How To Produce Better Content With Collaborative AI

John (00:08): Hello, and welcome to another episode of the Duct Tape Marketing Podcast. This is John Jantsch. My guest today is Kate Bradley Chernis, former rock and roll, DJ, turn founder and CEO of lately, AI tool that uses proprietary language models to craft personalized social media messaging. Lately, AI ensures data privacy by not relying on public data sets. Kate’s also been a guest speaker at numerous industry events and organizations including Walmart, Ericsson, and Harvard University, and I don’t know, maybe third time back here on the Duct Tape Marketing podcast as well. Welcome back, Kate.

Kate (00:45): Hey, John. So great to lay eyes on you. I feel like it’s been a little while.

John (00:50): It’s been

Kate (00:51): So,

John (00:52): I know you get tired of telling this story because it’s the first thing everybody always asks you, but I know people are going, wait a minute, rock and roll, DJ now founder of a company. How do you do that?

Kate (01:01): I’ve gotten better at telling the story too, which is important I think. And my co-founder teases me. I do often bury the lead. So yes, guilty is charged. I was broadcasting to 20 million listeners a day with XM Satellite Radio. I was the first music director for a channel called The Loft. But what was interesting to me about radio was the Theater of the Mind, which a lot about being in podcasting, but to clarify for everybody else. So the theater of the mind is the act of the imagination playing a role either when you’re listening or reading, not when you’re watching tv, for example. And what you’re doing is your imagination is filling in the blanks that you can’t see, right? You’re imagining what the characters look like or what they’re doing. It’s why the reason when you see a movie and you’ve already read the book, you’re kind of mad because it’s never as good as what you’d imagined, right?

John (01:55): Well, or a lot of people, I love to listen to baseball as opposed to watch it because baseball announcers are so much better at describing what’s going on because they have to.

Kate (02:04): They have to. And they’re so crafty. I mean, that’s a real sport in itself. Exactly. Great point. And when I was in radio, I’m old enough so that there wasn’t social media when I first started or the internet and you couldn’t look people up. And so we would kind of mess around and play tricks on the listeners and make up these scenarios. It was fun. And I had written hundreds of commercials because I learned quickly that was how you made money in radio. And I was a fiction writing major, and I saw these parallels between wielding the mic and wielding the pen and listening and reading. And my boss, I was number one in our format, which was very rare because it was called AAA or Adult Album Alternative. There’s only a handful of stations in the country, but

John (02:53): We were, and even fewer women.

Kate (02:55): And even fewer women. Yeah. It was just totally random thing I fell into. But country and rock, those stations are number one. And so my bosses were like, what are you doing? And I’m like, well, I did know what I was doing. I threw out their playlists and I was running the whole show because all the content was produced by me, all the commercials, all the drops, everything during my time, but I looked into it more. This is a long story. I hope it’s interesting, and I read this book called This Is Your Brain on Music. You guys remember, I think Daniel Leviton did that hard read. It’s a F read, but it’s about the neuroscience of music and music listening. And I learned something interesting about the parallel of music listening and theater of the mind. All of this relates to lately somehow, but I’ll share it. So when your brain listens to a new song, John, what songs do you like, by the way? Are you classic rock guy like I am?

John (03:48): Yeah. Yeah. I mean, I listened to The Loft and occasionally jump over and listen to Earl Bailey.

Kate (03:53): That was my station. Yeah, I love Earl. He’s so great. Yeah, so great. Oh my God, you’re making me reminisce. So great quality music, rock and roll. That’s for I’d say intelligent rock and roll. Let’s say what I was learning from Daniel was that when your brain listens to a new song, it must instantly access every other song you’ve ever heard before. And it’s trying to index that new song in the library of the memory of your brain. This is happening in a moment, right?

John (04:22): Makes sense. And of

Kate (04:22): Course, in order to access all that memory, it’s pulling on nostalgia and emotion, obviously memory, all those things that create trust and trust is why we buy now. Guess what? The theater of the mind kicks in. Same thing happens, nostalgia, memory, emotion, trust. And when you’re doing a good job on the mic, John, you actually make your listeners feel as though they’re talking. You’re to them directly that this one way street is a two way street and they have ownership in the conversation and writing is the same thing, and it’s a complicated feat to do it well because you’re talking to nobody but also somebody specific. That’s the magic. So I took these ideas to a little company called Walmart and I got them 130% ROI year over year for three years with what became the prototype for lately.

John (05:19): Screw it. We’re not going to talk about lately. Let’s just talk about the Jay Hawks latest

Kate (05:24): Novels talk. Oh my God, you’re so funny. I took out their last, well, I dunno if it was their last record, but the one with cloud something cloud, it was their last really poppy record from the nineties. And I had that in my, I still have a CD player in my car and I was rocking to that record. I love it so much. Any other Jayhawks fans listening to us, I wonder

John (05:47): Of a certain age

Kate (05:49): Of a certain age. Yeah, for sure. But I love that record and it got panned for being too poppy, but I think it’s a real lot of gold in there.

John (05:57): Absolutely. So let’s talk about content marketing. AI seems like daily is changing. I mean, content marketing has changed dramatically over the last decade or so, but certainly AI seems to be changing it every day. What’s your take on how it’s changing really the whole landscape of content marketing?

Kate (06:14): Well, I mean, thanks a lot chat GBT, because now everybody can make more garbage than ever before. They made our jobs a lot harder. The task for marketers, the challenge has always been how do we cut through the noise? And now there’s just so that certainly has changed the landscape. One thing that I’m seeing, and I wonder if you are, which is astonishing to me, the laziness is not changing. So it’s specifically regenerative AI and text generative ai, which is where I live. People still hate writing. They don’t want to do it. But also their value behind it is save time as opposed to be more effective. That’s shocking. And honestly, I’ll ask our own clients this question all the time and the CEO or the CRO, they want to make more money, but the actual users are thinking of save time and getting them to be aligned is a challenge.

John (07:11): My take on this a little bit, I agree with you, at least we’re in this phase right now of more noise, but I think eventually all things people are going to go, it’s pretty easy to separate noise from signal, maybe even more so now, right? Because writing quality content still takes strategic thinking. That’s right. And I think it makes people who do strategic thinking even more valued, even though right now there a lot of ’em are feeling sort of undervalued.

Kate (07:39): So on parallel with that, so there’s this symbiotic relationship between AI and humans who can think strategically and analytically and they rely upon each other and it’s called collaborative ai. This is the year of collaborative ai, in my opinion. We built collaborative AI into lately from the beginning, which, but it’s the idea of a human analyzing and course correcting what the AI generates so that it can boost the learning. Fascinating about what you said, which is the number one vacuum of skills across the globe is guess what? The ability to analyze. And the reason that is is because we have, this is back to laziness too. We’ve become a culture unable to identify problems because for so long, especially in corporate life, it was like, don’t bring me a problem, bring me a solution. So even when I have a friend who has some teenage daughters and when they need to go, they know they can Google the answer to anything, but they don’t know what to type in, right?

John (08:49): Well, and that’s so many things brought up there. What I tell people all the time is what we’re left to provide is context. That context cannot be provided by chat GPT. And so to your point of the search, I mean a lot of it is the right context produces the right answer, but these machines are basically just going into a database and saying, here’s what I think the answer is. Whereas we are saying, well no, here’s the real problems the customer is telling us they’re struggling with and why our solution or whatever it is we’re selling is the answer for them. And I think short of having that understanding, it’s a crap shoot what you’re going to get back.

Kate (09:29): Yeah, I mean that, thank you. I’m going to steal the context because it is so true. Someone was just asking me the other day, well, should I second guess everything that lately generates for me? And I said, yes, you are still the king, the humans. We are still the king of the food chains. Of course. And he’s like, well, won’t AI know better than me? And I’m like, never. No. All AI is good at doing is now is synthesizing scale really,

John (09:56): Right? Yeah. In fact, I’ve been for a long time because I don’t really think it’s AI yet to be truthful. It’s not really artificial intelligence, turn it around. It’s more informed assistance is kind of how I talk about it. It’s my pleasure to welcome a new sponsor to the podcast. Our friends at ActiveCampaign. ActiveCampaign helps small teams power big businesses with the must have platform for intelligent marketing automation. We’ve been using ActiveCampaign for years here at Duct Tape Marketing to power our subscription forms, email newsletters and sales funnel drip campaigns. ActiveCampaign is that rare platform that’s affordable, easy to use, and capable of handling even the most complex marketing automation needs. And they make it easy to switch. They provide every new customer with one-on-one personal training and free migrations from your current marketing automation or email marketing provider. You can try ActiveCampaign for free for 14 days and there’s no credit card required.

(10:57): Just visit active campaign.com/duct tape. That’s right. Duct Tape Marketing podcast listeners who sign up via that link will also receive 15% off an annual plan if purchased by March 31st, 2024. That’s activecampaign.com/duct tape. Now this offer is limited to new active campaign customers only. So what are you waiting for? Fuel your growth, boost revenue and save precious time by upgrading to active campaign today. Alright, so I’m going to ask you a really hard lately question. You and I started talking about lately three or four years ago at least, and at that 0.3 or four years ago, what lately was doing was very cutting edge. People didn’t necessarily understand it, but it definitely produced a result that was very cutting edge. Fast forward to today, everything you buy now has AI in it, supposedly like your detergent now has AI in it. I think if you buy it, so what’s the differentiator or what’s your stay ahead cutting edge play.

Kate (11:59): Yeah, yeah. So I told you it was going to be a hard question. Great question. Well, a couple of things. Yeah, it’s a hard one and there’s functional answers and then there’s kind of aesthetic answers I’m going to call them. But so functionally we’re the only generative AI that I know of where we have a continuous performance learning loop plugged in so that the results that we generate for you are always tied to your personal analytics or the analytics of your company, which is to say it’s never out of thin air. All other generative AI doesn’t know you in any way and can never give you results that are essentially really meaningful. The other component there is the collaborative ai. So because we built that into the product, the whole product originally we have kind of pole vaulted over everybody else. Harvard Business Review just released an article about collaborative AI citing lately as a leader and one of the studies they did show that collaborative AI outperforms AI alone two to seven X every time.

(13:02): But on the sort of aesthetic side, again, what’s really interesting to me is that the save time piece. So of course we save time everywhere else, but that is not our cutting edge leg up. The leg up is we show you why it’s effective, what’s the DNA of the messaging that will get you the highest response. And I think what we’ve done a poor job of is actually leveraging how well we do that and what that value is. So I’ve called upon my engineering team for this year to actually do a better job of getting people to understand this information and how to use it. Fascinating to me is I can show you these words, John. I can show you the ideas, the send, the structures, all the things that will get you the most engagement, but people then don’t know what to do with it, which is like to me, duh.

(13:58): But that is the crux, right? The last answer to your question is going even deeper here. So I’m planning an integration with my friend David Allison, who owns the value graphics database and value graphics are identifying how to group people by what they care about as opposed to demographics, which is more insightful, radically more insightful, and they consult the United Nations. And so some characteristics would be like if I care about the environment and you’re selling me lipstick, you want to sell me lipstick that talks about how great it is for the environment, or if you’re selling me lipstick and I care about family, you want to mention that the family company has been around for a hundred years passed on by daughter, whatever. And so we’re working on a way of integrating these values inside lately, so you can get more of that why and understand who your target audience is and why are they responding to the content we’re generating for you? Kind of nerdy, I don’t know. I’m excited about

John (15:02): This. No, but I think you’re absolutely right. I mean, I’ve been saying for years, I mean my target market is based on behavior, not on how old somebody is. It’s what they value. It’s do they invest in the community? I mean do they invest in their industry? Those are how we actually identify some of those behaviors. And I feel like that’s a way to actually niche down is to focus on behaviors. So having obviously tools that, and I’m guessing that you’re going to go into personalization at some point with that level of segmentation as well. Let me ask you, I said in the intro, in your bio I mentioned the idea of data privacy. I know it’s a big deal. How much has the market perceived this idea of public data sets versus privacy versus, I don’t know what you’re talking about.

Kate (15:46): Yeah, pretty huge. And it’s another arena of AI where everybody thinks they know all about it but they don’t, which is kind of the whole trend of the last year. So some companies come to us with an AI task force and that has to be part of one of the initial calls where we’re checking the boxes for them, the safety boxes with their legal and IT teams, which we check a lot of those boxes. Then there’s other companies like PWC who’ve gone full on into AI and they don’t seem to really care, which is, and then there’s a lot of companies where they know people are using it even though there’s maybe a ban on Chachi PT throughout the company, but people are using it anyways. So there’s nothing consistent for sure. My husband actually just bought Chachi PT for his phone because he didn’t want to put on his work computer, but he wants to be able to use it for work to help him do things faster, smarter, better. Of course, I think there’s a lot of that going on. What I understand,

John (16:52): I mean it’s stupid. It’s stupid. I mean it is like I can write a formula in Excel or I can just dump all this in and say give me the answer. So I use it all the time for stuff like that. But maybe we better back up just a minute because I asked that question assuming a lot and assuming that people knew what that really meant. So when I go to chat, GPT, everything that’s being put in there is helping teach the entire language model and everything from my statistics to my Google analytics that I’m getting analyzed. I mean that’s all just being fed. And then theoretically in some fashion, anybody has access to it that, I mean not specifically to it, but it’s feeding the machine that then is going to produce something. Whereas the private data set that is if I come to lately and I put that same kind of information in there, it’s only going to be used to build my personal model. Is that the way to sort of explain it? That’s

Kate (17:46): Right. That’s correct. Yeah. We don’t take any of your information and muddy it with anybody. And the one thing we can see is the patterns that if things are working well for you and they’re working well for another customer, we can see those patterns but we don’t share them with you individually. We would take the knowledge and share it at large. And so that’s been a real win for us by the way, because I’ve been asked to actually give courses on AI to educate companies on why that exact thing matters. I think David, not David Allison, David Meerman Scott who was investor and friend, he put it succinctly where to help people understand, he was like, listen, there’s only two questions that matter whose data and whose math with chat CBT, it’s the world’s, it’s your data, it’s the world’s data, everybody’s data and general math like a general generic math with lately it’s your private data and then our math on top of it.

John (18:49): So I need two more questions. I’m going to ask you the first one just because we haven’t, you and I have talked a number of times, it’s that idea of like, oh yeah, we have listeners too. If somebody came to you and said, lately I kind of heard of that, what does lately do? How would you describe what lately does?

Kate (19:05): Oh, I’m so bad at this. It’s like the shoemaker has no shoes, but I’m evolving. So lately learns the patterns of when you write well, what helps you do that. And it also learns the patterns of what your unique audience will actually reply to on social media. And then we help you evolve that model by repurposing long form content and identifying what part of that content will actually get you the highest possible engagement on social. Awesome. How did that go? That was

John (19:35): Great. So the output could be a blog. No, that was very good. It was still maybe a little philosophical. It’s long.

Kate (19:42): I know.

John (19:43): So the output, the end output is a blog post or is a LinkedIn post or is a X post, right?

Kate (19:49): Yeah, the output is a social media post and so much more. I mean it’s really the insights to know this is why we have investors like you and David Merman, Scott and others, is like there’s so much potential in what we’ve identified. How can we evolve the product to really give you more

John (20:07): So an entrepreneurial question to send us out. Do you think, I know you haven’t done this a hundred times, do you think that working in an industry that is evolving so quickly makes it even harder to evolve a business?

Kate (20:24): Oh, I mean the challenges are, yes, for sure, but there’s so many other smaller challenges that I didn’t expect that seem to me to eclipse that some of it’s being a female entrepreneur, let’s be honest. Some of it’s working through a pandemic. I think the way that we come at this from radio, from this totally unfathomable background gives us a huge insight as a company, not just me at how we go at AI and we go at it very humanly. That’s just how we did it from the beginning. So I love that what we’re able to see about the benefits of it our often inside out of what everybody else is seeing. And I feel really proud about that.

John (21:10): Yeah, that’s really interesting because I do think a lot of people approach this as what can the machine do? And I think that you’re actually saying our point of view is how do we get the output that’s going to have the most impact from a neuroscience point of view? And I think that’s a harder one to explain probably, but it’s certainly more impactful than a machine view for sure.

Kate (21:33): I just got a Kennedy chill, so not what can machine do for you, but what can you do for your machine?

John (21:40): I like it. I like it. Okay. T-shirts. Start printing right now. We have to. Alright, Kate, it’s great catching up with you. Obviously we’ve mentioned lately do AI numerous times. Is there anywhere else somebody should connect with you?

Kate (21:55): They can find me in all the places LinkedIn. I’m just playing Kate Bradley and tell me that you met me with John and that we can be friends.

John (22:03): Okay, awesome. Well it was great catch up with you again. Hopefully we’ll run into you soon out there on the road.