even, you and I able even able to catch up with like the developments and everything else there's always something new. So we just need to be fine with it to a degree. But how I think about it is there's like core fundamental differentiators, core fundamental values that we have as a company. And as long as we communicate to the market, the small details will always change. So we just need to be fine with the changing details.

[00:12:26] Emir Atli: Um, so it's like a, it's not like, um. It's a framework to think about the company and think about the market. And as long as you have that framework, you can complete the gaps and you can complete the, you can fill any gaps doing, like if it's sales, if it's marketing, or if it's going to market in general, you can fill in those gaps.

[00:12:44] Emir Atli: But, um, if you have that framework as a company, then everything else becomes easier and you just need to be fine with mistakes. 

[00:12:52] Steph Bian: Yeah, definitely. And because the market is moving so quickly, like you mentioned, how do you think about building [00:13:00] trust and authority in a space that's always moving and before the product is necessarily like completely ready?

[00:13:07] Emir Atli: Yeah. Um, so there are a couple different ways if so we have, we have a good customer base. Um, so we, what we do is we do weekly biweekly product updates and do a ton. I mean, we are really. Hands-on with our customers. So we take that really seriously. We have like Slack channels with every single customer or email threads if Channel and Slack or teams, all that communication channels.

[00:13:35] Emir Atli: We take that really seriously, being really close to customers. We have lots of pilots with customers to get their feedback. Um, and. Yeah, our customer success managers, AEs, are always with the customers to understand what else can we build, what, how else can we provide more value with and doing lots of pilots with them yeah, that is the biggest thing. And then in terms of like going to market, [00:14:00] we are changing our approach from traditionally we have been always very, very strong in like content and we had a really high velocity. Uh, right now we are shifting that towards a little bit less velocity. I think that is also where the social media, I mean, LinkedIn's our biggest, LinkedIn is also moving towards that.

[00:14:17] Emir Atli: So we are prioritizing stuff like this where we actually talk about like what's going on behind the scenes. Uh, we are prioritizing high value production. We are prioritizing high value content versus velocity, which traditionally worked really well for HockeyStack and a lot of companies. But we are prioritizing thought leadership and a little bit less velocity in our content, in our go-to market.

[00:14:39] Emir Atli: Um, and in our sales process, we also prioritize more discovery and more like, so essentially a lot of companies come to us and they, this is shifting day by day, but traditionally, companies used to come to us, even like inbound or abound, and then say, I want X, Y, Z, I don't know. I want dashboards, I want [00:15:00] workflows, I want email automation, some stuff like that.

[00:15:02] Emir Atli: More like feature focus Now. A lot of companies, especially in the large enterprise, they come to us and they say, we wanna implement AI in our go to market. And that's the only thing. So you need to shift, like you basically need to shape the way that they're thinking about it. So that's what we are trying to do with our content, with our sales process, our product marketing demand gen, everything.

[00:15:23] Steph Bian: Mm-hmm. That's a really interesting point. Going off that, is there something that is like quite fundamental that you believe about AI and go market that companies aren't thinking about or maybe thinking about the wrong way at the moment? 

[00:15:39] Emir Atli: I wouldn't say the, maybe not the wrong way, but different way of thinking about it.

[00:15:43] Emir Atli: I believe that there's gonna be like one vertical AI company for each vertical. It's gonna take the majority of the market. So go to market is a little bit behind. Where it's like a legal space where we have Harvey [00:16:00] basically destroying the legal space. We have like G Clean destroying the search.

[00:16:04] Emir Atli: Space. They're like different AI companies and cursors during the product development space with ai. I think there's gonna be companies that are basically taking a lot of the market share versus like horizontal companies where you can go in and build everything like, you know, AI companies where it is very, very, very horizontal.

[00:16:22] Emir Atli: Like you can build literally everything on those com on those platforms, which I think there's space for them too, but I just don't know. Don't really think that it's gonna be embedded in the company's workflows. They're incredible companies like Rider for example. It's more like in hybrid approach.

[00:16:37] Emir Atli: You can go really deep and you can build a really, really customized, uh, you can even write code on the platform. There's, uh, definitely space where companies like that, but I believe mostly in vertical companies. 

[00:16:52] Steph Bian: Is that because of. Like with data that's underlying those companies or what makes it a stronger [00:17:00] approach?

[00:17:00] Emir Atli: Yeah. Um, because if you're horizontal you just don't have like the focus in a vertical. So like every single day we wake up and we are worked on quarter market challenges versus if you're like horizontal companies, you are mostly focused on making the company more horizontal and expanding the service area and.

[00:17:19] Emir Atli: If you're a, like, if you're a horizontal company and your roadmap is essentially impacted by what the people wanna do on the platform, and it can be a variety of things versus if you're a vertical in one, one category or like one vertical, then you can go really, really deep and you can solve bigger challenges then which increase your A, C, B because you were solving a bigger challenge for a team or vertical.

[00:17:42] Emir Atli: Um, and it's just more valuable I think, again, going back to the point solution. Products, point solution AI features. The reason that they're not really valuable is because it's solving like one minor thing in your day. Uh, maybe saving you 10 minutes every day, but [00:18:00] it's not super valuable, uh, because it's not valuable.

[00:18:03] Emir Atli: Your a CB is lower 'cause your is lower. You don't get to market. I mean, it's not really back going back to market share thing, but essentially AI needs to be more valuable and more, it becomes more valuable if fair in one vertical.

[00:18:19] Steph Bian: Considering that you think we're going to move to vertical solutions, do you think the GTM function might merge near product or data as their solutions evolve? Or do you think that there'll be like definitely a solution for each team? 

[00:18:37] Emir Atli: I think your go to market is gonna merge into, like, marketing and sales are gonna be very, very close.

[00:18:42] Emir Atli: And I think sales and account management are already close, but they'll be very close as well. I think if you look at like a traditional enterprise sales team, you have like, you know, like a sales engineer, you have an au you have an SDR. You have like roles between the SDR and AE where they [00:19:00] qualify to qualify the account.

[00:19:02] Emir Atli: And some teams, they also divide sales engineers into different, different roles. Um, and then you have like a post-sales solutions consultant type of person. I think more and more those roles are gonna merge into maybe from like five roles to two roles. 'cause you can do much more. Right now, before it was like the sales engineer would take the demo from an ae.

[00:19:25] Emir Atli: Do like technical demo with a customized POC instance, and then you then push it off to like a, I don't know, someone a bit more technical to go over like onboarding scope and stuff like that. And then you kick it off to an account manager, customer success manager, and implementation consultant kind of person.

[00:19:41] Emir Atli: So all of those roles are gonna be merging into different, like a little bit more generalized roles, in my opinion. Uh, because essentially you can do much more. And I think the AE role is gonna be more, I mean, we, we always took this approach, but in a lot of companies, they don't really do demos. They are more like project [00:20:00] managers.

[00:20:00] Emir Atli: So I think. A lot of companies, and this is what I'm hearing from a lot of sales leaders as well, I think a lot of sales leaders, a lot of companies will prioritize salespeople, especially AEs with a lot of product knowledge, can do their own demos, can get out from AI and take the deal from like stage one all the way to stage three, four, maybe get out from one person, so like a sales engineer or a little bit more technical person that can do like a customized POC instance.

[00:20:26] Emir Atli: But otherwise, I think most of the sales process is gonna be led by one person. 

[00:20:31] Steph Bian: Where do you think we are at currently in terms of like AI agents being able to affect like those roles? 

[00:20:40] Emir Atli: I think we are very, very close. I think this is like a 12 to 18 months. Like we have like probably like 12 to 18 months to be seeing a decrease in all of these roles and like job postings and stuff like that.

[00:20:55] Emir Atli: I think there's gonna be a wave that's starting right now that I also see from [00:21:00] my circle, my network, basically, if they'll prioritize these types of AEs and once they see success with them, it's gonna be decreasing. And on the marketing side, I think if you take, I. Especially reporting out of way um, a lot of marketers can do much more with their day to day. So I think we have, we are like 12 to 18 months, uh, from this cycle. And I think right now we are seeing the earliest companies emerging as like less headcount, more revenue per employee, and much more efficient. 

[00:21:30] Steph Bian: Mm-hmm. Super exciting and going into a bit your role as CRO at Hockey Stack, you lead both sales and marketing, which is not always the case for other companies. How do you think that changes the way that our team operates? 

[00:21:50] Emir Atli: Yeah. I think it changes the way that we are focused on, like everyone has focused on one goal. Two goals, qualified pipeline revenue. We don't really care about anything else. We don't really [00:22:00] report to our board about anything else. Like we don't really report on leads, MQLs, pipeline, stage one, anything like that.

[00:22:06] Emir Atli: So we are only focused on two goals. Another thing is nobody like I. Blames another team or another department. If we like, if you have less pipeline, it's all like everyone's problem. If you have less revenue, it's everyone, everyone's problem. If there's any type of problem, it's literally everyone's problem, not just one team.

[00:22:23] Emir Atli: Um, and that's because I take account accountability for literally the entire revenue function. I don't really know if it's the right or wrong option. Um, there might be companies where the other option might be better or, I don't know. I feel like a really, really high volume PLG company. Maybe two people like C-M-O-C-R-O would make some more sense.

[00:22:46] Emir Atli: Um, yeah. But I think for our current stage and current like priorities, having one person leading the entire revenue function is much more efficient and, uh, more effective. [00:23:00] 

[00:23:00] Steph Bian: When you see signs or sales and marketing being misaligned, how do you course correct that? 

[00:23:05] Emir Atli: Yeah, I think, um, we don't really have this that much because maybe I can talk about why we, we don't have it that much because essentially all of our priorities are from coming from our sales it's coming from sales. So we always go backwards from how can closes more revenue or what is not allowing us close more revenue, and then we go backwards from it. Um, so literally everything that we do day to day is based on our models and based on revenue so we are always aligned on revenue and how that goes backwards is, for example, like we do every Friday we do, speak the board reporting, and then we look at like five metrics for that peak, uh, that are coming from like qualified pipeline and revenue, which is like stage coer rates closed, lost regions and stuff like that. And then we go backwards from it. For example, if, if one week we have more closed lost [00:24:00] deals because of a specific competitor or a specific reason or like budget deals being stalled, then we go backwards from it for marketing, sales, customer success to solve those in a week or a week.

[00:24:13] Emir Atli: We improve on those metrics and we are hyper focused on those. We're obsessed about those metrics. And because of that, we almost never have misalignment. 

[00:24:24] Steph Bian: Okay. Just one final question for you. What are you most excited about when you look at the market, where it's at, at the moment, and looking at Hockey Stack, particularly Nova and Odin, what are you most excited about, maybe in the next six months of what could be possible?

[00:24:43] Emir Atli: Yeah. So our biggest goal is. Right now mean these are, this is a really, really big technical challenge. So if we have like Odin, which is like our analyst, and you have Nova, our sales agent right now in its current state, these are treated and used [00:25:00] as two separate products. The goal in the next six months is to make it.

[00:25:05] Emir Atli: So that they communicate with each other. Right now the communication layer is a little bit missing which is again, a really big technical challenge, which essentially means, so we just update our homepage. On the homepage you see daily insight. So say daily insight is, you get more pipeline if you invite people to an event, once they become stage three opportunity, how can we automatically communicate that to Noah so that we can invite people to our next event and no one knows our next event and where it is and then we already know where people are, like in terms of location in our pipeline. So how can we make them communicate with each other without a human being in the loop? And then you set like, uh, certain rules, certain rule sets in the product. That is the most exciting thing for me, is essentially seeing a product communicating like two products, communicating with each other without me or anyone else in the, in the loop and doing stuff for the [00:26:00] human.

[00:26:00] Emir Atli: Yeah, that is, that is really exciting. 

[00:26:05] Steph Bian: Well, thank you for your time. Thank you so much. Um, yeah. And all the insights that you shared with us today. 

[00:26:09] Emir Atli: Thank you.