Three years into the ChatGPT era, the dust hasn't fully settled on what AI means for sales enablement—but the picture is getting clearer. In this milestone 50th episode, we get into what's actually happening on the ground with go-to-market teams, what the hype is getting wrong, and where we think the biggest opportunities still lie.
In this episode, we share our honest take on the state of AI in sales and enablement—from the tools that are actually getting traction to the patterns we think will define the next few years.
Enjoy the show!
Eric Doty: Hello everyone, welcome back to Grow and Tell. I'm your sometimes co-host, Eric Doty, head of marketing at Dock with Alex Kracov, the CEO and co-founder of Dock. Hello Alex, how are you?
Alex Kracov: Doing good. What's up everybody?
Eric Doty: And so today we're gonna finally talk about the topic that our audience has been absolutely clamoring for our huge listenership, the current and near future state of AI and sales enablement. I'm sure it's not the last time we'll talk about it. It's not the first time, but this is our first fully dedicated episode to the topic. Before we get started, I did want to call out Alex. It's our 50th episode. We made it. I know.
Alex Kracov: Wow, we've been doing this for that long. That's crazy. Yeah.
Eric Doty: When you start a Dock, you're like, I hope I start a podcast that reaches 50 episodes. And there you go. You achieved your dream.
Alex Kracov: Yeah, I mean, I am proud we've like stuck with it. I mean, it's so easy to, I wonder how many podcasts get to 50 episodes, you know? Like, I feel like a lot of people start and stop because it's hard and we're getting there. Yeah. Okay. Nice.
Eric Doty: Your head of marketing did this research, okay? I looked it up. Apparently it's like approximately 9%. It feels too high, but I feel like, you okay, we're in the top 10%. Yeah, I also, yeah, we'll get there. All right, okay, back to AI. To set the stakes here, we're like three years into the chat GPT era. It's crazy, it's already been that long. And not to say the dust has completely settled, but.
Alex Kracov: Okay, okay. Yeah, I thought I'd be in the 1% at least, you know? Yeah, but I'll take it. We'll get to 100 and then we'll be good.
Eric Doty: I feel like there was maybe two years of chaos where we're not sure what the AI future was gonna look like. And I feel like we have like a little more clarity of the vision now. And so, you and I have talked a lot off air about AI. And in my opinion, like you've had a very kind of like level-headed take this whole time on like what the AI will do to the sales enablement industry, like what it will do for Dock and like obviously what it'll do for SaaS in general. So I feel like you're a good person to talk to about like the future of AI. And I think our goal we were just talking before we started is I think like the average sales enablement person listening to this show will probably feel like a little uneasy or uncertain from like a career perspective, right? Of like, what should I pay attention to? Like what's a waste of my time? What tools should I use? What should I not use? And so we just maybe not answer all those questions, but at least like give a bit more sense of calm to that sort of uncertainty, if that makes sense.
Alex Kracov: Yeah, excited to get into it.
Eric Doty: Cool, so like right off the bat, I think one thing you've said to me previously is that AI is kind of both overhyped and underhyped right now. And I think it's helpful if you just talk about that for our listeners.
Alex Kracov: Yeah, it's something I really believe and I'm glad we're starting with this question because it's easy to feel, and I feel this all the time as a founder of a SaaS AI company, right? That like we're super, super behind, right? Like that we've already missed the AI wave. Like if you go on X, Twitter, LinkedIn, you see all these messages of all these amazing things that people are doing with AI and it feels like people are super advanced. But then like you actually start to talk to customers. Both, we talk a lot on those podcasts with different enablement leaders or customers as we're selling Dock and it's like still super underutilized and people are still really figuring it out how it works. And it's a really nascent technology. And the people we talk with are usually some of the most forward thinking folks kind of, you know, in the world when it comes to AI and actually even one little tiny little anecdote is I was at a wedding, you know, last weekend and somebody was talking about illness or whatever. And I was like, you could just ChatGPT it, you know, and she had made a comment like, but it'll just tell me I have cancer. Right. And like that is such an old way of thinking that was like the WebMD joke, right? Like no matter what I search WebMD is going to tell me I have cancer. I have a headache, I have cancer. And it's like, no, ChatGPT actually gives really nuanced, you know, answers around medical stuff. And like this person lived, you know, not in San Francisco or New York, maybe they're more like middle America. And it's like this technology and the power of it just hasn't gotten to most of the world. Even though if you live in San Francisco, if you work in technology, it can feel like you're really behind. And so it feels like, you know, over-hyped in the short term of, you know, like everyone's using it, it's changing the world, like I'm gaining so much revenue and it's doing all these amazing things, but it's still such early days and there's still so much to build as someone who's spending most of my time trying to build these different features, like, we're just not there yet. There's so much things when it comes to the agentic infrastructure. Everyone's talking about agents, agents, agents, but if you're actually trying to build these things, we have a long way to go to kind of get there. I don't know, the dot com era, right, is the easy analogy. And that boomed and bust, but then a lot of amazing stuff was built in the 20 years after that. And so we got a long way to go to kind of like take this nugget of technology and make it incredibly useful for all of society, but then also like to go to market organizations.
Eric Doty: Yeah, for sure. I had a similar experience recently where a teacher friend of mine was doing report cards and you know, it's one of those tasks that it's very templated in that like there's only so many things that her board of education or whatever lets her say. And there's only so many like variants of what you, you know, there's like five buckets of what you can say for each. And she was like, I tried ChatGPT in the past. It doesn't work. And I was like, well, what if you like, pre-built a spreadsheet where you're like, here's the things I want to say. And, you know, I sort of walked her through the use case and then she came back to me a day later and she was like, I vibe coded this whole thing with Claude and I was able to do my report cards in two hours. And so I think there was this like consumer experience of even just trying ChatGPT for the very first time three years ago to like what's possible now. And so I think that like each person has experienced that like over hype under hype cycle too.
Alex Kracov: Totally. There's two different dynamics happening. One is people are learning how to use these tools better, right? And that template example, and how do you do it, and how do you use source material and guide the AI to what you want it to do? But then the models are just getting better themselves. I mean, in the engineering world, coding has blown up since, I think, was the Sonnet update in November of last year. And it's crazy how much it's transformed engineering. And so the things that you could put in Claude Code or Cursor, a year ago that weren't possible and buggy and didn't work, work insanely well today. And so if you're the type of person where you sort of rely on that past experience, like, that's how it used to work. It's just not true. Like these things, technology is evolving so fast. And that's like, I think it's hard for human brains to wrap their heads around it. We use something, we're like, that's how it always works. But the pace of change is exponential. And we, as humans, are bad at kind of understanding exponential change.
Eric Doty: Yeah, especially with software, right? Like you download a tool and you try it and you say, I don't like it. I don't like the way it works. You're not going to pop into it a year later and say, I hope it's better now. Like that's not a thing. It's not a way that we used to experience software.
Eric Doty: And so that does lead me into my sort of first big question for you here. Like, can you just talk about the AI landscape right now? Like the sort of tool stack in terms of like when you look at the market where we're in the end of March recording here at 2026. Like what are the big use cases that you see are like taking the most hold already? And then maybe like, where's there still lots of opportunity where a lot of enablement teams haven't gone yet.
Alex Kracov: Yeah, so we've been interviewing a lot of enablement folks trying to get a sense of, like how much are they using AI at their companies? And maybe we'll separate it into like two buckets. There's the way that the enablement teams themselves are using AI, and then there's the way that their sales teams are that they're enabling are using AI. So on the sales side, I think there's like a lot of stuff happening at the top of the funnel, right? Outbound emails, right? The classic, I think the first use case was like, help me write a better outbound email or help me take some data and then let's personalize it with AI. And there's a bunch of interesting companies like, you know, Unify or Gong who are doing different things there where, you know, trying to take different signals, maybe you're using AI to sift through all those signals, write better outbound messaging and do different triggers there.
And I think that's working pretty well, although what's interesting, I think, is as humans, we're getting pretty good at detecting AI slop versus a handwritten human email. I think, I don't know, it's an interesting dynamic that we're still wrapping our heads around. I think that's one big use case that we see in sales teams. The other one is all of the account research that happens. We've talked to a lot of folks where it's like, OK, their AEs are getting AI written reports with here's all the meetings that you have this week. Here's the status of those deals. Managers are getting those, right? A lot of account research. Or maybe it's more like, hey, we read through all of the stuff about this company on the internet. Here's a little thing that talks about their business goals and stuff like that. So a lot of the account research that you would do to prep for a new deal or an existing deal is being automated by AI with some of the most advanced teams. We talked to the enablement leader at Lagora, and they're doing some pretty exciting stuff with Notion and their agent builder. The other thing that's kind of similar to that is like handoff notes. So, yeah, like transitioning from sales to customer success, salespeople would have to write a bunch of handoff notes and handwrite that, and now you can automate a lot of that, or at least give them a head start to the CS. And then maybe the third bucket where I see a lot of usage is like CRM updating.
Nobody wants to update Salesforce and AI can really help you do that. Actually, one more which I think Dock is helping to lead the way on is content generation. And so you have tools like Gamma who are doing this in the slide deck world. But I think where Dock has found a nice little niche is around customer-facing collateral, taking things like your Gong call and then creating a business case or a customer success plan or an executive summary or a project plan.
And that's one way like our customers are really using Dock today as they work with customers, right? The sales are upside. And then there's the enablement side of things where I still feel like it is somewhat nascent, although the big pattern is like Gong and conversational intelligence, right? Like we just hear that again and again on calls, on these podcasts, like calls are the central artifact of a lot of the AI usage.
And Gong is an example that does keep coming up again and again. But I think it was really hard to do enablement back in the day if you had 1,000 sellers on calls and you had no insight into what was actually happening on those calls. You would do a training, then these calls would happen and you wouldn't really know if they were doing a good job or not. When I started my career at Yelp, I would have managers barging into my call, which was very distracting. That was how you did it. But yeah.
Eric Doty: The call was in person. It was just a meeting.
Alex Kracov: Yeah, totally. But now enablement leaders can get insights using a tool like Gong across a thousand different calls with all their sellers and get to get a sense of how well people are doing. And are they like the product certification that you just did on the new product? Are they actually following through with that? Are they, I'm a new rep. Am I actually doing a good demo? The conversational intelligence is definitely the main tool enablement folks are using. And a bridge off of that, and I think we'll talk more and more about this, is AI role play, which is the enablement folks are really excited about AI role play. Sellers, I think there's different opinions on how excited they are to actually use it. But I think that's a really interesting dynamic that we hear a lot about.
Eric Doty: Perfect, and then in terms of up and coming opportunities or areas where you feel like are maybe underutilized, maybe, I'm thinking of CMS for example, people probably aren't clamoring for, I want my CMS to be more AI powered, but there's probably lots of gains there to be made to people who are like, you have your workflow so you're used to it, so you're not necessarily needing AI in your mind, but there's gains there. Like, are those kinds of categories too?
Alex Kracov: Yeah, I mean, to be honest I think we're still figuring it out. But I think the broadest category, not, maybe taking a step back from like exact like enablement categories, but I think there's a lot of lessons to learn with what is happening in like the vibe coding or like, you know, real coding space where you have these agentic editors and you can run a bunch of agents in parallel and they run off and do a bunch of things, right? Or you have, you know, like Linear has a thing where if, which is issue tracking, I create an issue tracker, and then it goes and tries to actually fix the issue and then a human will actually review the code. So there's a lot of this AI agents running out into the world and doing things for you, which to some degree is happening with the deal review stuff and whatever. But I think there's a long way to go in the types of technologies that companies can build where you have a bunch of little go-to-market agents running around in the background.
Trying to do things, alerting people, like both creating work and maybe even doing selling or customer success themselves, but then also sharing information. And a lot of that hasn't been built out or imagined yet. You know, like one thing that we are really trying to invest in at Dock personally, and I'm spending a lot of time on, is, you know, I guess one of the barriers to entry to kind of getting started with Dock, although it's pretty easy, we have templates, is getting your stuff into Dock, right? Like you need to build a deal room template or an onboarding template.
And so one of the things we're investing a lot into is how do we just like take a dump of your source material, your marketing website, your pitch deck, your videos, all the stuff, you just throw it into Dock and then boom, our agents will go and scan through all that, splice it out and then create, propose a deal room template, create it for you and then you can edit it via agents. And so we're spending a lot of time on like that type of thing and that's very unique to Dock. But I think that type of pattern of less you clicking the UI, more agents doing things is going to take over all of go to market, right? It's not going to be just on the dock side of things, but it could be like, hey, help manage my entire CMS. Help me spin up a training, right? Like, you know, hey, go listen to all the Gong calls, take all that, figure out the things that everyone's getting wrong, propose a training outline, make a perfect training, make a quiz, send it to people. Like that type of thing starts to get a lot easier. Whereas in the past you would have to like, it would take weeks. You'd have to listen to calls and figure it out and propose and meet and stuff. And it's going to really cut that time from idea of something I want to do to activation. And I think this agentic infrastructure everyone's building out is going to be the huge opportunity over the next couple of years.
Eric Doty: So I think that leads me to my next kind of category of questions, which is, like, if you're an enablement leader thinking about your tech stack, like, I think the big question right now is like, where should the AI live or like, how is this going to shake out in terms of like, is the tech stack ultimately going to be a bunch of data with a bunch of like internally vibe coded things with Claude or ChatGPT or are the SaaS companies going to build some layer like, how do you see this actually shaking out? And then I think a question we talked about too before we started recording is like, where does rev ops fit into this? What does, like who should own this stack? I know I'm throwing a lot of questions at once, but I think like, can you paint a picture of how might this world shake out?
Alex Kracov: Yeah. So I think like one of the most fundamental things with AI and getting all these agentic workflows working is the data, right? Like none of this works that well without having a really strong data pipeline, right? Like AI could hallucinate or give you wrong answers. Or, you know, if you're trying to send outbound emails based off AI, if your data is bad, the email is going to be bad. Or if you're trying to set up AI role plays and it doesn't have good data to back up like real life scenarios, it's going to be bad. So data is the fundamental fuel, right, behind all this agentic stuff. And so that's like the first step to kind of get right in a lot of the things to build a lot of the product experiences. And that, you know, traditionally has been a little bit more of like a rev ops, you know, function, right. But also cross into engineering and stuff. And so oftentimes what we found is rev ops teams are the ones who are really owning that data layer. They own the data and then they will partner with enablement to start to build sort of like the productized experiences for the internal team. And then like some enablement people are calling themselves enablement, but they honestly start to feel a little bit more like rev ops, right? You know, and what's interesting, so that's one pattern. Another pattern I've seen is a lot of companies are developing these like AI committees or AI excellence teams or whatever.
You know, the CEO is super excited about AI. They're like everyone needs to use more AI and they'll spin up a committee to do this because there also is this interesting dynamic of going back to the data is does the data all need to be centralized across the entire company? Can there be some silos between go-to-market and the company? And how does that work? And should we buy a universal agentic thing like a Glean that sits across all company data or do we have more specific systems for specific teams. And in some ways, there's like, we're all figuring it out. There's no quite right or wrong answer. There's pros and cons to each approach, and depending on how your team sets it up and the different workflows and different tools you use. But I imagine like a future where, you know, SaaS is not going away. I don't think, I don't believe in the meme of everyone's going to be vibe coding everything. I think it's a security nightmare. I think it's a lot of work. It's a lot to maintain. For all the reasons you didn't build custom software yourself in the future, you might build like little apps, but the big things, you're not vibe coding Salesforce, right? Like that's just a ridiculous notion. And yeah, so I don't think that's the future. I think the future looks more like SaaS products really change. If you're a SaaS product who isn't adapting to this AI future and generative AI, you're not building those kind of agentic workflows native to your core products, you're probably going to die. And then you're going to have a bunch of SaaS products with agents that are all talking to each other. Whether it's an MCP or API, I think we will all figure it out. But yeah, it'll look like the Dock agent is talking to the Salesforce agent and the HubSpot agent and different things like that. I think in some parts it'll be more bundled together, like every, what's that guy's name? There's a famous professor that's like every software cycle is bundling and unbundling. And we're probably going back a little bit more towards a bundling cycle because of that data, it makes it a little easier. But then at the same time as we build out all this infrastructure, it will get unbundled again and people will be able to talk to each other through these agents. And I think one really weird paradigm is it's like, you need to build software to make it appealing to agents, not humans. Like, that's really weird. Like you need to get it so your agent is good at talking to another agent, not good at talking to Eric, not always. And so that's a funny dynamic that I think we're still figuring out.
Eric Doty: Yeah. And that reminds me of what I'm going through in content marketing right now. It's like, you also have to make content that AI bots will be able to crawl and understand and like, you know, make sure you're writing in small chunks and all these weird tips coming out. I think the same will apply to software. I did want to, you were talking about the data, and how like data is the base of all this. I think what's interesting is that like what data should sales enablement own and the way I'm thinking through this is like, the base of like your training information and maybe like content in terms of product information, competitor stuff, like things that rev ops might help with the infrastructure, but they might not, like your rev ops team is probably not gonna create your competitive documents, right? Or like they might help you build something that can crawl the internet or whatever, but you know, like what are the sort of like data points that sales enablement might own in this future.
Alex Kracov: Yeah, it's a really good question. I'll start with the rev ops team, I think definitely should own and then the enablement. I think rev ops team or just the operations data team at the company, like the CRM data, stuff, close rates, win rates, all the stuff in Salesforce, that type of stuff I think needs to be probably on the rev ops side and then your product data and that connection, which is how it's always been. Those teams have really always owned that. And I think where enablement starts to own it more is like the call data.
Although there's definitely a partnership between the call data and that CRM data because you want to get the call data into the CRM and you want to do triggers like the call is such an important sort of currency and fuel in this world of AI. So that's a place where I think there's some overlap. But then, you know, there's all the client facing collateral, which is definitely owned by enablement and product marketing, I would say, kind of share that responsibility. So that's all your pitch decks, all your one-pagers, your competitor battle cards, all your product one-sheeters, right? There's a ton of really useful messaging and data and things in there that can use to answer questions and fuel your AI agents. Then there's things like customer-facing collateral in the sense of business cases and QBR decks and customer success plans and project plans. That's an important artifact that is used between the enablement person and then the field leaders, the sales leader, the customer success leader. And then, what else am I forgetting? Like training data. I think we, as an industry, think like e-learning and the LMS world, it's like this necessary evil, I feel like. People want to kind of get away from it. Micro-learning was a big trend and still an important way. It's way better to sit through a short training as opposed to a really long-winded one. That makes a ton of sense. And now AI role play is a really hot category, which is definitely an extension of learning. And every enablement person you talk to still is like, I still need an LMS. I still need a place to put my trainings and different things like that. And there's obviously a lot of data that gets created via that. And I think that we keep saying this word data, but I think one of the really interesting things with AI is like, content is data. You know, LLMs, right, large language models are sifting through all the content and that's really interesting data that you can use to create things. And so all of the things we're talking about are like these artifacts that you can then use to fuel the different agents you want to create across the company.
Eric Doty: Yeah. And I think, you know, we launched the Dock AI agent that reps can ask questions that can get answers quickly, but that still needs the source information of, you know, what is the information about our product that the reps need. And so that still requires like the content creation step at the very beginning. Like you need your source of truth content. And I think as a marketer too, I'm interested in how this plays out. Like it feels like every company will need this sort of like, I don't know if it's built in a Wiki or something that then connects via MCP to something like Dock, but this like, okay, here's all of our products. Here's the messaging for each product. Here's the competitor information. Like all that will need to live somewhere and we'll need to connect to all these other systems that train reps, right? So maybe someday you don't need to build the course and the LMS can automatically build the course based on this source data, but like right now it still seems like the source of truth ends up being these courses or playbooks or wherever that you choose for that content to live.
Alex Kracov: Totally, and it's like changing so fast, like the different paradigms, the patterns, what tools are hot, which ones you should use. And so, I don't know, it's both an exciting time to be building and buying software, but a really hard time too, because it's all changing. And one thing that is the coolest, hottest way to do something six months ago is completely different now. And so that makes it really hard. Like it's hard to stay on top of things. Yeah, which is exciting too. That's like where I see opportunity as an AI SaaS CEO.
Eric Doty: Yeah, it's funny. I see people say like, this tool will scrape your website and have all that data available to your team. And I'm like, well, our website is outdated because our product is developing so quickly because our developers are now using AI to build so many things so quickly. And so like as the marketer, I can't keep up and I'm trying to enable our team with materials. And so I'm trying to like keep up at the pace of the product we're building as well. So it's like this doubly difficult problem.
Alex Kracov: Totally, and even on that little example, I mean, I feel like the future is like, I'm like, hey, we've launched this new thing and then boom, agent goes, takes a new thing, hooks up to Figma, makes the graphics, hooks up to Webflow, publishes on the website, right? And then you can approve it or not, right? I think that is kind of where that type of workflow is where the world's heading, but we're totally not there yet, right? Like that type, no one's hooked all of that up, but it feels like we're heading in that direction.
Eric Doty: So there's been a lot of hype around AI role play. I think you were just at the Sales Enablement Collective conference and AI role play was the thing on the tip of everyone's tongue. Like, how are you feeling right now about that category? And like, I think one interesting question is we've been deciding like, should you build it into Dock? Should you not? And can you just talk through how you're thinking about that?
Alex Kracov: Yeah, I mean, it's clearly a hot category. People are very excited and intrigued by it. I think there's like there basically was a bunch of exciting point solutions that have been working on this for a couple of years. You know, like our friends at HyperBound have been friendly with the CEO, I think is one great example of it. And they seem to be growing fast and working well. So yeah, there's this interesting dynamic of it's definitely like the hot category. You know, from talking to different enablement leaders, I hear a few things like one, enablement people are excited about it because I think, I mean, on some level, like the CEO is yelling at everyone to say, do more AI. And this is like a really cool kind of sexy AI thing that works well, right? Like it can work well for certain organizations. And I think where I've heard that it works well is, you know, during onboarding. Like new, it's like basically more immature reps, I think is where it works well, right? During onboarding or BDR cold calling, like those are the really strong use cases.
When I've talked to this dynamic of if you talk to like a more mid-market rep or a more senior rep, they're like, oh, I don't want to do the role play. You know, like they don't want to kind of do that. And it's sort of less useful than they want to learn on the job. They want to learn with coaching via Gong calls and things like that. Like sitting around talking to a bot can be weird, but it works well again on sort of like you're a younger sales rep. And then if it's a super transactional sale, right? If you're saying the same things and it's a scripted sale, then it also, I think, is a really strong use case for it. But enterprise sales, where it's so much relationship building and nuance and stuff, it's like a little less of a use case where it works well and maybe the models will evolve and stuff. The other really interesting dynamic is who should build this technology, which you kind of touched on. A bunch of the point solutions, at least to my knowledge of how they work, it's a lot of prompts that you stitch together with automation and things.
But we just talked to Stacey, the head of enablement at Gong yesterday. And I think one of the really interesting things with Gong Enable, who's doing role play is that their role play that you talk to on the other side is actually is like based off of all of your calls. So if you're talking to the scary procurement person, you want to practice that they're actually building a model based off your different calls and using them in role play, which seems amazing. You know, like that it's actually based on real data as opposed to prompting.
Eric Doty: Yeah.
Alex Kracov: So I think that's a really exciting model and I think it really makes sense for companies like Gong to do this and might make less sense for big enablement platforms like Dock to do it. But we'll see. I think it is an interesting technology and as a CEO you're always trying to find what's everyone buying, right? You obviously want to be a part of it. The other little anecdote going back to what I hear from AEs is one funny dynamic is AEs sometimes are just trying to figure out how they pass the role play as opposed to getting better at their job. We heard that from somebody. So thought that was kind of a funny dynamic with this. It's definitely cool technology and it's a great kind of like initial AI use case, which will be exciting. I think where it gets more interesting is can we take AI role plays and actually turn into AI sellers, I think is a really kind of futuristic thing, which would be cool.
Eric Doty: Yeah, I can imagine if you have this internally motivated seller who is like, how can I get more reps on calls without having more calls? Like, I really want to get better. I want to get better at this one thing. Then AI role play can play a big role for them. But if there's someone who's like, this is a checkbox I have to do to get through my quarterly training, whatever, it might be more of an internal rep motivation thing than the availability of the tool.
Alex Kracov: Totally, and like you listened to a lot of the demos on people's website and it's like still feels AI-y. You know, it's one of those things where we started the conversation where it will get better. The role plays you're doing today, 10 years from now are going to be radically different. And so that's one of those where, yeah, I don't know, the model development will really change it.
Eric Doty: Yeah. So if you mentioned this with respect to AI role play, and I think it's why it's catching on is if a CEO comes to the enablement team and says, we need to be doing more with AI. And I think part of it too is like, here's $20,000 go buy some AI. Like AI role play was an easy thing to add to the budget. Right. But if you were an enablement leader right now and your CEO came to say that to you, like, what advice would you have for those people? Like what does a healthy approach to AI look like right now.
Alex Kracov: Yeah, I wouldn't just go buy software that checks the box, honestly, as much as I'd love for you to go buy a Dock that just checks the box. I don't think that's the right approach. I think you actually, you need to start with the data. You need to kind of get your data in place and what is most important and like start with that layer and partner with the rev ops team and sort of come up with a game plan to make sure it's usable because any of the systems that you want to potentially use are going to kind of require that. So I think that's kind of step one.
Eric Doty: Yeah.
Alex Kracov: And then I think it's like what every enablement person should be good at, which is prioritization is what are the things that are going to make the biggest impact on the outcome of the business as opposed to vanity metrics, right? So it's not just trainings completed, but it's like, you know, what are the goals that sales leadership wants to do? Is it close rates? Is it, you know, sales velocity? Is it, you know, close rates related to a new product launch and different things like that? And what are the things we can do to get better at that? So, you know, let's say you want to increase win rates and you're really bad at doing business cases, maybe a place where you want to start and sort of that roadmap is AI generated business cases. But maybe a pain point is that your sales team is spending too much time writing handoff notes, right? And they're wasting a lot of time and there's just wasting time and there's too much process. Like that may be a place to start. So I think it goes back to what enablement people are good at anyway, which is coming up with what are the business outcomes, coming up with a priorities list, and then thinking through some ideas on how AI can solve it. And then there is just a lot of experimentation. If you have the resources on your team, I think the VPs of Enablement are not going to have time to mess with these tools. And so you need to empower the people on your team to bubble up different ideas. I think I feel really strongly about that. So giving them access, letting them play around with things, whether it's Claude Code or other stuff, to kind of come up with ideas as it's all changing, I also think is really important.
Eric Doty: Yeah. And if I could give some advice from like, I've been very AI forward from the second the tools were available to me being the only marketer at Dock. It was like, okay, AI is going to be the thing that, that I'm going to do. That's going to make us be able to, you know, sort of compete with large marketing teams. And I think at the beginning I rushed out to build all kinds of workflows and tools for myself. And then now that we're kind of two years into this, I realized that data infrastructure layer that you talked about is like really what I need to succeed with AI because like the tools will keep getting better that I can put on top of things. But what I really need is like a very good base layer. So for example, we have this podcast we're recording now and I have it in a database. I have the transcripts, transcripts in there, all the video clips we make. And now I've made like a database that can like make use of that content. And that connects into my like content updating engine. But before, when it was just like, okay, make a Dock writer that writes in our writing style, like that wasn't that useful because they didn't have anything to feed into it. And so now we have the podcast. I've built like a really good competitor database. They've got all this like baseline research. I've been trying to standardize like where our product information sits so that it can feed into this. So like once, now that I have that layer more built out, it's like a lot easier for me to make better use of AI as other tools come up.
Alex Kracov: Yeah, and like the easy place for enablement people to start is like calls. Like if you don't have a call recorder, like go buy that immediately and then figure out. And then once you have it, I'm going to assume most mature organizations do. That's a really easy place for you to start to kind of connect the dots when it comes to enablement and just, you know, the reinforcement on the things that people are learning and those goals that you talked about with the sales team. Like this is like, there's no better place than your call recorder, whether it's Gong or Fathom or whatever, to try and start to figure out and all these tools start to have their own AI tools now. You can start asking questions in the Chorus, Gong, whatever it is, to start to kind of connect the dots. So that's like a super easy place to start.
Eric Doty: Okay, final question, bit of a meta one is like selling AI. So we've been selling AI at Dock for let's say over a year now. What have you learned in terms of what it takes to sell AI products that was different from more like predictable SaaS products?
Alex Kracov: Yeah, and I'll preface this with like, still figuring it out. But I think one interesting dynamic with AI products is they can all kind of look the same on the surface. You know, like support, if you're selling a support chat, right? Like they all have the same little chat window and you can be like, our AI is better than this one. And so I actually heard Des, who's one of the founders and the CPO at Intercom talk about this problem where, you know, these are iceberg products, right? Like they all look the same, but what's behind them is super different.
And so I think one of the interesting things he had mentioned is they kind of give like a list of questions to ask their chat bot versus somebody else's and there's some gotcha questions that like Intercom's better at answering than whoever the next competitor. So I thought that was super interesting. The other challenging dynamic with AI as we've talked about this a lot is the data layer. So, you know, on a demo it's really hard and like to prove that your AI is better if you don't upload the customer's data into your system but there's this catch-22 of like they might not want to upload their data because of security concerns if you're not an approved vendor yet. And so how do you overcome that dynamic in the sales process is incredibly interesting. The easy way to do it is you do more pilots, right? And you say, okay, let's do a small subset, whether that's a long pilot, short pilot, whatever, but you do some subset of the data to sort of test and see if it's working, right? And you give the vendor just enough data to kind of get a sense of, yeah, like what's the overall output going to be. But that's another really interesting challenge. And then maybe the last one I'll mention is pricing, I think is incredibly interesting. You know, like paying by token and usage-based pricing is definitely like the hot, exciting thing.
And maybe the more exciting thing people talk about is a lot of like outcome-based pricing, right? Like you're not just paying for, you're paying for the work, not just like seats, right? But I like, I think this works really well in some industries. It works really well in actually engineering because engineering and infrastructure are always used to doing this. It has always been kind of usage-based. Like our server bills go up if people use it more and we're used to doing that and the whole industry is used to doing that. And it's worked well for like companies like Clay because it's often, those tools are often like single player in the sense where it's like one rev ops person or maybe a couple are just in there like scraping leads and enriching data and they have a credit card and they can just keep buying more as they run into problems, like those are empowered. But at least in our dynamic, I don't think CFOs want to give every AE access to a credit card with tokens and they want more predictable pricing. And so this is like a really interesting dynamic that we have as we sell Dock and we've decided in the short term just include it in sort of our flat rate fee as we include it and take a lot of the complexity out because as soon as you start to try and sell tokens and things that gets really complex because people are like, wait, how much token are you gonna use? And you're like, I don't know, like you've never used it before. Like that's a really tricky dynamic. And so we've opted to sort of not do that. But then there's the cost side of AI where you do need to be charging because the stuff can be expensive and needs to protect your margin. That's something I think we're still figuring out, people are figuring out. But yeah, the pricing dynamic has been interesting. I'm really happy actually with like going for flat rate pricing in the short term because it's made sales honestly a lot easier.
Eric Doty: Yeah, it's hard to have a flexible line item in the budget. It's not classically how sales enablement has worked. So perfect. Well, that's a great place to end today. Alex, thanks so much for your thoughts on AI. I'm sure this won't be the last time we talk about it. It'll probably come up in every single episode going forward, but yeah, it was really fun to chat today. Thanks everyone for listening to Grow and Tell. You can catch us at growandtellshow.com, YouTube, Spotify, Apple, all those great things. Thank you for listening and have a great day.
Alex Kracov: Thanks everybody.
Alex Kracov is the founder and CEO of Dock. He was the third employee and VP of Marketing at Lattice, where he spent five years helping grow the company from zero to over $50 million in ARR—ultimately building it into a $3 billion business. Before Lattice, Alex worked at Blue State Digital (the political adtech firm that helped elect President Obama) where he led projects for Google, and he started his career as a sales rep at Yelp.
Eric Doty is the Head of Marketing at Dock. He joined Dock over three years ago as the company's first and only marketing hire and has been building the content engine ever since. Eric has been the first marketing hire at three different startups. He's also Head of Community at Superpath, a community for content marketers.
Three years into the ChatGPT era, the dust hasn't fully settled on what AI means for sales enablement—but the picture is getting clearer. In this milestone 50th episode, we get into what's actually happening on the ground with go-to-market teams, what the hype is getting wrong, and where we think the biggest opportunities still lie.
In this episode, we share our honest take on the state of AI in sales and enablement—from the tools that are actually getting traction to the patterns we think will define the next few years.
Enjoy the show!
Tools mentioned:
Eric Doty: Hello everyone, welcome back to Grow and Tell. I'm your sometimes co-host, Eric Doty, head of marketing at Dock with Alex Kracov, the CEO and co-founder of Dock. Hello Alex, how are you?
Alex Kracov: Doing good. What's up everybody?
Eric Doty: And so today we're gonna finally talk about the topic that our audience has been absolutely clamoring for our huge listenership, the current and near future state of AI and sales enablement. I'm sure it's not the last time we'll talk about it. It's not the first time, but this is our first fully dedicated episode to the topic. Before we get started, I did want to call out Alex. It's our 50th episode. We made it. I know.
Alex Kracov: Wow, we've been doing this for that long. That's crazy. Yeah.
Eric Doty: When you start a Dock, you're like, I hope I start a podcast that reaches 50 episodes. And there you go. You achieved your dream.
Alex Kracov: Yeah, I mean, I am proud we've like stuck with it. I mean, it's so easy to, I wonder how many podcasts get to 50 episodes, you know? Like, I feel like a lot of people start and stop because it's hard and we're getting there. Yeah. Okay. Nice.
Eric Doty: Your head of marketing did this research, okay? I looked it up. Apparently it's like approximately 9%. It feels too high, but I feel like, you okay, we're in the top 10%. Yeah, I also, yeah, we'll get there. All right, okay, back to AI. To set the stakes here, we're like three years into the chat GPT era. It's crazy, it's already been that long. And not to say the dust has completely settled, but.
Alex Kracov: Okay, okay. Yeah, I thought I'd be in the 1% at least, you know? Yeah, but I'll take it. We'll get to 100 and then we'll be good.
Eric Doty: I feel like there was maybe two years of chaos where we're not sure what the AI future was gonna look like. And I feel like we have like a little more clarity of the vision now. And so, you and I have talked a lot off air about AI. And in my opinion, like you've had a very kind of like level-headed take this whole time on like what the AI will do to the sales enablement industry, like what it will do for Dock and like obviously what it'll do for SaaS in general. So I feel like you're a good person to talk to about like the future of AI. And I think our goal we were just talking before we started is I think like the average sales enablement person listening to this show will probably feel like a little uneasy or uncertain from like a career perspective, right? Of like, what should I pay attention to? Like what's a waste of my time? What tools should I use? What should I not use? And so we just maybe not answer all those questions, but at least like give a bit more sense of calm to that sort of uncertainty, if that makes sense.
Alex Kracov: Yeah, excited to get into it.
Eric Doty: Cool, so like right off the bat, I think one thing you've said to me previously is that AI is kind of both overhyped and underhyped right now. And I think it's helpful if you just talk about that for our listeners.
Alex Kracov: Yeah, it's something I really believe and I'm glad we're starting with this question because it's easy to feel, and I feel this all the time as a founder of a SaaS AI company, right? That like we're super, super behind, right? Like that we've already missed the AI wave. Like if you go on X, Twitter, LinkedIn, you see all these messages of all these amazing things that people are doing with AI and it feels like people are super advanced. But then like you actually start to talk to customers. Both, we talk a lot on those podcasts with different enablement leaders or customers as we're selling Dock and it's like still super underutilized and people are still really figuring it out how it works. And it's a really nascent technology. And the people we talk with are usually some of the most forward thinking folks kind of, you know, in the world when it comes to AI and actually even one little tiny little anecdote is I was at a wedding, you know, last weekend and somebody was talking about illness or whatever. And I was like, you could just ChatGPT it, you know, and she had made a comment like, but it'll just tell me I have cancer. Right. And like that is such an old way of thinking that was like the WebMD joke, right? Like no matter what I search WebMD is going to tell me I have cancer. I have a headache, I have cancer. And it's like, no, ChatGPT actually gives really nuanced, you know, answers around medical stuff. And like this person lived, you know, not in San Francisco or New York, maybe they're more like middle America. And it's like this technology and the power of it just hasn't gotten to most of the world. Even though if you live in San Francisco, if you work in technology, it can feel like you're really behind. And so it feels like, you know, over-hyped in the short term of, you know, like everyone's using it, it's changing the world, like I'm gaining so much revenue and it's doing all these amazing things, but it's still such early days and there's still so much to build as someone who's spending most of my time trying to build these different features, like, we're just not there yet. There's so much things when it comes to the agentic infrastructure. Everyone's talking about agents, agents, agents, but if you're actually trying to build these things, we have a long way to go to kind of get there. I don't know, the dot com era, right, is the easy analogy. And that boomed and bust, but then a lot of amazing stuff was built in the 20 years after that. And so we got a long way to go to kind of like take this nugget of technology and make it incredibly useful for all of society, but then also like to go to market organizations.
Eric Doty: Yeah, for sure. I had a similar experience recently where a teacher friend of mine was doing report cards and you know, it's one of those tasks that it's very templated in that like there's only so many things that her board of education or whatever lets her say. And there's only so many like variants of what you, you know, there's like five buckets of what you can say for each. And she was like, I tried ChatGPT in the past. It doesn't work. And I was like, well, what if you like, pre-built a spreadsheet where you're like, here's the things I want to say. And, you know, I sort of walked her through the use case and then she came back to me a day later and she was like, I vibe coded this whole thing with Claude and I was able to do my report cards in two hours. And so I think there was this like consumer experience of even just trying ChatGPT for the very first time three years ago to like what's possible now. And so I think that like each person has experienced that like over hype under hype cycle too.
Alex Kracov: Totally. There's two different dynamics happening. One is people are learning how to use these tools better, right? And that template example, and how do you do it, and how do you use source material and guide the AI to what you want it to do? But then the models are just getting better themselves. I mean, in the engineering world, coding has blown up since, I think, was the Sonnet update in November of last year. And it's crazy how much it's transformed engineering. And so the things that you could put in Claude Code or Cursor, a year ago that weren't possible and buggy and didn't work, work insanely well today. And so if you're the type of person where you sort of rely on that past experience, like, that's how it used to work. It's just not true. Like these things, technology is evolving so fast. And that's like, I think it's hard for human brains to wrap their heads around it. We use something, we're like, that's how it always works. But the pace of change is exponential. And we, as humans, are bad at kind of understanding exponential change.
Eric Doty: Yeah, especially with software, right? Like you download a tool and you try it and you say, I don't like it. I don't like the way it works. You're not going to pop into it a year later and say, I hope it's better now. Like that's not a thing. It's not a way that we used to experience software.
Eric Doty: And so that does lead me into my sort of first big question for you here. Like, can you just talk about the AI landscape right now? Like the sort of tool stack in terms of like when you look at the market where we're in the end of March recording here at 2026. Like what are the big use cases that you see are like taking the most hold already? And then maybe like, where's there still lots of opportunity where a lot of enablement teams haven't gone yet.
Alex Kracov: Yeah, so we've been interviewing a lot of enablement folks trying to get a sense of, like how much are they using AI at their companies? And maybe we'll separate it into like two buckets. There's the way that the enablement teams themselves are using AI, and then there's the way that their sales teams are that they're enabling are using AI. So on the sales side, I think there's like a lot of stuff happening at the top of the funnel, right? Outbound emails, right? The classic, I think the first use case was like, help me write a better outbound email or help me take some data and then let's personalize it with AI. And there's a bunch of interesting companies like, you know, Unify or Gong who are doing different things there where, you know, trying to take different signals, maybe you're using AI to sift through all those signals, write better outbound messaging and do different triggers there.
And I think that's working pretty well, although what's interesting, I think, is as humans, we're getting pretty good at detecting AI slop versus a handwritten human email. I think, I don't know, it's an interesting dynamic that we're still wrapping our heads around. I think that's one big use case that we see in sales teams. The other one is all of the account research that happens. We've talked to a lot of folks where it's like, OK, their AEs are getting AI written reports with here's all the meetings that you have this week. Here's the status of those deals. Managers are getting those, right? A lot of account research. Or maybe it's more like, hey, we read through all of the stuff about this company on the internet. Here's a little thing that talks about their business goals and stuff like that. So a lot of the account research that you would do to prep for a new deal or an existing deal is being automated by AI with some of the most advanced teams. We talked to the enablement leader at Lagora, and they're doing some pretty exciting stuff with Notion and their agent builder. The other thing that's kind of similar to that is like handoff notes. So, yeah, like transitioning from sales to customer success, salespeople would have to write a bunch of handoff notes and handwrite that, and now you can automate a lot of that, or at least give them a head start to the CS. And then maybe the third bucket where I see a lot of usage is like CRM updating.
Nobody wants to update Salesforce and AI can really help you do that. Actually, one more which I think Dock is helping to lead the way on is content generation. And so you have tools like Gamma who are doing this in the slide deck world. But I think where Dock has found a nice little niche is around customer-facing collateral, taking things like your Gong call and then creating a business case or a customer success plan or an executive summary or a project plan.
And that's one way like our customers are really using Dock today as they work with customers, right? The sales are upside. And then there's the enablement side of things where I still feel like it is somewhat nascent, although the big pattern is like Gong and conversational intelligence, right? Like we just hear that again and again on calls, on these podcasts, like calls are the central artifact of a lot of the AI usage.
And Gong is an example that does keep coming up again and again. But I think it was really hard to do enablement back in the day if you had 1,000 sellers on calls and you had no insight into what was actually happening on those calls. You would do a training, then these calls would happen and you wouldn't really know if they were doing a good job or not. When I started my career at Yelp, I would have managers barging into my call, which was very distracting. That was how you did it. But yeah.
Eric Doty: The call was in person. It was just a meeting.
Alex Kracov: Yeah, totally. But now enablement leaders can get insights using a tool like Gong across a thousand different calls with all their sellers and get to get a sense of how well people are doing. And are they like the product certification that you just did on the new product? Are they actually following through with that? Are they, I'm a new rep. Am I actually doing a good demo? The conversational intelligence is definitely the main tool enablement folks are using. And a bridge off of that, and I think we'll talk more and more about this, is AI role play, which is the enablement folks are really excited about AI role play. Sellers, I think there's different opinions on how excited they are to actually use it. But I think that's a really interesting dynamic that we hear a lot about.
Eric Doty: Perfect, and then in terms of up and coming opportunities or areas where you feel like are maybe underutilized, maybe, I'm thinking of CMS for example, people probably aren't clamoring for, I want my CMS to be more AI powered, but there's probably lots of gains there to be made to people who are like, you have your workflow so you're used to it, so you're not necessarily needing AI in your mind, but there's gains there. Like, are those kinds of categories too?
Alex Kracov: Yeah, I mean, to be honest I think we're still figuring it out. But I think the broadest category, not, maybe taking a step back from like exact like enablement categories, but I think there's a lot of lessons to learn with what is happening in like the vibe coding or like, you know, real coding space where you have these agentic editors and you can run a bunch of agents in parallel and they run off and do a bunch of things, right? Or you have, you know, like Linear has a thing where if, which is issue tracking, I create an issue tracker, and then it goes and tries to actually fix the issue and then a human will actually review the code. So there's a lot of this AI agents running out into the world and doing things for you, which to some degree is happening with the deal review stuff and whatever. But I think there's a long way to go in the types of technologies that companies can build where you have a bunch of little go-to-market agents running around in the background.
Trying to do things, alerting people, like both creating work and maybe even doing selling or customer success themselves, but then also sharing information. And a lot of that hasn't been built out or imagined yet. You know, like one thing that we are really trying to invest in at Dock personally, and I'm spending a lot of time on, is, you know, I guess one of the barriers to entry to kind of getting started with Dock, although it's pretty easy, we have templates, is getting your stuff into Dock, right? Like you need to build a deal room template or an onboarding template.
And so one of the things we're investing a lot into is how do we just like take a dump of your source material, your marketing website, your pitch deck, your videos, all the stuff, you just throw it into Dock and then boom, our agents will go and scan through all that, splice it out and then create, propose a deal room template, create it for you and then you can edit it via agents. And so we're spending a lot of time on like that type of thing and that's very unique to Dock. But I think that type of pattern of less you clicking the UI, more agents doing things is going to take over all of go to market, right? It's not going to be just on the dock side of things, but it could be like, hey, help manage my entire CMS. Help me spin up a training, right? Like, you know, hey, go listen to all the Gong calls, take all that, figure out the things that everyone's getting wrong, propose a training outline, make a perfect training, make a quiz, send it to people. Like that type of thing starts to get a lot easier. Whereas in the past you would have to like, it would take weeks. You'd have to listen to calls and figure it out and propose and meet and stuff. And it's going to really cut that time from idea of something I want to do to activation. And I think this agentic infrastructure everyone's building out is going to be the huge opportunity over the next couple of years.
Eric Doty: So I think that leads me to my next kind of category of questions, which is, like, if you're an enablement leader thinking about your tech stack, like, I think the big question right now is like, where should the AI live or like, how is this going to shake out in terms of like, is the tech stack ultimately going to be a bunch of data with a bunch of like internally vibe coded things with Claude or ChatGPT or are the SaaS companies going to build some layer like, how do you see this actually shaking out? And then I think a question we talked about too before we started recording is like, where does rev ops fit into this? What does, like who should own this stack? I know I'm throwing a lot of questions at once, but I think like, can you paint a picture of how might this world shake out?
Alex Kracov: Yeah. So I think like one of the most fundamental things with AI and getting all these agentic workflows working is the data, right? Like none of this works that well without having a really strong data pipeline, right? Like AI could hallucinate or give you wrong answers. Or, you know, if you're trying to send outbound emails based off AI, if your data is bad, the email is going to be bad. Or if you're trying to set up AI role plays and it doesn't have good data to back up like real life scenarios, it's going to be bad. So data is the fundamental fuel, right, behind all this agentic stuff. And so that's like the first step to kind of get right in a lot of the things to build a lot of the product experiences. And that, you know, traditionally has been a little bit more of like a rev ops, you know, function, right. But also cross into engineering and stuff. And so oftentimes what we found is rev ops teams are the ones who are really owning that data layer. They own the data and then they will partner with enablement to start to build sort of like the productized experiences for the internal team. And then like some enablement people are calling themselves enablement, but they honestly start to feel a little bit more like rev ops, right? You know, and what's interesting, so that's one pattern. Another pattern I've seen is a lot of companies are developing these like AI committees or AI excellence teams or whatever.
You know, the CEO is super excited about AI. They're like everyone needs to use more AI and they'll spin up a committee to do this because there also is this interesting dynamic of going back to the data is does the data all need to be centralized across the entire company? Can there be some silos between go-to-market and the company? And how does that work? And should we buy a universal agentic thing like a Glean that sits across all company data or do we have more specific systems for specific teams. And in some ways, there's like, we're all figuring it out. There's no quite right or wrong answer. There's pros and cons to each approach, and depending on how your team sets it up and the different workflows and different tools you use. But I imagine like a future where, you know, SaaS is not going away. I don't think, I don't believe in the meme of everyone's going to be vibe coding everything. I think it's a security nightmare. I think it's a lot of work. It's a lot to maintain. For all the reasons you didn't build custom software yourself in the future, you might build like little apps, but the big things, you're not vibe coding Salesforce, right? Like that's just a ridiculous notion. And yeah, so I don't think that's the future. I think the future looks more like SaaS products really change. If you're a SaaS product who isn't adapting to this AI future and generative AI, you're not building those kind of agentic workflows native to your core products, you're probably going to die. And then you're going to have a bunch of SaaS products with agents that are all talking to each other. Whether it's an MCP or API, I think we will all figure it out. But yeah, it'll look like the Dock agent is talking to the Salesforce agent and the HubSpot agent and different things like that. I think in some parts it'll be more bundled together, like every, what's that guy's name? There's a famous professor that's like every software cycle is bundling and unbundling. And we're probably going back a little bit more towards a bundling cycle because of that data, it makes it a little easier. But then at the same time as we build out all this infrastructure, it will get unbundled again and people will be able to talk to each other through these agents. And I think one really weird paradigm is it's like, you need to build software to make it appealing to agents, not humans. Like, that's really weird. Like you need to get it so your agent is good at talking to another agent, not good at talking to Eric, not always. And so that's a funny dynamic that I think we're still figuring out.
Eric Doty: Yeah. And that reminds me of what I'm going through in content marketing right now. It's like, you also have to make content that AI bots will be able to crawl and understand and like, you know, make sure you're writing in small chunks and all these weird tips coming out. I think the same will apply to software. I did want to, you were talking about the data, and how like data is the base of all this. I think what's interesting is that like what data should sales enablement own and the way I'm thinking through this is like, the base of like your training information and maybe like content in terms of product information, competitor stuff, like things that rev ops might help with the infrastructure, but they might not, like your rev ops team is probably not gonna create your competitive documents, right? Or like they might help you build something that can crawl the internet or whatever, but you know, like what are the sort of like data points that sales enablement might own in this future.
Alex Kracov: Yeah, it's a really good question. I'll start with the rev ops team, I think definitely should own and then the enablement. I think rev ops team or just the operations data team at the company, like the CRM data, stuff, close rates, win rates, all the stuff in Salesforce, that type of stuff I think needs to be probably on the rev ops side and then your product data and that connection, which is how it's always been. Those teams have really always owned that. And I think where enablement starts to own it more is like the call data.
Although there's definitely a partnership between the call data and that CRM data because you want to get the call data into the CRM and you want to do triggers like the call is such an important sort of currency and fuel in this world of AI. So that's a place where I think there's some overlap. But then, you know, there's all the client facing collateral, which is definitely owned by enablement and product marketing, I would say, kind of share that responsibility. So that's all your pitch decks, all your one-pagers, your competitor battle cards, all your product one-sheeters, right? There's a ton of really useful messaging and data and things in there that can use to answer questions and fuel your AI agents. Then there's things like customer-facing collateral in the sense of business cases and QBR decks and customer success plans and project plans. That's an important artifact that is used between the enablement person and then the field leaders, the sales leader, the customer success leader. And then, what else am I forgetting? Like training data. I think we, as an industry, think like e-learning and the LMS world, it's like this necessary evil, I feel like. People want to kind of get away from it. Micro-learning was a big trend and still an important way. It's way better to sit through a short training as opposed to a really long-winded one. That makes a ton of sense. And now AI role play is a really hot category, which is definitely an extension of learning. And every enablement person you talk to still is like, I still need an LMS. I still need a place to put my trainings and different things like that. And there's obviously a lot of data that gets created via that. And I think that we keep saying this word data, but I think one of the really interesting things with AI is like, content is data. You know, LLMs, right, large language models are sifting through all the content and that's really interesting data that you can use to create things. And so all of the things we're talking about are like these artifacts that you can then use to fuel the different agents you want to create across the company.
Eric Doty: Yeah. And I think, you know, we launched the Dock AI agent that reps can ask questions that can get answers quickly, but that still needs the source information of, you know, what is the information about our product that the reps need. And so that still requires like the content creation step at the very beginning. Like you need your source of truth content. And I think as a marketer too, I'm interested in how this plays out. Like it feels like every company will need this sort of like, I don't know if it's built in a Wiki or something that then connects via MCP to something like Dock, but this like, okay, here's all of our products. Here's the messaging for each product. Here's the competitor information. Like all that will need to live somewhere and we'll need to connect to all these other systems that train reps, right? So maybe someday you don't need to build the course and the LMS can automatically build the course based on this source data, but like right now it still seems like the source of truth ends up being these courses or playbooks or wherever that you choose for that content to live.
Alex Kracov: Totally, and it's like changing so fast, like the different paradigms, the patterns, what tools are hot, which ones you should use. And so, I don't know, it's both an exciting time to be building and buying software, but a really hard time too, because it's all changing. And one thing that is the coolest, hottest way to do something six months ago is completely different now. And so that makes it really hard. Like it's hard to stay on top of things. Yeah, which is exciting too. That's like where I see opportunity as an AI SaaS CEO.
Eric Doty: Yeah, it's funny. I see people say like, this tool will scrape your website and have all that data available to your team. And I'm like, well, our website is outdated because our product is developing so quickly because our developers are now using AI to build so many things so quickly. And so like as the marketer, I can't keep up and I'm trying to enable our team with materials. And so I'm trying to like keep up at the pace of the product we're building as well. So it's like this doubly difficult problem.
Alex Kracov: Totally, and even on that little example, I mean, I feel like the future is like, I'm like, hey, we've launched this new thing and then boom, agent goes, takes a new thing, hooks up to Figma, makes the graphics, hooks up to Webflow, publishes on the website, right? And then you can approve it or not, right? I think that is kind of where that type of workflow is where the world's heading, but we're totally not there yet, right? Like that type, no one's hooked all of that up, but it feels like we're heading in that direction.
Eric Doty: So there's been a lot of hype around AI role play. I think you were just at the Sales Enablement Collective conference and AI role play was the thing on the tip of everyone's tongue. Like, how are you feeling right now about that category? And like, I think one interesting question is we've been deciding like, should you build it into Dock? Should you not? And can you just talk through how you're thinking about that?
Alex Kracov: Yeah, I mean, it's clearly a hot category. People are very excited and intrigued by it. I think there's like there basically was a bunch of exciting point solutions that have been working on this for a couple of years. You know, like our friends at HyperBound have been friendly with the CEO, I think is one great example of it. And they seem to be growing fast and working well. So yeah, there's this interesting dynamic of it's definitely like the hot category. You know, from talking to different enablement leaders, I hear a few things like one, enablement people are excited about it because I think, I mean, on some level, like the CEO is yelling at everyone to say, do more AI. And this is like a really cool kind of sexy AI thing that works well, right? Like it can work well for certain organizations. And I think where I've heard that it works well is, you know, during onboarding. Like new, it's like basically more immature reps, I think is where it works well, right? During onboarding or BDR cold calling, like those are the really strong use cases.
When I've talked to this dynamic of if you talk to like a more mid-market rep or a more senior rep, they're like, oh, I don't want to do the role play. You know, like they don't want to kind of do that. And it's sort of less useful than they want to learn on the job. They want to learn with coaching via Gong calls and things like that. Like sitting around talking to a bot can be weird, but it works well again on sort of like you're a younger sales rep. And then if it's a super transactional sale, right? If you're saying the same things and it's a scripted sale, then it also, I think, is a really strong use case for it. But enterprise sales, where it's so much relationship building and nuance and stuff, it's like a little less of a use case where it works well and maybe the models will evolve and stuff. The other really interesting dynamic is who should build this technology, which you kind of touched on. A bunch of the point solutions, at least to my knowledge of how they work, it's a lot of prompts that you stitch together with automation and things.
But we just talked to Stacey, the head of enablement at Gong yesterday. And I think one of the really interesting things with Gong Enable, who's doing role play is that their role play that you talk to on the other side is actually is like based off of all of your calls. So if you're talking to the scary procurement person, you want to practice that they're actually building a model based off your different calls and using them in role play, which seems amazing. You know, like that it's actually based on real data as opposed to prompting.
Eric Doty: Yeah.
Alex Kracov: So I think that's a really exciting model and I think it really makes sense for companies like Gong to do this and might make less sense for big enablement platforms like Dock to do it. But we'll see. I think it is an interesting technology and as a CEO you're always trying to find what's everyone buying, right? You obviously want to be a part of it. The other little anecdote going back to what I hear from AEs is one funny dynamic is AEs sometimes are just trying to figure out how they pass the role play as opposed to getting better at their job. We heard that from somebody. So thought that was kind of a funny dynamic with this. It's definitely cool technology and it's a great kind of like initial AI use case, which will be exciting. I think where it gets more interesting is can we take AI role plays and actually turn into AI sellers, I think is a really kind of futuristic thing, which would be cool.
Eric Doty: Yeah, I can imagine if you have this internally motivated seller who is like, how can I get more reps on calls without having more calls? Like, I really want to get better. I want to get better at this one thing. Then AI role play can play a big role for them. But if there's someone who's like, this is a checkbox I have to do to get through my quarterly training, whatever, it might be more of an internal rep motivation thing than the availability of the tool.
Alex Kracov: Totally, and like you listened to a lot of the demos on people's website and it's like still feels AI-y. You know, it's one of those things where we started the conversation where it will get better. The role plays you're doing today, 10 years from now are going to be radically different. And so that's one of those where, yeah, I don't know, the model development will really change it.
Eric Doty: Yeah. So if you mentioned this with respect to AI role play, and I think it's why it's catching on is if a CEO comes to the enablement team and says, we need to be doing more with AI. And I think part of it too is like, here's $20,000 go buy some AI. Like AI role play was an easy thing to add to the budget. Right. But if you were an enablement leader right now and your CEO came to say that to you, like, what advice would you have for those people? Like what does a healthy approach to AI look like right now.
Alex Kracov: Yeah, I wouldn't just go buy software that checks the box, honestly, as much as I'd love for you to go buy a Dock that just checks the box. I don't think that's the right approach. I think you actually, you need to start with the data. You need to kind of get your data in place and what is most important and like start with that layer and partner with the rev ops team and sort of come up with a game plan to make sure it's usable because any of the systems that you want to potentially use are going to kind of require that. So I think that's kind of step one.
Eric Doty: Yeah.
Alex Kracov: And then I think it's like what every enablement person should be good at, which is prioritization is what are the things that are going to make the biggest impact on the outcome of the business as opposed to vanity metrics, right? So it's not just trainings completed, but it's like, you know, what are the goals that sales leadership wants to do? Is it close rates? Is it, you know, sales velocity? Is it, you know, close rates related to a new product launch and different things like that? And what are the things we can do to get better at that? So, you know, let's say you want to increase win rates and you're really bad at doing business cases, maybe a place where you want to start and sort of that roadmap is AI generated business cases. But maybe a pain point is that your sales team is spending too much time writing handoff notes, right? And they're wasting a lot of time and there's just wasting time and there's too much process. Like that may be a place to start. So I think it goes back to what enablement people are good at anyway, which is coming up with what are the business outcomes, coming up with a priorities list, and then thinking through some ideas on how AI can solve it. And then there is just a lot of experimentation. If you have the resources on your team, I think the VPs of Enablement are not going to have time to mess with these tools. And so you need to empower the people on your team to bubble up different ideas. I think I feel really strongly about that. So giving them access, letting them play around with things, whether it's Claude Code or other stuff, to kind of come up with ideas as it's all changing, I also think is really important.
Eric Doty: Yeah. And if I could give some advice from like, I've been very AI forward from the second the tools were available to me being the only marketer at Dock. It was like, okay, AI is going to be the thing that, that I'm going to do. That's going to make us be able to, you know, sort of compete with large marketing teams. And I think at the beginning I rushed out to build all kinds of workflows and tools for myself. And then now that we're kind of two years into this, I realized that data infrastructure layer that you talked about is like really what I need to succeed with AI because like the tools will keep getting better that I can put on top of things. But what I really need is like a very good base layer. So for example, we have this podcast we're recording now and I have it in a database. I have the transcripts, transcripts in there, all the video clips we make. And now I've made like a database that can like make use of that content. And that connects into my like content updating engine. But before, when it was just like, okay, make a Dock writer that writes in our writing style, like that wasn't that useful because they didn't have anything to feed into it. And so now we have the podcast. I've built like a really good competitor database. They've got all this like baseline research. I've been trying to standardize like where our product information sits so that it can feed into this. So like once, now that I have that layer more built out, it's like a lot easier for me to make better use of AI as other tools come up.
Alex Kracov: Yeah, and like the easy place for enablement people to start is like calls. Like if you don't have a call recorder, like go buy that immediately and then figure out. And then once you have it, I'm going to assume most mature organizations do. That's a really easy place for you to start to kind of connect the dots when it comes to enablement and just, you know, the reinforcement on the things that people are learning and those goals that you talked about with the sales team. Like this is like, there's no better place than your call recorder, whether it's Gong or Fathom or whatever, to try and start to figure out and all these tools start to have their own AI tools now. You can start asking questions in the Chorus, Gong, whatever it is, to start to kind of connect the dots. So that's like a super easy place to start.
Eric Doty: Okay, final question, bit of a meta one is like selling AI. So we've been selling AI at Dock for let's say over a year now. What have you learned in terms of what it takes to sell AI products that was different from more like predictable SaaS products?
Alex Kracov: Yeah, and I'll preface this with like, still figuring it out. But I think one interesting dynamic with AI products is they can all kind of look the same on the surface. You know, like support, if you're selling a support chat, right? Like they all have the same little chat window and you can be like, our AI is better than this one. And so I actually heard Des, who's one of the founders and the CPO at Intercom talk about this problem where, you know, these are iceberg products, right? Like they all look the same, but what's behind them is super different.
And so I think one of the interesting things he had mentioned is they kind of give like a list of questions to ask their chat bot versus somebody else's and there's some gotcha questions that like Intercom's better at answering than whoever the next competitor. So I thought that was super interesting. The other challenging dynamic with AI as we've talked about this a lot is the data layer. So, you know, on a demo it's really hard and like to prove that your AI is better if you don't upload the customer's data into your system but there's this catch-22 of like they might not want to upload their data because of security concerns if you're not an approved vendor yet. And so how do you overcome that dynamic in the sales process is incredibly interesting. The easy way to do it is you do more pilots, right? And you say, okay, let's do a small subset, whether that's a long pilot, short pilot, whatever, but you do some subset of the data to sort of test and see if it's working, right? And you give the vendor just enough data to kind of get a sense of, yeah, like what's the overall output going to be. But that's another really interesting challenge. And then maybe the last one I'll mention is pricing, I think is incredibly interesting. You know, like paying by token and usage-based pricing is definitely like the hot, exciting thing.
And maybe the more exciting thing people talk about is a lot of like outcome-based pricing, right? Like you're not just paying for, you're paying for the work, not just like seats, right? But I like, I think this works really well in some industries. It works really well in actually engineering because engineering and infrastructure are always used to doing this. It has always been kind of usage-based. Like our server bills go up if people use it more and we're used to doing that and the whole industry is used to doing that. And it's worked well for like companies like Clay because it's often, those tools are often like single player in the sense where it's like one rev ops person or maybe a couple are just in there like scraping leads and enriching data and they have a credit card and they can just keep buying more as they run into problems, like those are empowered. But at least in our dynamic, I don't think CFOs want to give every AE access to a credit card with tokens and they want more predictable pricing. And so this is like a really interesting dynamic that we have as we sell Dock and we've decided in the short term just include it in sort of our flat rate fee as we include it and take a lot of the complexity out because as soon as you start to try and sell tokens and things that gets really complex because people are like, wait, how much token are you gonna use? And you're like, I don't know, like you've never used it before. Like that's a really tricky dynamic. And so we've opted to sort of not do that. But then there's the cost side of AI where you do need to be charging because the stuff can be expensive and needs to protect your margin. That's something I think we're still figuring out, people are figuring out. But yeah, the pricing dynamic has been interesting. I'm really happy actually with like going for flat rate pricing in the short term because it's made sales honestly a lot easier.
Eric Doty: Yeah, it's hard to have a flexible line item in the budget. It's not classically how sales enablement has worked. So perfect. Well, that's a great place to end today. Alex, thanks so much for your thoughts on AI. I'm sure this won't be the last time we talk about it. It'll probably come up in every single episode going forward, but yeah, it was really fun to chat today. Thanks everyone for listening to Grow and Tell. You can catch us at growandtellshow.com, YouTube, Spotify, Apple, all those great things. Thank you for listening and have a great day.
Alex Kracov: Thanks everybody.
