Scott Brinker, VP of Platform Ecosystem at Hubspot and Editor at chiefmartec, shares why marketers should be cautious with AI agents and embrace the growing martech landscape as a sign of innovation, not overload.
Scott Brinker, VP of Platform Ecosystem at Hubspot and Editor at chiefmartec, joins the show to break down what not to do with AI agents, and why the ever-expanding martech landscape is actually a win for marketers.
Scott and Sara also dig into the transition from siloed databases to more integrated data warehouses, the growing role of AI in enhancing customer service and content production, and the importance of maintaining human oversight.
Scott:
“I understand the complex landscape is challenging to navigate but if you pull back the cover of the capitalistic engine that's at work, it is this incredible span of entrepreneurship that is focused on empowering marketers. That is a gift to marketing, not a problem.”
Links & Resources:
Timestamps:
00:35 - The Evolution of Data in MarTech
02:15 - AI in Customer Service and Content Production
04:10 - The Role of Humans in AI-Driven Marketing
06:00 - SEO, AI Agents, and Content Validation
10:00 - Marketing Technology Landscape Report
0:00:00.2 Sara Faatz: I'm Sara Faatz, and I lead community in awareness at Progress. And this is 10 Minute Martech.
0:00:05.2 Scott Brinker: Somebody said, like in the old marketing, like the worst thing you could do is spray and pray. Here in the age of AI, like what you don't wanna do is deploy and annoy all these agents.
0:00:18.8 Sara Faatz: That's Scott Brinker, vice President of Platform Ecosystem at Hubspot and editor at chiefmartech. Let's get started. So, Scott, other than AI, what is your burning martech idea, concern, or question right now? What keeps you up at night?
0:00:40.4 Scott Brinker: Related to AI but has been a growing revolution in martech here for a while now is changes in the data layer, and the short version is right. It used to be that all these different martech apps had their own little siloed database and integrating data between things was just an absolute nightmare. Over the past five years plus we've seen the shift towards more and more cloud data warehouses, lake houses and whatnot that are just making it, I'm not gonna say easy, but easier for all these different applications across the business to share data. And I think that whole direction is both super interesting, but also has a lot of value to unlock.
0:01:18.0 Sara Faatz: So what on the AI side, what does keep you up at night there?
0:01:23.3 Scott Brinker: You know what they say, the spiderman thing, with great power comes great responsibility and there's definitely a segment of the marketing world out there that's like, oh, great, well, we've got these agents, we can just set them loose to like email people and call people and text people, and we don't even have to have humans doing it. I think the reaction to that is that's one might anticipate not gonna be joy and delight on behalf of the buyers. And so I hope we sort of get past that more in mature phase of what's possible with agents quickly.
0:01:56.4 Sara Faatz: Yeah, you definitely don't want AI agents running amuck causing shenanigans left, right, and center.
0:02:01.4 Scott Brinker: Somebody said, like, in the old marketing, like the worst thing you could do is spray and pray. Here in the age of AI, like what you don't wanna do is deploy and annoy all these agents.
0:02:13.0 Sara Faatz: Yes. Yeah, 100%. Are you seeing a good practical application of AI today based, taking that say just deploy and annoy what is working well right now from an AI perspective?
0:02:25.2 Scott Brinker: Well, there's a lot. I mean, on the customer facing side, I think the maturing of customer service agents so that when someone comes to your website and they're in a mobile app, and they need help because generative AI has really allowed a lot of these customer service agents see a lot smarter. If you're plugging into the right data and the right content, I think that's a real benefit for customers. Again, you always want to give people the option to get to human if they don't feel their needs are being served by the agent. But boy, if the agent can take care of a resolution for me in a matter of like a minute, that's fantastic. I think on the backend, for marketers, of course there's a lot of discussion around leveraging generative AI in content production. And I do think it has a role to play in like accelerating both not only the speed of production and reducing the cost of production, but it's given more people a wider range of this ability to take ideas and even like prototype them out that previously, oh, now we're gonna need to get graphic designers and we're gonna need to get a programmer, to like go from the idea of like, oh, I have an idea. Can I even bring this to life in a way to like see, is this even credible? I think that is incredibly powerful and we're just barely beginning to tap that potential.
0:03:43.5 Sara Faatz: Yeah, I would agree. I think though, kind of like, don't deploy and annoy, you also have to make sure you're using the tool correctly because the last thing you want is to create digital waste. You wanna make sure you're using it in a way that's providing value back to the business, value back to your customers and creating content that's interesting, engaging, accurate, just letting AI run off and do its own thing without that human in the loop, without that human interaction, I think is... Where do you think the role of the human is in all of these things that you talked about, whether you're talking about the data layer, when you're talking about AI or just martech in general?
0:04:19.1 Scott Brinker: You know it's funny, I did a map of this of like where do we spend calories in marketing? And I roughly put it into three buckets. There's strategy and creative, which has always been the pinnacle of like, why do people go into marketing? No, because it was strategy and creative. But the truth is, for a long time the amount of calories we could spend really on the strategy and creative was very limited. Most of our calories went into a bucket I would just broadly call production and analysis. Like, it took a lot of work to be able to bring those strategy and creative ideas to life. And then there's a third bucket, where I live and love, the marketing operations and marketing martech capabilities that support that. And to me, what's very exciting about AI is I think it has a lot of value to bring in reducing the cost and the time associated with that production analysis stage.
0:05:08.7 Scott Brinker: And if we turn around and we use that opportunity to lean in to the strategy and creative to run more experiments to like try and bold new ideas, and again, like, it's not like every idea is a winner. Not everyone works, but one of the things about marketing is it's very hard to predict what's gonna work in advance. But if you can accelerate this loop of being able to experiment and try new ideas and get that feedback, I think that's actually an incredible competitive advantage. And so I'm hoping, yes, marketing as a whole is really gonna take advantage of this moment to say like, wow, could we take strategy and creative to a whole new level with this new tool set?
0:05:47.0 Sara Faatz: When we think about AI today and the output, there are concerns that people have about hallucinations, there are concerns that people have about what kind of information it's providing and how do we see ourselves surfacing within that world. Can you talk a little bit about your views on SEO and AI agents and AI search today and how that might be changing the way we think about that content and that production that you're talking about and the experiments that you're talking about people taking?
0:06:19.2 Scott Brinker: Yeah. Well, what's interesting about this is you've got AI on both sides of the equation. You've got AI on the seller side that can be used to generate more content. It can be used to like dynamically generate content. And again, I totally hear you on, yeah, be careful of just letting that thing run loose, but with human supervision and connecting more and more of our data into these engines, like yeah, you can actually produce some really useful content and information that's available. There's the other side of it, which is yeah, the buyer's side, and typically that's what Google appeared to be. But it's interesting because over time, I don't know, you know the whole meme of like, well, I read it on the internet, therefore it must be true, is I think society sort of matured to a level of the recognition that like, okay, yes, actually it turns out anyone can publish anything.
0:07:09.8 Scott Brinker: AIs aren't the only things that hallucinate. Humans seem to have a miraculous ability to hallucinate all sorts of things. But I think one of the things that excites me potentially about these AI engines on the buyer's side is like the Google algorithm at the end of the day, it was, I don't know the full algorithm, but the core algorithm here was like, oh, well, let's just keep track of where links are going, and the more sources that get more links, we'll treat that as authority. It was a very useful algorithm for what it was. But from my perspective it never went deeper into like, okay, can I actually cross check what I'm seeing here and does it hold true elsewhere? And one of the things that's really fascinating with some of this current generation of these AI assistants and those that are leaning into these deep research, is it can go and not just read one source, but it can be going and looking at multiple sources and cross checking and cross validating. And so this is not to say that there aren't still risks in all of these generative AI technologies associated with hallucination, but I think this ability for the reasoning component of these AI engines to actually start to provide a bit of a crosscheck on the validity of information that's out there on the web might actually, I think it'll actually be more of a net benefit in helping us find what is real and what is not.
0:08:30.0 Sara Faatz: Do you think it's there? Do you think the promise of what that can offer us is here today? Or do you think that there's still more that we'll be seeing and evolving from from an agent perspective?
0:08:41.9 Scott Brinker: The height and AI continues to far exceed the actual reality, but I think one of the things that's really different in this current environment compared to anything I've seen before in my career, is the time it takes for reality to close the distance with the hype, it's just collapsing a lot faster than ever before. So it's like, you'll see these things that get produced as like, oh yeah, this is hype, I can't really do that. I tried it, didn't work that well, and then a few months go by and you're like, oh, wow, actually, yeah, no, this works pretty, pretty well now. And so I think on one hand, like you have to maintain that skepticism of like, okay, what's actually real that I can rely on? Particularly if you're trying to look at implementing things for your business or things that are gonna have an impact on customers, I think skepticism there is a healthy trait, but boy, I don't think you wanna fall over the line into like this pessimistic view of like, nah, nah, nothing to see here. It's all like, there's some like article that's passing around on medium or something like this. Like, oh, you're just a fool if you believe in this AI agent craziness. I'm like, well, maybe you're a fool today, but I'm not sure you're gonna be a full six months or a year from now. And so keep an open mind.
0:09:53.5 Sara Faatz: So let me ask you one more question, actually two more questions. What is your martech hot take right now?
0:10:00.4 Scott Brinker: We're coming to this time of the year when we publish the martech landscape, and we've been doing this now for my goodness, like 14 years, and every year it grows and every year people say, oh, that's a terrible thing this keeps growing and it should just collapse. Like, won't it please consolidate? Please, Lord, make it consolidate, make the bad chart go away. But my hot take is, this is actually a great thing. All these companies out here who are building these products, they are all competing with each other to be able to serve you as the marketer. And I understand large, complex landscape, it's challenging to navigate, but if you really pull back the cover of the capitalistic engine that's at work there, it is like this incredible span of entrepreneurship that is all focused on empowering marketers. And so I don't know, I take a much more optimistic view that, yeah, for all the new companies that keep appearing in the martech landscape every year, that is a gift to marketing, not a problem.
0:11:01.8 Sara Faatz: Awesome. That's a great, great response. I love that. One last question. Who do you follow for inspiration or information?
0:11:08.9 Scott Brinker: So many great people. I would say on the AI side, one of my favorite dev remains, Ethan Molik, the University of Warden professor who just every week covering more things and helping involved in a lot of academic research or around okay, what's actually really doable with these engines? Great fan of his.
0:11:28.3 Sara Faatz: Well, thank you so much for your time today, Scott, always it's wonderful to talk to you and I really appreciate the discussion.
0:11:35.2 Scott Brinker: Thanks so much for having me, Sara.
0:11:38.4 Sara Faatz: Listeners, thanks for tuning in. To never miss an episode make sure you like and subscribe wherever you get your podcasts. Until next time, I'm Sara Faatz, and this is 10 Minute Martech.