Loading summary
Transcript112 lines
- [00:00]
Ty Degrange
Foreign. Welcome to another edition of the Always Be Testing podcast. With your host, Ty degrange, get a guided tour of the world of growth, performance marketing, customer acquisition, paid media and affiliate marketing. We talk with industry experts and discuss experiments and their learnings and growth markets, marketing and life. Time to nerd out. Check your biases at the door and have some fun talking about data driven growth and lessons learned.
- [00:36]
Alexis Contos
Thank you everybody. We're going to get started with our next panel which is about leveraging AI to drive better results. You heard a little bit about AI and the role of storytelling in the last panel. Now we're going to talk about leveraging AI to drive more measurable results in marketing. So with that, I would like to call up Ty, Corey, Jennifer and Alexis to the stage. Hi, I'm Alexis Contos. I'm the CMO of nea. We're a tier one venture firm based in Silicon Valley in New York. We invest at the intersection of technology and healthcare. You know, some of our companies are leading the AI revolution. We led the series A in Perplexity and we invest in all the infrastructure down to the foundation models for a lot of these AI companies. So seeing amazing, incredible companies come across our desk every day. And as a marketer and leading marketing for the firm, I play around with a lot of these companies and the tools that they have.
- [01:33]
Ty Degrange
I'm Corey Trefletti. I like her job. That sounds awesome. I'm CMO for a company called Rembrandt. We use AI. We have a process we call Regenerative Fusion which we use to insert brands in post production into creator content, CTV content, et cetera, to create realistic in scene media and virtual product placement. And we've been doing it for about a year and a half now. I've been a CMO a few times. We've sold a couple different companies with the same team and I'm happy to be here as well.
- [02:01]
Jennifer Barkley
Hi everyone, I'm Jennifer Barkley. I'm the VP of Commercial Marketing for Visa. On the acceptance solutions side, most of my marketing career has been going into legacy organizations like Sabre, IBM, Visa, Orgs that have been around for a really long time and helping them rethink and reimagine how they go to market. So I'm on the acceptance solutions side, meaning we are on the card side. This is how the technology and the infrastructure runs that helps merchants and banks and financial institutions and fintechs accept payments. So good to be here with you all.
- [02:35]
Corey Trefletti
Amazing. I'm your esteemed moderator, Ty degrange. I'm the CEO of Round Barn Labs. We're a performance marketing agency. We've lean in heavily to affiliate and influencer marketing and we've been fortunate to work with some great tech brands over the years from Nextdoor to Metaquest to Grant and Atlassian. So tech is definitely something we are near and dear for us. And I have the pleasure to host my always be Testing pod right here at the Capital Factory and so it's fun to see familiar faces and the Capital Factory team roll out the red carpet for everybody that's in the audience tonight. So hopefully there's some good insights for everybody to take away. And speaking of insights, let's do a quick poll. Who here is a startup Fortune 500 bigger company, don't be shy. Maybe mid tier in between. And on the AI side, who is adapting and utilizing AI tools right now? Amazing. And who is utilizing them in their marketing? Love it. So wanted to set the stage by talking just some of the observations we've had as a group and talking through some of the things that are we're witnessing now before we jump into the questions of the panels. So HubSpot's saying 80% of startups using it. We're seeing that here in the panel. I don't know if you've seen Kieran Flanagan on HubSpot, an amazing SVP in terms of his knowledge of the space and how it's impacting content marketing. It's starting to certainly take over his workflow. He's using it a ton on my part. I observed Intellimize they were acquired recently. They're an AB testing tool fully using AI into their workflows. So you can a B test on any website. I think many of us know and are well aware of what's happening with Meta and Google in their ad targeting and using LLM multiple language models, especially with performance Max. We deal with that a lot in our in performance marketing. So it's certainly here on the creative side there's Pencil mid journey they're actually using full on just straight up using an ads. We're actually seeing it start to happen with TV ads AI where connected TV is starting to get ads fully utilized and created with AI. So some observations we're seeing. We talked about this a lot. I just wanted to set that stage and it sounds like this group is starting to delve into that. So hopefully we'll have some good questions as we come into the end here. Let's kick it off here. Where are we at in the cycle of AI in terms of the kind of Gartner hype cycle we've talked about. Where are we at currently, in your view?
- [05:00]
Alexis Contos
I think we're just starting to reach the peak. I think we are probably. Maybe the reason all of you are in this room is to just try and deduce some of this information into things that can be a little bit more actionable. I myself feel incredibly overwhelmed with all the tools that are coming our way, and I'm trying as many as I can. And I think evaluating them is personal perspective as well as trusting your community and seeing what works and what doesn't. Because the thing that we. The thing that we can't risk is our time and time and efficiency. And so I think we're just starting to reach the peak and probably within the next like 12 to 18 months, we'll start to see the dip of people just being exhausted, trying to figure out what works best and potentially some consolidation in the market.
- [05:47]
Jennifer Barkley
I'm already exhausted. Just going to LinkedIn exhausts me in terms of AI. So I think it's. It's quite personal, honestly. I'm just depending on, you know, what size organization you're in, what you're using it for, how bombarded you may be. So it's just case in point, my experience and, you know, Alexis would be completely different. And I think that it's a means to an end. And so to the extent that you've already identified what it is that you need to do better right around your core in your marketing programs, if you have had that clear roadmap of these are the gaps that I have and I know what I need now, you are leaps and bounds ahead, right? Versus there's AI, how, and you're all over the place and trying to figure out how you're going to. That would be going backwards. I think that that's a much more exhausting place to be. Not because of Alexa. She has a different job. So for a different reason, but for a larger organization looking and assessing the different tools and, you know, just individually how. How comfortable we are. I look across my own team, you know, 50, 60 of us and all around the world, and looking at how some people are adopting it and they're much more comfortable with it and then some aren't. But overall, I think that you'll find some places where, you know, slip of disillusionment and then we're gonna. Some people have already found a. They're monetizing it and they're good to go. So I think it's quite a personal way to plot it really.
- [07:11]
Ty Degrange
Yeah. And I'm probably sitting between both of you, literally and figuratively, because I think it depends who you're talking to. If you talk to people in the media who write, I get to write for media posts and I get to write for other people too. But the people who write about it are already in the trough of disillusionment. They're saying, well, this company's complete out in front of their skis and everything. I think that individuals, we're at the stage where it's either a habit or it's not. And I think that's a really big note to take, to take into consideration. Because if it's become a habit, like the panel that was here before us, those folks are using it every single day and it's streamlining their processes and helping them find new ways to get stuff done in a faster clip. Now, whether it's better or not, it's up to the individual because hopefully you're still adding some of your own value into it. But I think if you're at that stage where it's a habit, then you're totally on board and you're actually constantly finding new tools. If you're at the stage where maybe you ran a big company and you're probably the exception in a big company to some of what I have seen, which is that your company is so far against allowing you to use any of this stuff that you're like, ah, forget it, I'm not going to, I'm not. I can't put it on my work computer, so I'm not going to use this stuff. I'm just going to go back to doing it the way that I've been doing it for the last however many years. So I think on an individual basis, it just depends on whether or not you've actually leaned in already and you found the benefits of these tools or not.
- [08:26]
Alexis Contos
I think there's kind of, there's multiple ways to approach this. But what I found, and I'm curious to hear from y'all, how you're approaching this on a day to day basis. There's the specific triage of finding a hack where you're like, man, I don't want to do this task anymore, so I'm going to find a hack to solve that. And I've done that multiple times. Then there's also approaching it based on first principles like clean slate, try and approach new workflows that you've just never done before. What is this enabling that you've just never had the opportunity to do before. I come from creative brand building strategy, so I'm less on the performance side and more on the brand building side. And it's unlocked so much for us and our portfolio companies and our individual marketing team to find new ways to generate content that still hold onto the creative process. And to me that's first principles is like you still hold onto your strategic values of what you need to deliver on, but you're finding new ways to approach it in ways that you never have before. So I'll give you an example. We work with a company called dxm. They're an early stage seed AI like software service model where you feed your brand guidelines into the platform and so anything you generate onto the platform. So it's the kind of a hyped version of midjourney, because you can still generate, but it's generating based on your brand guidelines. It has completely unlocked the type of content that we're able to generate. Either we would have to hire a designer, they haven't gotten into animation. But that time and value that you're getting from a designer, to be able to have it do on branded content and have 20 different versions in a matter of minutes to me is first principles. Like we just didn't have that opportunity before. So I think that's an important thing to keep in mind is like find your hacks, but also try and approach things from like clean slate. What are the new things you want to learn?
- [10:11]
Corey Trefletti
Team will want to test that. So we need a QR code for everybody to go and start testing it.
- [10:16]
Alexis Contos
Yes.
- [10:16]
Corey Trefletti
I just want to call out to what you shared earlier as well, the habit formation. What a great call out for. You've got so much great startup energy in the room. Getting that habit. Right? Getting that habit flow going. I mean, that's what you need to do for your customers at this time. And so what a great analogy between getting the habit dialed in for your own work and AI, but also as a startup.
- [10:36]
Ty Degrange
Well, you know, the fact of the matter is that everybody here is human. I hope.
- [10:40]
Corey Trefletti
I hope so.
- [10:41]
Ty Degrange
Okay, so how many of you, when you have for now, how many of you, when you get in a thrust into a stressful situation and your CEO is asking you 100 questions and everybody's asking you 100 questions, you default back to what you know, which is just to start doing stuff, or maybe you're a delegator, you start delegating stuff to change that habit and say, well, maybe I'll go use this tool because it'll allow me to do something faster. I think that's one big key thing to realize that when you're in that situation, you have to force yourself to change your habit. The second component, and I only discovered this in the last week, to be brutally honest, because I've been using some of these tools. But within our company, we finally got access to Gemini. Gemini is baked into the Google suite. Oh, my God. It's so much faster and it's so much easier to be a habit now because I no longer have to take my copy, go drop it somewhere else, play with it, bring it back. It's right there on a dual screen. All of a sudden, everything is significantly faster. It changed a habit for me almost overnight, for literally everything I'm doing in G Suite. So those kinds of integrations and the fact that you can catch yourself in a moment of stress and instability and say, I'm going to try this something differently is huge.
- [11:49]
Corey Trefletti
Our team just added it like three weeks ago, and we're still gathering information to say, is this. And it's kind of a segue. We talked a little bit about what's hot, what's not, maybe better phrased for the group. You know, we talked about what's working, what's not. You have some observations on to date, like what's actually working versus what's kind of missing the mark.
- [12:08]
Jennifer Barkley
So we have a custom instance of ChatGPT and Visa that we put in place very quickly. Did it launch in November 2022? So it's just under two years old, which is kind of crazy to think about. That has layers to it. So you can query, you know, the trained database, you can query the kind of augmented database that's learning from within the firewall, if you will. And then you can even just upload your own mini document library and get in there and just, you know, do it for very specific tasks. And when you're working with subject matter or any sort of industry, it could be financial services, it could be health care, anything that has so much compliance regulation, just, you know, layers and layers. We're not selling lipstick. That's a whole other. You know, when I think about AI for Revlon or whatever, it's just so. It's like so many cool, creative things you could do with imagery and video. We go the other direction, which is how do you keep your messaging consistent and get your teams on board? And so, particularly for, you know, startups, your goal is to grow, and now you're going to bring other people into your fold. You're going to need to get Them onboarded really quickly, understanding your messaging, your marketing team, your sales team, your product team and everybody singing from the same handle. This concept of being able to have that way to organize information, to have people go and find what they need when they need it, from an organized set of it's just been absolutely phenomenal. But the habits, you know, being able, it's not just an individual habit. Then how do you make sure your entire team is using it so that folks aren't off their love proactivity? But sometimes I'll see folks who've taken the initiative and it's like you didn't have to do that. It's beautiful work. So five years ago kind of a thing and just being gently and doing the coaching to bring people on board. But I think that's an example of just being able to have that single source of truth. Not just one truth, but something to be able to at least as a jumping off point in the springboard.
- [13:59]
Alexis Contos
How many of you consider yourselves part of the creative process, part of creative making? I've actually found depending on the size and stage of your company and how big your team is, there's lots of different tools that can scale to something very specific all the way through to a platform that can help you do a lot of things like write creative briefs, generate images, write copy, write blog posts. I played around with typeface AI and was pretty impressed with the breadth of it, understanding the creative process. And I think the overarching evaluation criteria I use is to protect the creative process that is so critical to keeping brand consistency, keeping that human connection, being able to ensure that there's an emotional resonance in your work. So that's kind of the first principle that I use, which is making sure you're protecting the creative process. And places like typeface can allow you to do that, but it also allows you to have collaborative processes like Google Docs or write creative briefs for your team in a place where that could take days to do or it might take five different people to do those. So playing around with tools where it's like a one stop shop, I think work very well depending on the size of your team and how fast you need to move. And then there's other things. Like last week we were just, Cory and I were just talking about this. I was fascinated with what Runway I launched with Act 1. Did anybody see the launch of Act 1? You're basically, it's a master puppeteer. So you're putting someone in front of a camera and I'm talking right now, you know And I give a script and then it's translating that into any form of animation and dubbing my voiceover on it. I mean, if you think about the production process and the money and the time and the team required to produce animation like that, it's mind blowing. It's truly mind blowing. There was a tweet from last week where there were like five different major AI announcements in one day. And my husband and I sat down on the couch and I put the tweet up onto our TV and we just watched all the demos and we were like, holy shit. Like, things are just rapidly evolving day after day. So anyways, those are two things that, That I find very interesting. But I have. I have a whole list of stuff that I'm happy to share afterwards if anybody wants to get in touch.
- [16:12]
Ty Degrange
Beyond what you're saying, what you're saying that the habit is the one component. The other thing is that there's. We live in a world where pretty much your team is distributed in different places, and the tools that we use are collaboration tech tools. And it's great when the AI is baked into the collaboration environment because it's not just you talking to the machine or you engaging with the machine, but everybody can, and it ends up having another person to further the conversation. So in addition to habit, it's collaboration. And then the last piece, which is the ability to set your ego aside, because I will tell you that in many cases, I'll sit down with something and I'll start writing, and then I put it in there and it comes back out. I'm like, damn, that was better than what I came up with. But the fact of the matter is, I'm like, I'm okay with that. It can be better because now I'm going to build on that. Whereas I think people get worried that, oh, you know, maybe this is actually going to be better than me. It won't be. It may be more creative in a spark environment to give you something to play with, but you're still the better mind. You're still going to take that, play with it, forge that into something and go use it. And so if you can get habit and if you can get collaboration and you can get your ego out of the way, it actually becomes a fantastic tool.
- [17:15]
Alexis Contos
Well, and to build on that, it's quality in is quality out.
- [17:19]
Ty Degrange
Yeah, for sure.
- [17:20]
Jennifer Barkley
So.
- [17:20]
Alexis Contos
And I know, I think that's probably the theme of a lot of the talks you're going to hear this week is it's true, whatever you put into it, is what you're going to get out of it. So I bet it's whatever your output was, was better, but it's because of what you.
- [17:30]
Ty Degrange
What I put in was crap, right?
- [17:32]
Alexis Contos
No, it was really good prompts.
- [17:35]
Corey Trefletti
You're only as good as your prompt.
- [17:36]
Alexis Contos
Yeah, but I would love to have like a two day course on like learning prompts. Like that's where I feel like I have, you know, people are like, I have writer's block, I have prompt block. Like I know I want to write something amazing, but I don't know what to write. So then I just like start writing random shit. But it actually knows what I'm saying. And it like, it's fascinating. Like it spits something out. But quality in and quality out is really important.
- [18:00]
Jennifer Barkley
It's interesting too because you know, organizations really the way I like to think of it is they, it's very rare that one will fail just out of spontaneous combustion. Like something very major goes wrong or you know, a PR nightmare or something. They will be made or broken at the seams. And so as marketers, if you look at where we sit in the organization, yeah, I think of it as the three legged stool with product because you have the product of the solution that you're marketing and then you have sales. You know, for some, if it's, you know, a D2C company or something like that, maybe not so much if it's just E commerce, but if you have a sales organization. And so the, you know, one can only be as successful as the other is successful. And for those who are marketing, right, like marketing and sales don't always get along in something so wild. I've really never kind of understood it, but one of the things that we found, we started to have a happy accident as we started to take a new approach to our campaigns where we were having like too many leads that more than our sales team could, could follow up on. And so we were able to take, you know, our ABM tool that we were using and then procure an intelligence platform for sales to use to bring them on board to help them think in terms of account based marketing. To say we have to do this together, we have to be a single revenue driving organization. It can't just be marketing over here doing cool stuff. And then if sales is, you know, the stool starts to get very wobbly. So I think that's another way to think about collaboration is to the extent that you have a data set that's hopefully clean because crap in, crap out, not just applies to prompts. But the idea that you can just become much more enlightened across the org when you have that single view of the buyer, of the journey, and you can eliminate some of the breakage that can happen between the gaps. And so as you're thinking about what AI looks like in marketing, I would encourage you to expand it beyond just marketing. It is a relationship and we have insights and we're at the epicenter of these insights with our customers and with the market. A lot of folks think marketing does campaigns. We are the relationship conduit with the marketplace that we're trying to succeed in. And so that's what gets me the most excited, is just how we can help bring that intelligence and that scope to other parts of our organizations.
- [20:13]
Corey Trefletti
I love that. Jennifer. I was speaking with a great channel partner. He's building out a series A software, so probably similar size, you know, startup. And I was blown away by how much he's already adopted. Basically utilizing Claude, getting existing customer data, poured it in. We talked a little bit about the mechanics of that. But basically being able to run, run queries against that just make. Make your querying game so much more easily facilitated across the organization to say, okay, is this person likely to retain. Is this person potentially churning? And for the early stage, I think that thinking in those terms and just really being willing and able to go further down that adoption, I think is really exciting for that earlier stage business.
- [21:00]
Alexis Contos
Yeah, I think the customization piece is important and I think you do it a lot at Visa because you have your own augmented pieces. But I think it's becoming more mainstream where within ChatGPT you can create your own custom LLM and then you could tie a rag into it, which is where you can put all of your data that's customized and unique and that's becoming democratized. Before that was large corporations could leverage rags and have development teams that could use that. Now it's becoming more mainstream. Like I watched a TikTok on like how to build your own rag and custom LLM last night. I was like, sweet, now I know.
- [21:30]
Ty Degrange
How to do it different than what my kids watch on TikTok.
- [21:32]
Alexis Contos
Right? That's true.
- [21:34]
Corey Trefletti
My husband was like, you're doing something.
- [21:35]
Alexis Contos
Right, but do your research on that stuff because it's becoming democratized where you can build that stuff on your own. And we all have droves of data that we want to leverage. But leverage the tool to allow you to gain faster, more efficient insights than needing an entire analytics team to do it.
- [21:53]
Corey Trefletti
Any other must have tools, maybe from the Audience, maybe anything that you all have observed, that is a must try, must test.
- [22:03]
Alexis Contos
I like Notebook LLM. Oh, yeah, Notebook lm. Yeah. Yeah.
- [22:09]
Ty Degrange
I mean, I used that the other day. I used it at your discretion. I think you mentioned it in our call. And then I loaded a bunch of documents into it. And then 30 seconds later, I had a full podcast and my wife, I played it for my wife and she's like, who are these people? I'm like, I don't know. They don't exist. And it was amazing. And it's incredible. It got 80% right. The other 20%, I'm like, I gotta change that because I can't have that out there publicly. But it's fascinating.
- [22:32]
Alexis Contos
Yeah, but you can keep feeding it like you. You just keep having a conversation with it and it just continues to optimize. It's really fascinating. If you haven't tried it yet, give it a try.
- [22:41]
Corey Trefletti
Other marketing campaigns that have been live or maybe early or where it's like, wow, AI was instrumental in making that a successful marketing campaign. Is there any examples of that? I feel like, Corey, you're kind of at the epicenter, rolling out that for brands.
- [22:58]
Ty Degrange
I mean, there's definitely a lot of campaigns where, I'll say this, there's a lot of campaigns where people made mistakes using AI. I think that the, the issue with AI and advertising and marketing has been people were trying to use it and not tell the audience that they were using it.
- [23:16]
Alexis Contos
Yeah.
- [23:16]
Ty Degrange
And then they were getting called out for it immediately because it still looks somewhat fake. You can tell whether it's a piece of writing or a piece of video, if an AI did it versus a person.
- [23:25]
Alexis Contos
Yeah.
- [23:26]
Ty Degrange
Now I think is more interesting. On the flip side of that is that the creative teams that I'm aware of now, including my own, is that I will go mock something up using one of those tools and give it to them, and they'll. Then they'll go and continue to refine it and they'll actually do stuff inside of the video editing softwares or in the. In Adobe, et cetera, to go and edit it so that it is exactly what we need it to be. But I've taken it to a point where, and I'm not a designer, I've taken it to a point where they don't have to read my mind as to what needs to be done. They can actually go and execute with their skill sets and their expertise. To me, that is fascinating. You know, on the flip side of that, like, I think one of the things that's also really interesting is that the AIs themselves don't do well with real physical constraints. You can't put a brand into it and say, I want this brand to be in this video doing X, Y and Z. It doesn't do that well. It hallucinates too quickly and too easily. And I was mentioning earlier that my CTO explained to me that if you put a prompt in saying I want a black cat in a living room in front of a window, you'll get that. If you sell it, you want my cat sitting in my living room in front of my window. You can't get that because this puts too many constraints. That's why brand advertising doesn't work. Yet. We're doing things that allow us to do a lot more real time physical constraints, because we're taking brand models and inserting them into existing content. But that's a different model than what a standard AI is being tasked to do. It's more like what your DXM is doing too, for those types of things.
- [24:46]
Alexis Contos
Yeah.
- [24:47]
Jennifer Barkley
But it starts to beg the question of if you can plug it into the tool and the tool can spit it out, Is it a brand or is it just an asset? Like what is the brand? Is the brand is having, you know, the tone of voice and the uniqueness and the point of view. And so I think this is something that's going to be really interesting. It's just more something to think about. Right. Not then like one way or another of a brand, I should have value. You know, I get amazed all the time. It's like, I could work for a brand and just the brand, not the company, could be worth billions of dollars. What made that human elements like trust. So I think it goes back to this idea of it's the means to the end. And if you get too carried away. My daughter is 11 and we were looking at Pinterest the other day. I was showing her, like when I was going to paint my face for Halloween. She asked me and I showed her a picture and she looks at me and her lip curls up, which she learned from me. So I can't get too turned off by it. But she goes, you know that's AI, right? And I looked at it and I was like, yeah, okay, maybe. And she's like, well, it's not gonna. It's not gonna look like that. You try to paint your face like that. Cause it's not even real. And as I was like, okay, good point. Not too worried about that. It was just for inspo. But thank you. But as the. As a marketer, I was like, wow, if you're working for a beauty brand or a clothing brand or dove or any of, like, this is a really big deal. And it starts to get into, you know, ethics and how we use it wisely, which is a whole other conversation that I think everybody should have and think about. But it's really, you know, it starts. You can put your guidelines in and it can get you so far. And this is where we start talking about the roles will start to, you know, morph and change. But that idea of, like, where is it a brand and where is it the asset? And being really cognizant of that, I.
- [26:35]
Alexis Contos
Think it comes back to quality standard. I mean, my definition of brand is the sum of all touch points a company has with its customers. So that runs across the entire organization. It runs across customer service all the way to product, everything. So if you're thinking about how do you sustain the brand that you've built or what you're trying to build, you have to look across everything. So customer service AI has revolutionized the customer service model. But you have to be able to hold the standards and know what those standards are and then be able to create that across the organization. To say, we are holding these standards for our brand. And then, you know, there's maybe at some point it would be interesting to create like an advisory board that is able to. And maybe it's not the executive committee, it's like an advisory board of. As we start to leverage these tools across the organization, what standard are we holding and how are we evaluating this? Because as your company grows and you're from 50 people to 100 people to 150 people, it can start to fall apart real fast. So those are things to really start to think about as you scale and grow is like, I always use like the 80, 20 rules. Like that 80% needs to be kind of on lockdown a little bit to keep that consistency. And then there's that 20% to, like, test and play. But, you know, you don't want to stifle creativity, but you also kind of need to put some guardrails around that. And so it's a slippery slope.
- [27:54]
Corey Trefletti
Yeah, it hits me, resonates hard with me when you talk about trust. I think we've observed it and our team as a performance marketing agency so much. There's just so much documentation and data to support that. A lot of trust, unfortunately, is kind of broken or, you know, needing repair. With a lot of institutions, a lot of brands And I think if you're a brand that's emerging, you're building something to be able to over index on trust for your customers and use AI in the right ways for your customers, I think is, you know, it can power, it can accelerate, it can get more insights. And I think, I think you nailed it. We're really have to focus in on that when it comes to like affiliate influencers that can go in bad ways if it's not monitored. But, but fortunately it's not just the brands talking to you, it's, it's third parties that people resonate with and trust talking to you just like a friend. Friendly suggestion. You want that? So the trust piece is so fascinating to me in this conversation around marketing because it's such a. Yeah, such a big one.
- [28:55]
Ty Degrange
Which is why so many people have already done it wrong. If you use AI and you try to mask it and pretend it was your team that, that did that, you're always going to get caught.
- [29:04]
Alexis Contos
Yeah.
- [29:05]
Ty Degrange
And then you've lost trust, you've eroded that trust.
- [29:07]
Alexis Contos
There's somebody, I can't remember where I read it, but it was like your audience is now a blend of human and technology. And it like really stopped me in my tracks. It was like so true. Is we are now, it's not just us humans who are the audience now. Like, the technology is now going to start to weave into who we're talking to, how people make decisions, like all of it, the whole journey, the whole process is now suddenly interwoven between the two. And something to really think about is how do you define, like, think about this? Like, how do you define your ICP in the next 24 months? Like it's not just a person. So what does that look like? And you know, what does that mean for your business and how you build.
- [29:46]
Jennifer Barkley
Product because you can manipulate it and you can create your own echo chamber. Like you, we all know, like what you like on LinkedIn is you're going to see more of that. It's like, oh gosh, don't linger too.
- [29:56]
Alexis Contos
Long on so and so's post because.
- [29:58]
Jennifer Barkley
I'm just going to see, you know, whatever for the next week straight. And that's, you know, and, and so are we having authentic, authentic engagements both ways? It's quite fascinating. And you know, Ty, I'm sure you see a lot too, like this idea of, you know, relinquishing control and there's so many different elements to that. You're relinquishing control to AI, you're relinquishing Control to influencers. You put the two of those together, it's like, I don't know, those are the things that don't get past procurement very easily.
- [30:26]
Corey Trefletti
Pisa's going to veto that one.
- [30:27]
Jennifer Barkley
Yeah, we've tried.
- [30:28]
Ty Degrange
It's an interesting question too. Just to go back to the topic itself. Right. Driving better results. If your audience is partially made of technology, how do you still measure the appropriate results for what we're trying to do? Because there's a portion of the audience that is just basically taking what we do, spinning it back to us in some way, shape or form. It's almost like your website becoming a knowledge base that's just getting you crawled by a new piece of technology, then to some information in what does a funnel look like?
- [30:54]
Alexis Contos
How does the, the consumer or decision making funnel change? Yeah, you know, like we've, I've heard of people prompting in like plan my trip to Napa. It's like, okay, well a computer just decided what hotel you're staying at. Right. So it, it can completely change and not in a scary way. Like I think we can embrace this as like product builders and brand builders. Like there is an opportunity here. But I mean I can understand how it gets a little questionable at times.
- [31:19]
Corey Trefletti
But we talked about this a little bit. Like the monetization piece hasn't yet been worked out for the perplexities and the clouds and the chat GPTs yet. And look at how much has changed on Google just in a short time.
- [31:30]
Alexis Contos
Yeah.
- [31:31]
Corey Trefletti
Some of the best SEOs in the world in the last eight months have already reported a 10% decline in SERP traffic. We're seeing it directly within affiliate because so much of affiliate has moved into the content review game, intersecting with that SERP and because of artificial intelligence overviews that they rolled out and also because of their core updates. So kind of what's your strategy to kind of integrate with Perplexity and these other tools and OpenAI once, once you can.
- [32:00]
Jennifer Barkley
There has to be. Oh, go ahead, Alexis.
- [32:01]
Alexis Contos
I was going to say Arvind, the CEO of Perplexity has made it very clear that the way he's going to eventually integrate the paid model into Perplexity is the opposite of Google. It's not a pay to play model. It is how is your knowledge base going to better the search process for what the end user wants? Now that is obviously going to evolve over time, but he has a very clear importance of not falling into the Google trap and there's a lot to learn from that. I mean SEO works very well and it has for many years. But I think all of that is about to change.
- [32:31]
Corey Trefletti
Easier said than done, but very compelling.
- [32:34]
Alexis Contos
Yeah, yeah.
- [32:36]
Jennifer Barkley
As the interface goes away as it is, then the whole model, like the fact that it's a bidding model and you're bidding for visual real estate and their visual real estate is being usurped by. I mean, I don't know about. You got like 80% of what I used to Google just less than a year ago. It just doesn't happen. I have conversations with my people in my pocket.
- [33:00]
Ty Degrange
Somebody had said it once, like, if you ever killed somebody, the best place to hide the body is on the second page of the Google search result because no one wants to see it. Now it's below the fold.
- [33:09]
Jennifer Barkley
It's going to be in Web3. And it's just a really interesting thing. So you have your, you know, how, what AI looks like five years or the next, you know, what do you want to do over the next 12 months, you know, 18 months in your fiscal cycles for your planning and your budgeting and everything. And then five year and then anything past that, it's. You have to start thinking about the fact that it's an interfaceless. I'm not saying I'm advocating for it, by the way, and it sounds a little extreme, but it is the way to start, you know, thinking about what that might look like if, if you're in. In it for the long run.
- [33:38]
Corey Trefletti
Yeah, we covered some awesome things. I think we're. We're all done here, gang. Ready for some questions?
- [33:44]
E
Yeah, my question is about the, like the issue that we have now with scam that are being also using marketing like AI. So basically we're going to get sooner or later to the point where there's more scam marketing material out there than real material. So whatever your product or service is, we'll get to the point when someone looks up for your product or service, the 10 first results will be scammers, some of which may actually impersonate you in a very convincing way. Do you have any sort of thought process as to how to deal with that issue that's upcoming? And perhaps another part to that question. You haven't really set in thinking about how to use AI to better target your individual customers. Engaging outside of public ads, having personalized ads for every single human individual that you want to target. That's something that you can do with AI and you haven't really mentioned that. Is that something you think you're thinking about or something?
- [34:42]
Jennifer Barkley
Yeah. So in terms of nurturing and Targeting and being able to get to that level of personalization, I think it's one of the biggest opportunities that organizations have to be able to meet the customer where they are based on their behavior, just then nurture them through based on where they are in the buying cycle. To make sure that we're not guessing. I mean right now we do a lot of guessing. You have to have some part of your awareness, some part of your consideration, some part of your choose running and firing at all time and then you're just hoping that people are finding the right thing at the right time. We're using firmographics and other performance marketing tools, but I think that that's an area that can just take out so much of the noise and make it so much smarter. And then the fraud, a huge part of my portfolio is for payment fraud right now. And just the opportunities for being able to have other solutions that are going to Some jobs are going to not be done the same way anymore and other ones will emerge. And I think the first thing that you hit on is an area that we're going to have to start saying, I don't have a solution for it today in terms of scam, but it's absolutely a problem that's going to need to be addressed.
- [35:42]
Ty Degrange
I would add, just having been in the ad space as long as I had been, it's been way too long. The if you look at the backside, AI and ML have been around in most of the platforms that you all use for the last 20 years to do targeting. The situation is interesting because in the last five years, with this impending cookie apocalypse, when people were starting to move away from third party data and only use first party data. You've got the situation now where actually more people are going back to contextual and away from pure data driven targeting or a combo of the two things using first party and contextual. So that is interesting in that it's going to allow you to go focus more on the context which can be then read and prescribed with AI better than it was 20 years ago when it was the sole way of targeting an audience. So I think that's really one interesting development.
- [36:26]
Jennifer Barkley
And the partnerships, the partnerships that allow.
- [36:29]
Ty Degrange
You to do the clean rooms and the data sharing, et cetera. On the flip side, the thing that you mentioned around the fraud, that's going to be really interesting because I see some potential future where if you look at how companies like Twitter were using the blue checks to do verification of real people as the handles, that model could be applied to the broader Open web in terms of content and companies where you could start to verify true companies as the source of content if companies like Perplexity allow that to be baked into its back end infrastructure. So there is a model where that verification component can come to the forefront across the entire open web. Because to the point that was said earlier, traffic is going to drop by 50 or 75% across the open web over the next few years because of the generative search. You don't have to go anywhere. Everything's going to come to you. And if that information comes to you like it does with Copilot now sourced with verification that it came from a legitimate company, that's a real big step forward to being able to debunk fraud.
- [37:25]
Corey Trefletti
Thank you guys for everything. Alexis, you mentioned the difference between using AI for hacks versus workflows.
- [37:31]
Alexis Contos
Yeah, I was curious if you guys.
- [37:32]
Jennifer Barkley
Could just speak to some of those.
- [37:34]
Corey Trefletti
Tools that we can use for hacks.
- [37:35]
Jennifer Barkley
Around more like video editing.
- [37:38]
Corey Trefletti
You're cutting like long form down to short form and trying to define some of that. And then you mentioned something around incorporating branded elements to.
- [37:46]
Alexis Contos
Yeah, over that.
- [37:47]
Jennifer Barkley
Can you just speak to some of.
- [37:48]
Corey Trefletti
Those tools that we're not familiar with?
- [37:50]
Alexis Contos
The branded tools, I think Corey would agree is like we're in the early stages of that. I think there's a lot of work to be done to integrate brands into content generation. I've used dxm. They're a small company. I'm happy to make intros to whoever would like that. Corey's company is doing some amazing work on the workflow side. Has anyone used Poppy AI? It's fascinating. So it is. It's basically a workflow board that you can feed with anything. You can feed it with LinkedIn posts, product information briefs, photos, and you put it all on the board and then you, you basically create a traffic pattern to what your output is. It could be a landing page, it could be a creative brief, it could be photography. And all of that information is feeding into the output, which is really incredible. I mean it's basically if you're, if you're like, I have all this information and I know that I need to get to X, you just put it all into the workflow and then you tell it what you want the output to be. So that's really cool. There's Runway, which is still fairly new, but I think the, the what it's capable of generating, it's not necessarily video editing, but I think what it's capable of generating is giving production to anyone in their Pocket. Like you would probably have to spend 20 grand to get some type of animation output where you can just do that by recording yourself and having that turned over. And then on the content creative side for writing, I've actually been spending more time with Claude and now there's Jasper and Copy AI and things like that. But I actually am so impressed with the creative caliber of Claude's output. I spent a lot of time with ChatGPT just because it was easy. And then I switched over recently to Claude and I'm really blown away. It's really impressive. And then obviously, like Anthropic made a huge announcement last week where Claude can take over your computer and just do whatever you need it to do, which is scary and fascinating at the same time. So the demo that I saw was they had Claude build a website. They said, go build a website. It went in, it generated the code, it downloaded the code and prompted it to create a website. So there's a ton of stuff out there, but those are just a few of them.
- [40:02]
Ty Degrange
I would just add too, that the AI built into stuff like Canva and VEED.
- [40:06]
Alexis Contos
Oh yeah.
- [40:07]
Ty Degrange
Like, I use VEED for all the video editing software and it's great because you push a button, it'll spit out a whole bunch of social clips.
- [40:12]
Alexis Contos
Yeah.
- [40:13]
Ty Degrange
Using the video. Those are great tools that are built into things you already use.
- [40:16]
Jennifer Barkley
Yeah.
- [40:16]
Ty Degrange
So it's a habit and it'll just spit out 10 different clips for you.
- [40:22]
Alexis Contos
Canva's done an amazing job of pivoting in the right way and centering all of their products and features around AI.
- [40:29]
Ty Degrange
I love that when you don't have to go somewhere else to find it, it's in there.
- [40:33]
Alexis Contos
There's also. I know this isn't like creative output, but I end up dealing with this a lot. There's a company called Gamma that creates PowerPoint presentations for you. You literally tell it the topic and it will create the content and design the PowerPoint for you. And so I've used that just as like a starter because again, like, who wants to stare at a. It's called Gamma. It's really cool and they have like a free trial so you can go in if you're like, I don't have time to build a PowerPoint presentation. At least just give it the content of what you want to talk about. And it's just like a really great place to start.
- [41:01]
Corey Trefletti
We use affiliate AI just to auto pull in insights and data into our Slack channel, Auto Populated, which is kind of a niche one. Jasper, you mentioned Cam is a great one.
- [41:11]
Alexis Contos
Oh, the other one is Napkin. AI takes text and generates it into charts. I am terrible at chart building, but if you just literally say this is the data, it will, it will create charts for you, which is really cool actually.
- [41:23]
Ty Degrange
I'll give you another one too. I love this game. Promo.com. you can go there and basically type in two or three lines of text and it'll spit out a whole bunch of social media posts with video or without video and animations. And you choose the ones you like, download them and you're off the races.
- [41:35]
Alexis Contos
Oh, that's cool. All right, we have time for one more question. We could go all day.
- [41:40]
Corey Trefletti
Yeah, it's going to have to be.
- [41:42]
Jennifer Barkley
On time for a part two.
- [41:43]
Corey Trefletti
Better stop us now.
- [41:46]
Jennifer Barkley
Thank you. I would love to hear from each of you. How is AI leveraged in the high level movement making process?
- [41:57]
Alexis Contos
Sorry, say that again. Decision making. High level decision making. Can you give an example of.
- [42:03]
Corey Trefletti
Sure.
- [42:03]
Jennifer Barkley
Just leveraging AI data for marketing strategy and then whether it's doing a campaign or product launch or roadmap.
- [42:12]
Ty Degrange
I mean, I'll be blunt, we don't use it to make decisions. Like it's a great way to basically quickly aggregate the information you want into a good visual way. But I'm never going to let it make a decision.
- [42:22]
Alexis Contos
I have used NotebookLM to aggregate data. So like I've had just droves of data or Excel spreadsheets that weirdly you can't upload in Excel to NotebookLM, so you have to turn it to a PDF. But I've used it just to synthesize like hundreds of cells of what should I be doing with this? And then it'll, you can just keep asking it like prompts to say how would I handle X Or what should I do with this? So you know, I've just used like aggregators to just try and distill data. But I agree is like you're using it to help you inform and make better decisions. But it's more of an efficiency play. Yeah, yeah, yeah. Save yourself some time.
- [43:01]
Jennifer Barkley
I'll use it to challenge my own biases. So if I feel like, like I got this like fast decision making, great. It's like, what are the things that I'm not considering? What's the worst outcome that could possibly happen to try to think through and try to augment decision making. It's not anything that's going to be completely outsourced. But I think that's one of the themes of the session is that you kind of have to know what it is that you're going after before you bring in these. These tools. So. Yeah, yeah, yeah, I'd answer that.
- [43:29]
Ty Degrange
Yeah.
- [43:30]
Corey Trefletti
I think the use case I mentioned earlier, where they're utilizing Claude and getting a lot of their early customer data in there and asking it questions similar to what Alexis said, I think is really helpful and helps you kind of dig deeper on, like, okay, what is this large data set? What are the trends? What are some of the data that we're seeing? What are some of the blind spots that we're not seeing, perhaps? And just making that gathering a little bit better and maybe cleaning it. You got to have clean data. You got to have good inputs. You got to be able to make some real. It's nothing if you can't make some conclusions around it or some determinations or some next steps. And I think you brought up a good point, and it kind of underscores that we've got to have the human judgment and the taste and the strategy and the vision. Without that, it doesn't really matter if you're using the best tool in the world.
- [44:20]
Alexis Contos
All right, well, thank you so much to our panel.