The New Marketing Moat

Julia Nimchinski:

And with that, we are transitioning to our all-star CMO panel. Welcome back to the show, Kelly Hopping, super excited to have you back. How have you been?

Kelly Hopping:

I’ve been great, thank you. It’s great to see you, Julia. How are you?

Julia Nimchinski:

Oh, I’m really, really… Excited for this. Let me share my screen, and let’s do a round of introductions. I was gonna go first.

Kelly Hopping:

I can start, we can go right in order if y’all want. Awesome. So, I’m, I’m Kelly Hopping, I’m the CMO at 6sense, I’m, I don’t know, maybe a 3- or 4-time host, of a panel or fireside chat or anything on this, with, this group, so excited to be back, and excited to meet all you guys today.

Richard Dumas:

And I’m Richard Dumas, I’m one of the co-founders and CMO of SlateCX, where we build a buyer engagement platform.

Laurie Pratt Ehrbar:

Hi, Laurie Ehrbar, I am the CMO of Salesloft.

Heidi Darling:

Hey everyone, Heidi Darling, CMO of Savo, a human intelligence platform, and we turn conversational AI interviews into, human contextual data, which is kind of the buzz right now, and I’ve worked with both Salesloft and 6sense in previous lives, so, nice to see your faces.

Ray Wizbowski:

And I’m Ray Wizbowski, CMO of ECI. We are a ERP for SMB manufacturers, distribution, companies, field service companies, and construction companies.

Kelly Hopping:

Awesome. Well, thank you guys for, for being here. We’re gonna go through, a little bit about the new marketing moat today. So we’re gonna talk about, with this evolution of AI and, and our go-to-markets, and what that means for where we’re placing our bets, what our tech stack looks like, how we’re evaluating, our protections, how we’re building our differentiation these days, and what that looks like. So we’re gonna… let’s just start really quick, and I’ll kind of hot potato at whoever wants to… to chime in, on this one, but I’d love to go around real quick.

I think all of us are at various stages of adopting AI in our go-to-markets, whether it’s using it actually in our go-to-market strategy externally, in the way we built our tech stack, or whether it’s internal, just that we have an… we have clawed writing blogs sometimes. Like, where are we in our journey? So, so I’d love to go around. What’s the one thing that’s, actually changed in your organization or in your marketing team, in the way that they operate On a daily basis, not a pilot, but the way they operate on a daily basis, using AI.

Laurie Pratt Ehrbar:

Oh, I can take it. So I… in case you missed it, we just did, like, a big rebrand, and and our glow-up involved using Figma, Adobe, Cursor, Claude, of course, right? It saved us, honestly, about 250 hours of work every month. Which freed up the team for actual thinking, production, and we did with smaller teams. So, I would say, like, that, we’re gonna just keep scaling that, and using that, going forward. So, yeah.

Ray Wizbowski:

Yeah, one of the things that we recently did, we fired Drift because they got hacked, and that was not a good thing. So we went out and brought in qualified, so, you know, an AI Piper, the AI SDR, and implemented that about. 6 months ago, took 2 months of learning, so it learns the entire system. You know, it looks at all the assets on your website, and pulls that in, and then becomes a dynamic conversational AI BDR function, and we’ve seen about a 70% conversion uptick because of implementing that. So it’s reduced the amount of that stuff that’s going to my SDR organization.

Kelly Hopping:

Well, the good news on that is that Salesloft has now, bought Drift and Clary, and is now, Fixing any of that stuff, so hopefully it’s, it’s all good.

Ray Wizbowski:

Sorry for that, but…

Heidi Darling:

I’m in kind of an interesting point, where I identified in my previous role exactly how I would start from the ground up, building a team, and, feel really privileged to have been able to implement that, which is that I started with an Agentic team, because I really wanted to see what was possible. having 4 agents, and then scaling that to 8 before I brought in… brought in humans in the loop. So, now have 2 humans on my team, in addition to 8 agents, and so that’s been a really fascinating and… and really, radically effective way to optimize.

So, and I… I did a… I was doing some analysis over the weekend, and we’re 3 times more… I, like, did… kind of hour raters, like, if you were doing interviewer raters, I did hour raters, and I… and I… so I did quant and qual, and for every 15 minute, hour-minute, 4 hour blocks of projects, we’re… were operating at, like, 3X efficiency.

Kelly Hopping:

Wow, that’s amazing. I mean, isn’t that… I don’t mean to interrupt the conversation, but I think it’s so fascinating. I came into a role 8 months ago, fairly similar to my role 3 years ago, and yet approached it completely differently. To your point, Heidi, like, you’re thinking Agentic first instead of thinking people first, and kind of working with what you’ve got. It’s just such a different mindset. You staff differently, you hire differently, your layering is different, your spans and layers are… Top-heavy instead of bottom-heavy, and…

Heidi Darling:

Every role is different?

Kelly Hopping:

Yeah.

Heidi Darling:

You know, like, you’re not like, should I hire a growth engineer or a product marketer? It’s like, yeah, and what do those look like together?

Kelly Hopping:

Yeah. Yeah, that’s fascinating. What about you, Richard?

Richard Dumas:

Yeah, I guess I’m also a little bit unique here, because we’re a startup, we’re just about a year old, so I’ve been building the GTM function from the ground up, and it is all Agentic, so I’ve got a handful of contractors that work for me, but most of the core functions, I’m using AI to build and to automate, and what I’m really noticing that’s changed so much is that many of the tasks that a person would do, you know, just task-oriented things of writing something or building a campaign can easily be automated now. So, for instance, when we get a new customer, we tend to build a new instance of our product for them.

And with Claude and a number of connectors, I can go in to SEMrush, understand what their most high-volume web pages are, use our own connector within Claude, and spin up an instance for them, and build the campaigns. With the HubSpot connector, I can connect that to all the HubSpot workflows and campaigns. So all those things, I would have needed, like, a marketing automation person, I would have needed, you know, a demand marketer, and whatnot. All those functions are basically automatable pretty easily now, so it’s an interesting new world we’re living in.

Kelly Hopping:

Yeah, yeah.

Ray Wizbowski:

Yeah, I’m on the other end… on the other end of that spectrum. I have a, you know, team of about 100 people, and I came in, was asked to grow, you know, 10%, you know, north of 10% year over year, but not ever increase my headcount. And so we’ve actually reduced the headcount and put in more automated processes. So, Richard, very similar to you, we have a go-to-market engine that identifies signals. And then takes us through the whole go-to-market motion. Now, there’s humans in the loop all through there because of… Of the scale that we’re trying to reach, but it’s changed the entire way that we look at the marketing stack.

Kelly Hopping:

Yeah. Yeah, it’s really interesting to see where you’re building automation in. I have lots of theories and thoughts about the actual data and intelligence that’s feeding those automations, because I think they’re so powerful as long as they’re built on, sort of, the right fundamentals. So we’ll go into that in a few minutes. I want to get to our topic, which is officially this new marketing moat. And I think the moat has kind of swung over the years. I have, I’ve sort of experienced it when it was all about brand, and I’ve experienced when it was all about the sort of, you know, the measurable marketing, as we talked about, sort of all the pipeline things.

Where do y’all think the moat is? I think it’s been brand, it’s been content, volume, it’s been budget, it’s been, you know, it’s been some product feature, some capability. What do you think the actual moat is now? That’s… that protects your brand out in the market. What do you think that can be?

Richard Dumas:

Yeah, I’ll take that one. So, I mean, you’re right, so much of what we do now in marketing can be automated, and everyone’s talking about the amount of AI slop that’s out there and being created. But I think it really is based on how well you know your customers. How well you know what questions they’re asking, and how you can build answers to those questions. So, we recently did a study where we looked at about 130 different B2B sites, and actually only 4% of them could actually answer many of the most important buying questions.

They could say things about, you know, we make lots of hyperbolic claims about what we can achieve, but we don’t show you how, we don’t show you what it looks like to deploy, we don’t tell you where your data is stored, you know, everything has been about forcing someone into booking a meeting for a discovery call, and now customers’ perceptions have completely changed. I mean, they want to self-serve, and they want to get that kind of deep set of answers about your business, and if they don’t get answers about you, they’re going to get it about your competitor.

So your ability to really understand what those questions are and have really strong answers that show up not only on your website, but when they look in AI answers or AEO or GEO, I think is a real marketing moat.

Laurie Pratt Ehrbar:

I expected.

Heidi Darling:

Yeah.

Laurie Pratt Ehrbar:

Oh, sorry, go ahead.

Heidi Darling:

plus one to that, and I, and I guess, like, the fine point that I would put on it is… we can all go fast, we can all automate everything, everything can be grayish. It’s… that understanding, that deep, nuanced understanding of your customers and your employees, because there’s definitely correlation there, that allows you to… to build trust and have authenticity, which I think those are the two things, trust and authenticity, that make you stand out today.

Richard Dumas:

Yeah.

Kelly Hopping:

Yeah, for sure. Laurie, did you want to say something?

Laurie Pratt Ehrbar:

Yeah, no, I was just gonna pile on that, because it’s proof can’t be faked, right? So, like, and it’s… and it’s, as we all just said, AEO, GEO, right? AI knows to look for the credibility. It’s gonna pull from, like, the signals, the validation, the unfiltered user-generated reviews. So, so yeah, just piling on the same. It’s the loyalty is the moat.

Richard Dumas:

Well, in the.

Ray Wizbowski:

And I’ll…

Richard Dumas:

I’m sorry, I didn’t.

Ray Wizbowski:

I’ll add just an adjacent comment to that, which is, we sell specifically to verticals, and the vertical specificity, the context of what we have to say, is far more credible because of the We know the business process for, you know, 20, 25 years. We’ve been working in these very specific niche and sometimes broader category industries. that as we apply, you know, any sort of AI to it, we can bring, both in our products and then how we go to market, we bring that vertical specificity that they’re looking for. And, you know, I think, Laurie, you just said that, which is.

Is when… when they’re searching, they’re searching for something very specific, and so getting down at the level that is what’s my pain point, what am I trying to solve for, is, you know, it’s far more important for them to hear the language that they’re asking in, and it’s answering the question that they’re actually looking to answer.

Richard Dumas:

That’s right, yeah.

Kelly Hopping:

Yeah, I see so much of a, I sort of think of this, like, balloon squeeze, right? I mean, we’ve seen the whole.

Ray Wizbowski:

Continue.

Kelly Hopping:

unless we just started leading marketing yesterday, like, we’ve seen this continuum of… of where to place the bets, and I think… I feel like the last decade has been so much on, pipeline, right? Just pipeline, pipeline, pipeline. What can we do there? And so things like press releases were all of a sudden a waste of time, because nobody looks at those. Or, like, just different, like, brand tactics, like sponsorship, like, any of those things kind of, like, fell off the wayside, even events to some degree. And then what you’re realizing now with the LLMs, like, brand carries such a different weight, and all of a sudden those reviews matter, those press releases.

I feel like now we issue press releases because we can… it’s the one place we can control the narrative. No one’s going to read the press release, but the LLMs are. And so at least they’re going to be trained on the right… the right positioning, the right messaging. So I feel this balloon squeeze to brand mattering more than ever. So when I think about the moat. I sort of think about, and maybe that’s tied to the loyalty and the customer experience and knowledge that y’all talked about, but I do think so much of that is tied to what is your brand. stand for and represent in that, because, they’re getting so many answers now through, through Claude or anything else, and so having those, you know, getting recognized there makes a big difference, so it’s, it’s a different world.

We’ve seen that, I think it’s, like, 90… 90% of folks come to their RFP with a short list of vendors. And then 83% of the time, they choose number one on that shortlist anyway. And so you realize the role of, like, the salesperson in converting them is very low. The role of the brand prior to engagement matters more than ever, and so figuring out how do we capture those folks, early in that. So, pretty interesting stuff.

Richard Dumas:

I think that’s interesting. I think one of the stats was that 80% or more of the evaluation is done before anyone ever talks to you.

Kelly Hopping:

Yeah, right? I mean, which is crazy, and so that means, like, all of a sudden, demos play a different role than they used to, and, product tours, the ability to play and touch the product, reviews matter more than ever, all those things that, because people aren’t, they can self-serve so much more than they used to be able to.

Ray Wizbowski:

And one of the things that we’ve done with my team is that we’ve built a messaging app that took our messaging framework, loaded it into… we use a couple different tools, but in this case, it’s Abacus. which is a, you know, an LLM multi-finder tool, and it actually creates a scorecard way of approaching our web content, so it thinks in scorecard. It thinks in the question, please provide the top three acts. And so we build the content on the website that matches what an LLM would look for to put us into a scorecard.

Heidi Darling:

I think that’s been one of the, like, most fun and fascinating challenges, is this split of marketing for humans and marketing for LLMs, where you want to keep the human at the center of things, and you want it to be… dynamic, a little bit different, stand out. Kelly, to your point about brand, but then also making sure that you’re… you’re marketing directly to exactly what the LLM needs to absorb, and all of the, you know, ticks and ties on that.

Kelly Hopping:

Yeah. Yeah, for sure. Question, how do you make… You talk about, like, sort of the balance of, of marketing to humans, marketing to LLMs. There’s also the just marketing quality that sounds like humans and sounds like LLMs, or sounds like AI. How do you make sure that your marketing material, the quality of your content, your blogs, your social posts, your, any of those, your events, anything that’s kind of feeling, outward focused. How do you make it sound more human? Or smarter, or less? I mean, Heidi, you’re doing it, you’re doing it with 2 humans and eight agents, so how are you making your stuff sound human?

Heidi Darling:

Yeah, it’s like, I beat the crap out of it every day. No, it’s an ongoing process, and I would say on a macro level, it’s the same as it’s always been, data in, data out, right? You put good data in, and you have the rigor and the discipline to update that data daily, so I do a batch mass of how I’ve edited Against what it provided. and provide that every Friday for 2… for 20 minutes. 2 hours, no way, 20 minutes. And then it gives me a takeaway of what it should add to, and I have 3 internal voices, like, internal communication, external communication, and then kind of social… social voice for each of my thought leaders.

So I’m, like, it’s, it’s a, it’s a structure that you build, and you have to, you have to update it weekly.

Kelly Hopping:

Yeah. Any other thoughts on that?

Laurie Pratt Ehrbar:

I do think, though, you still, like, to, add to that, I agree with all that we do the same thing, and then there’s always the piece that you still have to humanize it. Right? So you still need a person deciding on it to make sure it does still sound human, right? Because if we just took it and shipped it, we’d miss all that opportunity to make sure it resonates, right? Absolutely.

Heidi Darling:

Absolutely. And it slips. There’s drifts.

Ray Wizbowski:

you have to.

Heidi Darling:

Keep doing it.

Kelly Hopping:

Yeah.

Richard Dumas:

Huh?

Ray Wizbowski:

Yeah, and we have a human in the loop, on that, but we… so we do similar, things Going back to the industry specificity, like, part of of the messaging framework that is in, you know, the writing tools that we use is we’ve built out skills that are very specific to, you know, what does a construction company look for, and what… how does a construction… what’s their vocabulary? And so we’ve built those skills within both Claude and Abacus to make sure that we’re, you know, we pull in the skill, and then that becomes… begins the beginning of the process of the content creation piece.

And so, it’s, you know. We’ve automated most of that, but then at the end, there’s always a human that checks it before it goes live.

Kelly Hopping:

Hmm.

Richard Dumas:

Yeah, I think that’s right. I mean, we… you definitely have to have a human in the loop, or a supervisor. You can’t just let the AI run wild in content generation. But also, I think it’s the kind of content you generate, so I mean… we see so much slob is like, you know, 10 trends for 2027, you know, anybody can write that, any AI can write it, and I think that annoys people, but what doesn’t annoy them is when the content is about solving their specific problem. So, one is to focus in on, is it actually helpful and useful versus being kind of traditional noise that’s out in the market, and then humanizing it, as you guys have said, in the one step in the process.

Heidi Darling:

I think it’s that personalization, which has been a buzzword for, I don’t know, 10 plus years, right? But, now it’s just personalization at scale, and one of the things that I feel again, like, fun challenge, or feel so lucky to be able to do, is because SAVO’s a conversation that can help you know anything about anyone. it’s the perfect lead magnet. So, we get to hear exactly what the pain point is, and who the person is, and then return a product that’s personalized to them. And, That is… that has been a game changer in terms of conversion rates.

Kelly Hopping:

Is personalization creepy now, or do we think it’s okay? Like, does it feel, bizarre when something comes in and it knows so much about you when it… when you received a note, you received some sort of custom ad, any of that?

Laurie Pratt Ehrbar:

I think it can still be creepy. So, we’ve all had that, where you’re sitting at dinner, and you mention a product, and then it pops up on your phone. It’s creepy. I think when it’s referencing show you that it’s watching, that’s creepy, right? But when it… when it references something that’s, like. when it’s relevant. So, let’s say it’s sales and you’re referencing, like, a funding round, or a job change, or explaining why your timing is essential. Like, those are the things that mean something and make a difference, and that kind of personalization. Because it can build trust, and you can build a relationship from it.

But watching for watching’s sake is… is creepy.

Richard Dumas:

Yeah.

Ray Wizbowski:

There’s a… there’s a bridge in between there, because we have… we’ve just… this isn’t something that we’ve launched at scale, but we have a pilot running right now. It’s… it’s somewhat partially that we’ve built, and then we’re using a bunch of tools to accomplish the end goal. But it’s… it’s a signal watching, so we’re… we’re in Reddit, and somebody’s in a Reddit subreddit, and they’re… they’re writing, hey, I’m… I’ve got a problem with this, you know, with a… a specific pain point within their industry, we capture that, we see who that is, we enrich the data, then we send a personalized email and say, hey, we saw that you had this comment in Reddit, we didn’t want to, you know. blast in the form, so here’s some things that you could be thinking about, and it’s actually just starting this, and we’re getting some pretty positive responses from it.

Richard Dumas:

That’s… that’s really smart, Ray. That’s a good idea. I love that idea. I think what I feel creepy about is we’ve probably all been getting these outbound emails that try to be personalized by AI.

Ray Wizbowski:

Yeah.

Richard Dumas:

Hi, Richard, great blog post. You, you know, you wrote great comment on LinkedIn, saw you commented on so-and-so’s post, right? And you know, it’s like, oh my god, another one of these AI, quote, personal.

Kelly Hopping:

Oh, those aren’t real?

Richard Dumas:

So what I do think, we sell a buyer engagement platform, so if someone comes to your site and they want answers to what’s the ROI for my business, or what does the deployment cycle look for, like, for a team for us, you know, can you store my data in a particular geography? I mean, that kind of personalized answer actually creates value. It doesn’t sound like they’re eavesdropping on you, or just being kind of creepy and manipulative.

Kelly Hopping:

Yeah, I feel like when we send personalized notes, I try… I’m… I feel like I’m showcasing, because I work for 6sense, which is… Basically, contextual intelligence for go-to-market, so that you can kind of know who’s in buying, and their intent, and all that good stuff. And so, sometimes I feel like this is creepy, but then I also think… but it’s also showcasing that I know this information because I use Success, right? And, like, I know that you’re in market for this product because I can see it in our product, so it’s almost one of those, like, do billboards work?

They just did, you know, it’s one of those, like, kind of like the proof is in the fact that I just sent you this note. So yeah.

Laurie Pratt Ehrbar:

I was just gonna say that, Kelly, but it’s like you and I are on the both sides of that, so it’s like the marketing and the sales side. So, like, we do live in the personalization, but not in the creepy side of it, of the, hey, this is relevant to you, right? Like, let me bring you what’s relevant to you and show you that I care, and I’m paying attention, and I understand your business, not I’m watching you.

Kelly Hopping:

Right, yeah, there you go, that’s probably the differentiation.

Heidi Darling:

Yeah, and I think… just, I’ll just go back to macro level. I mean, I… look, I read the news, I’m… I love the environment, like, I’m… I… AI, who knows what’s gonna happen, right? But I am a little Pollyanna AI, I think, because in some ways, I think it’s restoring some sanity, which is like, okay, just reveal that you’re… you’re watching. Just tell them that you’re, you know, like, just tell the truth, just be authentic. Think about brand, like, don’t… be different. Do be different, don’t do what everybody else is doing. It’s kind of like, It makes you have to return to the basics and almost, like, the common sense of life, in a weird way.

Kelly Hopping:

Yeah. Yeah, I love that. That’s a… that’s a good, healthy… I like Pollyanna view, I like that. The, we just got a question in the chat I wanted to… to tee up before we keep going. The question is about the lines blurring between paid, earned, and owned media. We used to kind of structure our teams in that way, maybe we still do. But which of these, as those kind of blur together a little bit, which of those do you think are actually transforming Agentically faster than the others, and why? Like, within paid media, what’s happening faster, what’s happening slower, which parts within paid are keeping up with Agentic, or where are we using it successfully?

I’d love to hear y’all’s take on maybe your own case studies of how that’s worked for you.

Heidi Darling:

I think the process of automating and learning and optimizing paid is happening faster. I think the owned is more valuable than it’s ever been. Oh, actually, I’ll say… Earned is more valuable than it’s ever been. Owned is a close second.

Laurie Pratt Ehrbar:

I would… I completely agree with you. Yeah, earned and owned, for sure. Yeah, and we can automate so much on paid, right? And we can learn so much faster than we could have before. You know if something in paid is working much faster than you used to, and it’s only getting better. So using all those tools now for paid, I mean, that’s just… you can do that at scale. And you can… you can prove your pipeline much, much more rapidly than in the past. But yeah, earned and owned, for sure, all day. Yeah.

Ray Wizbowski:

Yeah, and I think earned in particular, because there’s so much more credibility placed on an earned piece of content out in the world. That, you know, the LLMs place a higher credibility on that. When it comes to owned, I think the… especially the long-form media, that… that, to me, that is where we’ve seen a lot more success in that getting picked up. And I agree with you that, you know, paid is something that we’ve optimized, and, you know, we optimize it internally, we have an agency that’s optimizing it as well. We’re learning in cycles that are, you know, 10 times faster than they’ve ever been.

Kelly Hopping:

Yeah, we run, multi-touch attribution, and I’m, I’m sort of a staunch advocate for it, and a geek about it, but our top 4 channels, right now, which have switched, like, in the last year, but our top 4 in the past 6 months. organic and owned. So, organic search, organic social. I say organic search, SEO, LLMs, AEO, whatever, all of it. Organic social. paid social and paid display, which both operate as sort of advertising vehicles in the paid channels. And it’s really, like, top of funnel brand and owned, which both feel like things.

Again, those are… a couple of those are paid, but they’re still brand awareness channels, versus, like, where we used to do the, like, conversion tactics. you know, demo forms, paid search, all the things that sort of do the bottom of the funnel conversion, those have dropped off considerably.

Richard Dumas:

I’m just curious to ask the group, so I think the question was, which is transforming agentically faster? So, do you have a way of transforming earned media agnically? Because I think that, when I thought about it, I was thinking, okay, well, Agentically, maybe it is more paid media, where there’s lots of… it’s easy to put the LLM in there and do a lot of analysis and optimization, but convincing a journalist to care about your solution. I don’t know if an LLM can do that, can they?

Heidi Darling:

There’s… well, Agentically, you can have PR agents that are scanning for what… well, I’ll give you one of my little secrets here. Scanning for what… What journalists are writing about, what they care about, and personalizing the pitches all day long. So you can scale your outreach. And then the other piece of that is, I think, reviews. So you can also do something similar with your customers with reviews.

Richard Dumas:

Interesting.

Heidi Darling:

review sites.

Kelly Hopping:

Any other thoughts on that? Alright. So be honest, so Heidi’s already shared, but how has your headcount, team structure, any of that changed in the world of AI? Like, how do you, structure? What do your spans and layers kind of look like? Where are you placing your bets? do you have a go-to-market engineer? Is that a thing? Are we now go-to-market engineers as CMOs? It sounds like with Heidi spending two and a half hours training like crazy on her agents, she might be a go-to-market engineer these days.

Richard Dumas:

But what is.

Kelly Hopping:

What does that look like? I’d love to hear how everybody’s thinking about their teams differently.

Richard Dumas:

I think I talked about this a little bit earlier, I mean, now I’m at an early-stage startup, before I ran bigger marketing teams for Five9 and Vonage and whatnot, and it’s just, I think the roles and jobs are really changing. Less from being coordinators and task doers, to being people who train the AI and supervise it. Right? So being able to create a skill, I think, as Heidi just described, to go out and find out how to do earned media, and then to supervise that. is a new job, really, right? Or a new capability. I think we’re gonna have more and more people that do that versus, you know, sending out emails, doing the pitches, right?

So I think it’s really, really changed there.

Ray Wizbowski:

I agree.

Laurie Pratt Ehrbar:

Well, I think the, that we always hear that same saying, AI won’t take your job, but someone who knows it will, right? We’ve heard it, like, a million times. But I think that’s real, so I do think our job as CMOs is making sure that someone is already on our team. And… and highlighting the real, actual use cases all the time, and making sure we’re training them, and training them to train the agents, so that they’re more technically savvy, so that they’re not just, like, executing like they used to.

Ray Wizbowski:

Yeah, I think for us, it’s a bit of a blend. We’ve spent a lot more… we’ve shifted into our operations organization, so I’ve hired go-to-market engineer. I have, you know, a couple people who are just application-specific folks that are watching those applications and, you know, whether training them or being the human in the loop to make sure that they’re operating the way they should be and not hallucinating or drifting, and so… There’s a big… much bigger emphasis on that part of the organization. On top of that, we… so because we… again, going back, we’re vertically specific, so I have teams that are focused on each one of the verticals.

And so when I look for a demand gen person or product marketing person, I’m now looking for people who are, you know, thinking about how do I apply an Agentic process to this function, as opposed to, how do I send out an email? I want them to be thinking about things differently, and so it’s actually created a significant amount of change within the organization. Over the last couple years, we’ve… we’ve retooled the entire team with that mindset, and we’ve been giving them the tools. And we haven’t got it all… we’ve got it wrong as, you know, as much as we’ve got it right, and it’s been a… it’s been an iterative process.

But the team… the headcount that I started with 2 years ago is still the same headcount I have today, and we’re… You know, we’re growing, you know, greater than 10% every year. So we’re able to scale the organization, get more efficiency out of our spend, more efficiency out of our organization. Without more people, we’re just leaning into technology.

Kelly Hopping:

I love that. I’m, like, writing myself a note as you, said it, so I went dark there for a second. Go ahead, Heidi.

Heidi Darling:

No, I was just gonna say, I think you… you really eloquently elaborated on just systems thinking, and I think that that is, you know, if you have to do it twice, can you automate it and think about it as a system versus a single execution or a single task? That’s been the biggest switch for me, is just systems thinking across the board.

Kelly Hopping:

Yeah. You know, I had this, like, leadership philosophy. I’ve done the CMO thing now a few times, and I’ve had… I had this philosophy kind of always coming in of, like, this sort of almost, like, religious commitment to, two spans and layers, right? Like, I had this kind of, like, one to four to seven directs, right? Like, I liked short, fat trees, not tall, skinny trees. Like, I didn’t need one person managing one person who managed one. I wanted, like, I’m gonna have four to seven, they’re each gonna have four to seven, and we’re gonna have this highly efficient team.

And I feel like… and even… I would say even a job ago, I might have been in that philosophy. Now I catch myself with sort of these, like. robust treetops and these little skinny legs to stick with my tree metaphor, because, like, I need, like, I want, like, studs, like, strategy, AI first, big thinkers in my directs. And then the teams under them are really small and lean, because it’s a combo of agents and humans.

I mean, a combo of agents and kind of a couple people, but it ends up being, like, this very… I picture, like, these, like, trees at the beach, you know, these big lofty tops and little tiny skinny legs, because I think that’s what it ends up being, and so the investment I’m making is sort of at the marketing functional leaders, the sort of VPs and senior directors of marketing, and then they have small teams underneath them now. It’s just a very different mindset. and instead investing in, like, what are the agent teammates that I need them to all have instead of… And how much of that am I driving versus them?

And then trying to figure out, you know, this go-to-market engineering function. I had this call with, Forrester. A couple weeks ago, they were doing research. And they said, hey, we just want to, like, feel out, like, is this… is this a real role? Is it going to be around? Is it going to be, like, a prompt engineer that disappears in, you know, 3 months? Or is this a role forever?

In other words, is a human going to be the one who’s doing this, or is the tech… and the other side of the coin is… or is the technology getting so advanced, almost like an app on our iPhone, that the technology just magically… like, works with itself, and so you don’t actually have to have these, like, go-to-market engineers that are architecting AI tech stacks, because the technology is so intuitive. Anyway, it was this fascinating conversation, thinking about what’s the role of a go-to-market engineer? Are they about helping the marketing and sales teams figure out Automation and agents that make their internal operations smoother, or are they the ones who are re-architecting our go-to-market tech stack?

To be a much more AI-first, MCP-driven, build-your-own kind of context layers, intelligence layers that power your whole ecosystem. Just interesting to think about, like, what that focus is. I feel like we all started internal with efficiency, but now we’re thinking this needs to go external as well.

Ray Wizbowski:

Well, yeah, and I think part of that is, again, from a go-to-market engineering standpoint, there’s a… there’s still an important part of translating what is actually happening into meaningful insights that aren’t… I mean, I love when I throw stuff in the clod, and I say, create a deck, and it comes back with these like, your pipeline is falling off a cliff. I’m like, come on, Claude. It’s not. It decreased by, you know, half a percent. You know, and we made it up over here. And it’s like, it’s… I still think there’s a… for a period of time, and I don’t know if that goes away, I don’t think it does.

What goes away is the point solutions. Like, we used to have somebody who only did SEO. that job’s gone. I mean, that’s all, you know, Agentic processes now. The go-to-market engineer who can look at it from, you know. you know, the first point of connection through conversion, through, you know, retention, and being able to extract insights, the pipeline insights, all the things that are going on, and making sure that while there’s connective pieces, that those connective pieces are actually telling you a story that give you insight, that actually allow you as a CMO or a senior leader within a marketing organization to make intelligent decisions and then push that back into the AI or Agentic processes, I think that’s that role right now.

And will that role change? It will absolutely change. Will it go away? I think it’ll be around for a while.

Heidi Darling:

Yeah, I think it also depends on, industry. Like, one of the things that we are… are hampered by, but still have found creative solutions around, is just PII. And, and… so you can’t connect everything. You… you do have to have some people connecting different motions,

Ray Wizbowski:

For sure.

Kelly Hopping:

Yeah, we’ve seen even that adoption on that sort of, like. AI adoption continuum. You can see it, even so clearly almost by industry. You’re talking about the PII side, Heidi, like, the highly regulated, the FinServ, the health cares, the government, like, those sides are going to be much different on the AI adoption curve than the others for all those reasons. So figuring out, like, if they’re gonna get there, if we’re always gonna have this kind of, like, two extreme ends of the continuum because of that. Yeah. Well, lots of good stuff. I like, I like hearing that.

What, what do you think that, two years from now, that’s gonna look like? So we talked about structure today. I think we’ve talked about the go-to-market engineers not going away for a while. So what does a winning B2B marketing team look like in the future? And what does a losing one look like? What, what, like, basically, what, structure or hiring habits do you think we have to let go in this new world, and which ones do we have to adopt?

Laurie Pratt Ehrbar:

I think for.

Ray Wizbowski:

No, go ahead, Laurie.

Laurie Pratt Ehrbar:

Yeah, I was gonna say, I think, like, for a winning team, they’re gonna always sit closer to the strategy, they’re gonna have well-trained, tested agents that they know are. They’re gonna be doing a lot more high-frequency execution underneath, right, with those agents. Humans are going to stay closest to the brand, the budget, right? We haven’t talked much about budget, but still need a human to stay close to the budget. And the customer relationships, because we said before, that’s the moat, right? The loyalty is the moat, so they’re going to have to stay close to the customer relationships.

The losing side of this, in my opinion, is automated content production, right? Like. you know, the fastest to show activity, right, and just more is always less. So I think that’ll be the losing side, the people who don’t use all this more structurally. I think that’ll be the two sides of the coin.

Richard Dumas:

Thank you.

Ray Wizbowski:

I think the winning side of… the winning side… sorry about that, Richard. I think the winning side of things are… are really, like you said, you know, it’s… it’s the gluing all the pieces together, and I’m gonna take a different tact on the losing side. I agree with you on content 100%. I think the losing side are point solutions that are trying to sell a specific thing, because we, you know, every AI tool that we’re… we’re looking to implement now, I won’t sign anything more than one year contract, and it’s… I tell them, you’re… this is a trial, and this… this is… we’re gonna… we’re gonna test you out for a year and see if you actually work within our process, within our flow.

But you have to prove your value, because after a year, the world will have changed that much more, and you have to be integrated into whatever’s coming next. Your point solution may not may not be relevant anymore.

Richard Dumas:

Yeah, I think I was gonna use a… I’m sorry, sorry, Kelly’s metaphor a little bit there, which is, of the tree, which is, I think, in the future, winning teams are gonna have a smaller number of real strategic thinkers.

Kelly Hopping:

Hmm.

Richard Dumas:

Because, I mean, what AI has allowed me to do, personally, and I think most is not all the mundane, routine things that I had to do before. I mean, think about how easy it is to do research, or to build an audience, or something like that. I mean, if you don’t have to spend your time on that, you can actually think about solving strategic problems, and I don’t know if that’s a seniority issue, or it’s just someone who has the mindset of thinking about the full solution and what it takes to be successful for your company or your product.

And then it’s like, oh yeah, I’ve got all these tools, you know, just like an AI agent that I can use to do this, but what am I actually trying to accomplish?

Kelly Hopping:

Yeah.

Heidi Darling:

I think maybe, to pull a thread on that, if I look at macro trends and what’s happening… I’m using the word macro a lot for some reason, apparently… is we’ve got this, what they’re calling silver tsunami, like, mass exodus of wisdom, knowledge, leaving the workforce, right? I think it was Deloitte that said, in the next 4 years, 76% of our, kind of, knowledge and wisdom is going to leave the workforce. And then we’ve got Either juniors who don’t have any of the experience, wisdom, and knowledge coming in, expected to be the human in the loop with Agentic, Agentic systems without the wisdom and kind of the context to run it.

So I think the winning side is if you’re able to really deeply understand what your company’s unique IP is. As it relates to their workforce and to their customers, and then the losing side is, if you don’t capture that, and you’re also not training and up-leveling, kind of, the next generation, then you’re just going to be stuck with this exodus of wisdom and agents, really.

Kelly Hopping:

Yeah.

Ray Wizbowski:

I had a conversation with my, with our COO, and she was saying, because we were talking about workforce evolution, and she goes, how does somebody become you? And, you know, I have two kids, one just graduated from college, one’s about to graduate this coming year, and the question is, how do they become me? Well, they don’t want to be me, but that’s fine. But my kids don’t, but I mean, really, as we’re looking at people coming into the workforce. I think you’re absolutely right.

There’s… there’s gonna be a… there’s gonna be a knowledge gap that, you know, as people leave, and you’re not… Going through the… and we’ve all went through the process of, you know, learning how to build a campaign and, you know, all the pains of doing all these things, which gives us perspective and the ability to look at things differently than somebody who’s just coming out of university. And so there’s got to be a process of helping them understand what we understand. To grow them up, and maybe grow them up faster in order to be able to, you know, have a career path forward.

Kelly Hopping:

Yeah, yeah, it’s fascinating. I have a freshman in college, and he just told me he was going to drop one of his classes and asked me if I would go through the catalog and find another one, because we’re trying to figure out… because he’s like, I don’t know what I want to major in, and I’m like, I can’t even tell you what to major in, these days, because, you know, we swung, again, that pendulum, right? We were all on the STEM thing 5 years ago, 10 years ago, like, you want to have a great job security, it’s all STEM.

Now it’s kind of like, well, no, you can actually automate a lot of that. stuff now? Where do you swing to the other side, and then what’s the job in between? We talk about trade school a lot, they haven’t been, but I’m always like, HVAC. HVAC is where it’s at.

Ray Wizbowski:

Haha.

Heidi Darling:

I literally just said I’m gonna buy an HVAC company as my backup man. Yes.

Richard Dumas:

Either that, or it’s actually the return of the liberal arts degree, right? Yeah. Because it teaches you how to think.

Ray Wizbowski:

Yeah, yeah, for sure.

Kelly Hopping:

So that’s a, yeah, my school’s at a liberal… my son’s at a liberal arts school, and I’m like, I constantly am like, I’m like, this is so weird to me, because I was an engineering degree, I mean, engineering major, so I’m like, I don’t even understand this, but now that I’ve, like, learned digging through his stuff, I’m like. This is actually probably, like, gonna change the world, because, yeah, being able to think and strategize and all that stuff, but to the point, how do you get From 22 to 40, or whatever, like, to, to, like, how do you get the experiences to be able to get to that, to use that strategy brain?

Kelly Hopping:

So any questions? Okay, so one thing really quick before… we’re gonna go into rapid fire here in just a minute, but one last question, is sort of the endless ROI question on AI. I’ve heard a couple of y’all mention about, like, three times productivity, or it saved us this amount, or our pipeline is twice as efficient, or whatever. How are you making, like, an ROI case, for, like, for your CFO, for your budget, to be able to invest in AI, to get your team, you know, Claude Code licenses, or to be able to… or whatever other tools that you need.

Where are you building that case from?

Laurie Pratt Ehrbar:

I’ve been building it from, like you had mentioned before, about using influence, like, I believe strongly in influence, but adding to that velocity. So, showing that instead of just trying to prove AI investment, like, source to deal, right? Showing that the accounts touched by the new, like, Agentic layer move faster and engage more of that buying committee, that’s been the one thing that’s resonating, with the CFO.

Kelly Hopping:

I love that.

Ray Wizbowski:

For us, it’s because we’re SMB, and some of our customers are really hard to find. You know, they’re not… there is no buying committee. The buying committee is the owner and his wife. And, you know, they’re… they’re… they’re debating whether they send their kid to college or buy a new machine. I mean, and sometimes that’s a real… that is a real conversation. And so… For us, the intent signals that I was talking about earlier, we have a much more robust way of seeing where people are engaging in a broader audience, and it helps us build our audiences.

And so being able to say, look, we weren’t able to find this kind of fire before that now is coming into our pipeline because we have invested in AI tools. Now, there’s a whole bunch of gray areas as well that we, you know, we… are able to put some dots together and say, look, this is, you know, in aggregate, we’re better because we’re doing these things, but there’s still a lot of work underpinning it that says, you know. there’s activity that are going on that I can’t quantify, but I know it’s getting better, and I tie those dots together to come up with a story that helps me get the funding I need.

Kelly Hopping:

Yeah, I love that. That’s great, I think we’re all gonna have very different scorecards, a year from now than we do today, which is good. Okay, we’re gonna do some rapid fire. I’ll throw some out, we can… you can kind of chime in, but sort of think of your answers in kind of, like, one sentence. Responses. What’s the one AI tool that you’ve actually gotten, speaking of ROI, that you’ve actually gotten ROI from?

Richard Dumas:

Claude plus connectors.

Heidi Darling:

Especially cloud design.

Laurie Pratt Ehrbar:

I agree, yeah. Well, distilling, like, customer call notes, telemetry data, like, using it for, like, truly, like, targeting. Yes, 100%.

Richard Dumas:

Aside from our own solution, that is, of course.

Kelly Hopping:

Yeah, feel free to shameless plug your own product right now.

Heidi Darling:

Last one, if you need the context,

Ray Wizbowski:

The, the thing… And… and sorry to bring this up again, the one thing that I have the most tangible ROI on is our… is our new, chatbot, you know, our… Our qualified chatbot.

Kelly Hopping:

Nope. What’s the most overhyped AI use case in marketing right now?

Laurie Pratt Ehrbar:

using Claude to make fully AI-generated thought leadership content.

Richard Dumas:

Automating things that annoyed people in the first place.

Kelly Hopping:

But now we know what to start.

Heidi Darling:

More bad stuff.

Ray Wizbowski:

We have… so, video production, you know, we do… we’ve… we’ve experimented with a lot of different tools, and, you know, the six-finger person, or the, you know, so it’s… it takes… it’s faster, but it still… it still takes some time to make sure that you’re looking at it going, wait a minute, that… it looks totally, you know, totally fake, so…

Kelly Hopping:

Yeah, for sure. What is, what’s one metric that you’ve stopped caring about now, that maybe you cared a lot about a year ago, 2 years ago, 5 years ago?

Richard Dumas:

booking meetings. I think that we’ve been so obsessed just on that as a metric. Like, how many meetings have I booked this week? And I think that’s completely changing because the dynamic of a discovery call or early meeting is changing, so we probably get fewer of them, because our buyers are more educated once we do engage with them.

Kelly Hopping:

Yeah.

Ray Wizbowski:

This might be controversial, but the MQL for us is a lot less important. It’s what actually turns into an opportunity, so the variation, especially across my business, of how many MQLs I need to actually get to real opportunities. I just… we hardly… we look at the number, but it’s not one that… we used to… we used to obsess about that number.

Kelly Hopping:

Yeah.

Heidi Darling:

I mean, it’s changed forms, but definitely search ranking.

Ray Wizbowski:

Yeah.

Kelly Hopping:

Yeah, I would say even, like, paid search click-throughs, like, I just realized the junk that comes through there is not the quality of what we want, or the ICP, as tight as we’d like to be, so I think that’s a… Now, the ones… I can pull it all the way through to ROAS or something, and actually see the ones that make it to pipeline. That I care about, but the actual click-through’s probably less. What do you think CMOs are doing wrong right now with AI adoption? Either internal… either in the way they take it to market, or the way they’re operating in-house.

Laurie Pratt Ehrbar:

Buying tools before they fix the workflow. So, if you use AI, like, you’re just gonna make your process, a broken process, broken faster, so… I think that’s… that’s the big one.

Ray Wizbowski:

I think… I think to that point, it’s not just… it’s the workflows, but it’s also the underlying data. Because if you’ve got terrible data, your AI is going to produce terrible results.

Kelly Hopping:

Yeah. Just does it faster. Yep.

Ray Wizbowski:

More mess.

Laurie Pratt Ehrbar:

faster.

Kelly Hopping:

Yeah, exactly.

Ray Wizbowski:

Yeah.

Kelly Hopping:

What is one… and this might be controversial, hopefully we don’t offend anyone listening, what’s the first Piece of tech you would cut from your tech stack right now, if you had to.

Laurie Pratt Ehrbar:

I don’t think I’d name one, I would just say any piece of tech that you cannot tell me the ROI on it in one sentence. Like, what we’re gonna get from it. then I would say, I don’t need it.

Ray Wizbowski:

I can tell you what we just cut, and I won’t… again, I won’t use the names, but it… we… we have a whole… we had a couple analytics tools, that would help us determine ROAS and other metrics, and those data tools are, you know, we’ve just built them ourselves now.

Kelly Hopping:

Yeah.

Heidi Darling:

I think this meeting is a really good example of, the way that AI has democratized things and stripped down some things to kind of the core essential to where maybe you don’t need a fancy webinar platform, Zoom is fine. And so… That would be an example of things that were cut.

Kelly Hopping:

I think Zoom would agree with that. Anybody else? Sales team. How’s your sales team, adopting AI? Are they… are they… are they adopting it? Are they threatened by it? Are they eliminating it? Where does it… where does sales sit on that journey?

Laurie Pratt Ehrbar:

I probably shouldn’t answer just because we’re sales loft, and obviously they use our product. So, if they didn’t use our product, they’d be in big trouble. So, so I will stay out of that one. But they are excited about it, they do use it.

Richard Dumas:

I think they’re excited about it, maybe depends where you are in sales. Maybe some of the early-stage SDRs may be a little bit more threatened by it.

Kelly Hopping:

And

Richard Dumas:

And then I think maybe for enterprise sales reps, they’re excited about getting more qualified discussions and meetings and everything, but they still want the focus on the fact that people buy from people. Right, and not to forget that, hey, I can automate the whole sales process, and you don’t really need an enterprise sales rep to have a relationship with a customer, so…

Kelly Hopping:

Yeah.

Ray Wizbowski:

I think for us, we… because we have, we sell into some very specific… like, we sell to hardware stores, and we sell to lumber yards, and we sell to construction, you know, companies. And manufacturers, and… and so, the… We have people who’ve been selling that… those solutions for 20, 30 years, and… and they don’t want… most of the AI that we want to give them. So we automate it in the background, you know, we listen to their calls, we populate the, you know, Salesforce with the right MedPick data and all of those sorts of things.

So we… that makes their life easier, because they weren’t doing it anyways. It just helps us on the information gathering side. Some of the new… so I would say our BDRs actually are super excited about the AI tools that we have, because we listen to all their calls, it has real-time coaching, they get… they get insights as they’re, you know, in a call. And that’s really improved their connections, or their conversion rates. So I think they’re the ones that are… digital native, they’re okay with AI, they’re using AI in their everyday lives, so for us, we have this spectrum of, you know, 250 people.

There’s some on one side of the spectrum that don’t… that doesn’t like it, so we try to build it in underneath. And for the rest, we’re, you know, we’re trying to mandate those processes, but it’s… it’s a… the adoptions, it varies across the… across the organization.

Kelly Hopping:

Yeah. Yeah, I’ll give a shameless plug really quick for BDRs. I’m such a huge fan, of, like, my BDR team drives, like I said, 60, I think I said earlier, 64% of our pipeline, which is a lot,

Ray Wizbowski:

Amazing.

Kelly Hopping:

Marketing, marketing source, of course, but it’s… but I’m not seeing the need to remove the humans, but augmenting them to be more efficient. Is what’s been so helpful. Giving them contact enrichment, giving them auto-dialing, giving them auto-follow-ups on inbounds, giving them, automated sequences that go on a list load after an event, all those things that kind of just help them be able to focus on both outbound and inbound, allows them to get better and more effective at their at their outbounding, and their follow-ups, and their sequences, and their technical product knowledge to be able to have a discovery call, or at least an initial call, has been so much more valuable, and so we’ve just seen that productivity tick up and up and up, not because we’re replacing them with AI, but because we’re sort of filling in the gaps with AI to allow them to focus.

Ray Wizbowski:

Well, I agree.

Kelly Hopping:

Yeah. Very cool. Okay, let’s see, we got… One more question, let’s see… Let me look at my thoughts here. So, we’ll sort of… let’s close out the way we started, and with a little bit more context now that we’ve had a conversation for… for 57 minutes. But finish the sentence again, and see if it’s any different. The new marketing moat is what?

Laurie Pratt Ehrbar:

I’d say loyalty, because proof can’t be faked. So…

Kelly Hopping:

Okay.

Ray Wizbowski:

I would… Okay, with my first answer, which is context, because industry knowledge and depth of insight into your customer can’t be faked either.

Heidi Darling:

Yeah. Let’s say…

Ray Wizbowski:

Yeah.

Heidi Darling:

nuanced human context, and at scale, too. Like, really getting the full… At scale. …full spectrum.

Richard Dumas:

Yeah, I think the new marketing moat is being able to answer and solve customer problems.

Kelly Hopping:

Awesome. Well, and I think all those are tied, very tightly with, with brand, understanding our buyers, understanding who they are, inspiring them that they become loyalists and advocates. And it goes back to, actually, authentic human relationships at the end of the day. Even if we’re using contextual intelligence to find them, to know about them, to understand who they are. It still is, it still comes down to the humans, which is ironic, given, the conversation we’re all having about AI. So, so I’ve loved this conversation. Thank you all for, for the time. I love the different perspectives from the different sized companies, different, different industries.

Heidi and I are gonna go by an HVAC company after this call, but the rest of us. I hope Hopefully all have.

Ray Wizbowski:

And we have an ERP solution to sell you for your HVAC company.

Kelly Hopping:

Perfect!

Heidi Darling:

Perfect.

Kelly Hopping:

Look at all this business being made here. I really enjoyed the conversation.

Julia Nimchinski:

Thanks, everybody.

Heidi Darling:

Kelly, thanks everyone.

Julia Nimchinski:

Thank you so much, Kelly.

Kelly Hopping:

Thanks, Julia.

Julia Nimchinski:

Thanks, everyone.