Speaker 1: (00:00)
Are you drowning in a sea of content, struggling to make your voice heard above the noise? Get ready for a game changing episode that almost became its own podcast. In this special edition of Talk Commerce, we're diving deep into the world of AI powered content creation with Kate Bradley, CEO of lately. That's Kate Lee from lately. I initially planned to launch a new show dedicated to this very topic, but realize the invaluable insights Kate brings are too good not to share with our talk commerce audience. Prepare to discover how neuroscience and artificial intelligence can transform your marketing strategy. Turning casual customers into passionate brand evangelists from a roots as a rock androll DJ to becoming a tech innovator. Kate shares her journey and reveals how lately dot AI's unique approach has helped clients achieve a staggering 12000% increase in engagement. Whether you're a seasoned marketer or a small business owner, this episode is packed with actionable strategies to revolutionize your content game. Don't miss this opportunity to stay ahead of the AI curve and supercharge your marketing efforts. Tune in on August 27th, 2024 to unlock the secrets of data-driven personalized content that will take your social media performance to new heights.
Speaker 2: (01:27)
You're listening to Talk, commerce, subscribe and download@talkcommerce.com.
Speaker 1: (01:37)
Welcome to this inaugural episode of what I'm calling content in commerce or content in ai or some variation of that. Today we have Kate Bradley from lately, Kate Lee from lately , and we're gonna talk about how machine learning is involved with content creation. So, Kate, go ahead. Give us an introduction. Tell us your day-to-day role, and give us a passion that you have.
Speaker 3: (02:03)
Oh, Brent, I, um, I adore you and certainly one of my passions is connecting with the people online, which, you know, we're both so good at. And, um, we've become friends, which is kind of amazing. But I am the CEO of lately and lately leverages neuroscience driven artificial intelligence to key into the exact messaging that'll get you the highest sort of performance in, in your marketing efforts wherever you are on social media. That's the short version, which is hard to do, ,
Speaker 3: (02:40)
And I'm testing it out. Um, let's see, any other passions? I, I'll tell you that. Okay, this is so crazy and snobby, but recently we, we've, we've become caviar, um, aficionados, but not like that. That sounds so terrible. Um, just Trout Row or Salmon Row, the kinds that are like bubbly that you can pop in your mouth, which is funny. And it, you know, you can have it over three nights, like a little jar of it with we cheat and we don't get bellinis. We, we get, um, brown rice crackers that are not salted, so it's a little healthier. And then they have, I live in, um, New York, but we found Vermont Creme Fresh, which I'm from Vermont, so Yes. And you just like, put a little dollop on there, put on your little spoonful of caviar and then like dill or something like chive from the garden. And we feel super fancy . And we've been doing it like, I don't know, every couple of weekends, the last few weekends.
Speaker 1: (03:39)
Wow, that's cool. Um, it, it is cool. Ikea used to have a ver you know, oh, the, the version of some version of Caviar. Yeah. It wasn't called Caviar, but it was a Swedish IKEA version. You had to assemble it. It came out flat
Speaker 3: (03:53)
.
Speaker 1: (03:54)
Uh, but in all seriousness, they, when they first opened, they had a, some kind of a version of of that. And, uh, I have, their market has completely changed. They, they opened and they had all kinds of Swedish stuff and now it's just, you know, generic American stuff that's made in China. So
Speaker 3: (04:09)
They know, they know us by now. . Yes.
Speaker 1: (04:11)
Yeah. Nobody bought the Luta Fisk and nobody bought the God. Yeah. Who would, they did have Luta Fis, that first couple of, of Christmases in Minnesota. Anyways, they had Lu
Speaker 3: (04:21)
Fisk. So Yeah, we saw it too. In, in, in was see in the tube. There was something in the Tube, remember?
Speaker 1: (04:26)
Yeah. What Luta Fisk is, is like a lie based fish. It's horrible smelling. So yeah. Anyways, I've heard . Um, so Kate, before we start, I'd like you to, I would like to tell you a joke. I'm calling, this is my free joke project. All you have to do is say, give me a rating for the joke of one through five.
Speaker 3: (04:44)
Okay.
Speaker 1: (04:44)
I'm ready. And I have a very, very poor joke today, but maybe we will, we'll revisit a different one that I have later, but here we go.
Speaker 3: (04:53)
I'm ready.
Speaker 1: (04:54)
Why couldn't the green pepper practice archery? Because it didn't have an arrow .
Speaker 3: (05:01)
I'm gonna do a four on that one. Okay. Maybe a 4.5. 'cause I really like puns. Um, yeah. And only on the delivery. I'm taking a point off on the delivery. Like, I, I think you could have paused a little bit before the, I'm just messing with you. I thought it was so great. Can I have that joke? Because, um, I've been feeding jokes into lately. Oh. Um, yeah, and they're dad jokes, so like, that's just perfect to kind like divvy up the results we give to people. Um, you know, sometimes you just need a funny break to jump in there and .
Speaker 1: (05:35)
Yeah.
Speaker 3: (05:36)
So I'm gonna write that down. It was a green, it was a green pepper.
Speaker 1: (05:40)
Yeah.
Speaker 3: (05:40)
Wait, what, what, say it one more time. Why didn't the green pepper to what,
Speaker 1: (05:44)
Why didn't, why couldn't the green pepper practice archery couldn't because it didn't have an arrow
Speaker 3: (05:52)
Practice archery. I'm,
Speaker 1: (05:53)
I put these, I put these on TikTok, so I have to, I try to get 'em done, done in, you know, 15 seconds.
Speaker 3: (05:59)
Do you deliver them? I'm, I mean, I'm assuming over video, are you dancing through it
Speaker 1: (06:04)
? No, just like this . Um, I, I was, I I would like to do a, um, a dad joke off. Yeah. So like, you know, it would be great to what, what one of the ideas that I came up with was let's do a, let's do a podcast, uh, for, you know, 10 minutes and just do joke offs. And then we will take the, I'll edit the video down and just come up with a, you know, a 32nd, a bunch of 32nd TikTok version. So I love it. Each person has to come to the, each person has to come with some jokes. So maybe five minutes of jokes , which is actually a long time. So, but
Speaker 3: (06:43)
That is a very long time actually. Um,
Speaker 1: (06:45)
Well, the whole point is editing,
Speaker 3: (06:47)
Right? Um, so, so couple things that I wanted to share with you. Number one of, of related note, we worked with a company called Levity Live, which they own the video rights to pretty much every comedian you've ever heard of. Um, and so they ran a Drew Carry special through our algorithm, and lately clipped up all the best jokes, . And I didn't realize it, but then someone else was like saying, oh, so have you like, literally found the, the algorithm for, for humor, you know? And I was like, well, that's a great, that's a great use of ai. I mean, that we haven't even, you know, thought of. But then the counterpoint to that, my husband was reading an article, I don't know where it was, um, yesterday, that said the, the, the people who are leading the AI movement or in creating companies are for the most part, um, liberal studies majors, which I was, because the one thing AI can't replace is that is the creativity, right? Which is where joke, joke telling also comes from, you know, and the, and the delivery. I think,
Speaker 1: (08:02)
Yeah, I think the, the problem with the problem that I see right now with content creation and AI for, for, and the creative process is that it does just water everything down. Yeah. Grammarly is a great, a great solution for helping you with grammar, but it also makes your text extremely bland. Yeah. And it wants to take out all the little nuances that make you who you are. So I I, I do agree. And, and I, I am, I'm very fascinated by all these new private models that are coming out for, for ai. And, and I did find somebody that had come up with a, if you Google AI private model or AI model that you can download and run on your local machine, there is a solution out there. It's just a paper. Somebody wrote like, nobody,
Speaker 3: (08:52)
There
Speaker 1: (08:52)
You go. Nobody has gotten past the point of finding that. And if you ask like chat GPT or Claude or any of them, any of the models, what could you gimme some good jokes? They'll give you horrible jokes. and I have great examples on my Talk commerce website of horrible e-commerce jokes that were written by chat GPT .
Speaker 3: (09:13)
Yeah. The, it can't, well it can't read the room. Right. And, and neither can Grammarly. Right? So Grammarly's boss is Emily Post, and you know, Emily Post hasn't evolved. She's, she's still, you know, back in the, those old kind of ways, elbows off the table, all the things. Um, you know, that's kinda like segue. But I was just thinking, I'm about to get on an airplane this week. And remember we used to dress up to go on the airplane. My husband still does. He always wears a suit. He always looks really super nice. That's because they always frisk him at the TWA thing for some reason. Like, he always gets flagged. W um, I usually wear something nice looking, but like athletic. 'cause I wanna be able to open that emergency door in case I have to throw someone out it or whatever. Um, but like, what I can't, and I'm sorry if I'm offending you or anybody else, what I can't get over right now is wearing pajamas to the airport. And , sometimes people are wearing their blankets and flip flops and you're like, good God. Have you no sense of like, you're seeing you're, these are strangers, you know?
Speaker 1: (10:21)
Yeah. Is that, is
Speaker 3: (10:22)
That old? I fly,
Speaker 1: (10:22)
I'm old. I mean, I, I did, I I used to fly to Europe more, and if you were to sit in, not that I would sit in first class very often, but , uh, there would be lots of people that would actually change into pajamas. And then as you're landing, they'd change back into
Speaker 3: (10:37)
Regular clothes back. There you go. That's totally reasonable. Um,
Speaker 1: (10:40)
I, I now fly to a lot to Hawaii and we, you know, we don't, we don't do a long haul flight, but, um, I do wear socks with my Birkenstocks as my dress up. There you go.
Speaker 3: (10:51)
That's allowed. Yeah, that works. That's, that's
Speaker 1: (10:54)
Allowed. You never know when you have to flip off your flip flops and you gotta walk through TSA with your bare feet,
Speaker 3: (10:58)
That's disgusting. . Yeah. I'm totally with you. . I, I, I've, anyways, I'm always astonished, like, the people watching is just like slob. It's slobs. Everybody does a slob, you know, and, and there's anyways that someone that's, you know, I don't know if AI can help us with this at all, but the dignity , I think we need to retain as humans for sure. Um, part of that, you know, on the AI note comes with the ability to analyze what it's doing and to make sure that it is representing us or you, or, you know, the work you want it to, um, put out for you in a way that's appropriate for whatever your audience is. And I think that's the, that's the challenge, right? That's where the human still must lead and fill the role. Um, however, as you well know, this, this s lobby, this spills over here as well. 'cause people are like, ah, it. , , let's just, whatever that thing said, let's just do that. You
Speaker 1: (12:02)
Know? Yeah. I, I think that laziness. And so I, I want to drill in on a couple points today for AI and content. The first one is what you were saying, the laziness in AI content and then the laziness in human editors as they look at it. Mm-Hmm. . I think that's a big problem. Yeah. And then obviously the content generation and creativity is the next one. I think what your solution and your solution has been around for a long time. Yeah. Um, can I say it's been around for
Speaker 3: (12:32)
10 years?
Speaker 1: (12:33)
Nine 10 years? Yeah. Yeah.
Speaker 3: (12:35)
Hard to believe , right? A decade.
Speaker 1: (12:38)
Yeah. And that was before the gen AI thing even became a buzzword. Uh, but just tell us a little bit about the difference between analyzing and reading and then breaking it into bits compared to, and learning a voice. I mean, I think the key point too, there is learning, right? All the models we're using aren't actually learning about us. That's right. Unless you That's right. Make your own little private model. So talk a little bit about how important it's for that voice to happen.
Speaker 3: (13:05)
Yeah. I mean, so couple of points I should tell people what we do. Um, so let me, let me do that, um, first and then we'll talk about some of these points, which I think are so important. So, so, you know this about me, Brent. Um, so my, I I used to be a rock and roll dj. My last gig was broadcasting to 20 million listeners a day for X XM satellite radio. And my uber power was turning listeners into fans or customers into evangelists. And that's a big difference, right? And that's what, as, as marketers, content creators, that's the holy grail. We wanna get stuff done, we want people to do what we want them to do, click buy, share, like, whatever. But then if you can get the beyond, like the bed, bath and beyond, , right? You want people to be your, your evangelists and to, and to share for you and, and build a, a natural flywheel.
Speaker 3: (14:01)
And, um, I started under trying to started to learn about the neuroscience of what makes that possible, at least in, in listening to music, you know. And then, um, I was able to understand what that did with, with writing as well. So when you, you know, become a voracious reader, what's, how does, how do you become so involved in the story that you feel like you have an ownership of it? And there was a parallel that has to do with the theater of the mind in the these two instances, right? So when you're listening or when you're reading your, your brain has to fill in the blanks. It's not the case with video, right? Video. You just give everybody everything and they lie back and it washes over them. And, and that's why we veg out, quote unquote, right? But so for the theater of the mind to kick in, your brain has to access memory, emotion and nostalgia to sort of kick in there.
Speaker 3: (14:57)
And similarly, when your brain listens to a new song, it's running down every other song you've ever heard in this quick moment. And it's trying to find, okay, well what are the familiar touch points? So I know where to log this new song in the memory of my mind. And when it does that, guess what? Same deal, nostalgia, memory, emotion, same parts of the brain. So there's a, a connection about that, that powerful feeling we get. And then the, the third thing I started to think about was like, alright, well, when Brent writes me an email and I read it, I'm actually hearing his voice in my head. And his job is to convey his voice. Like I have resting face and email, so I have to use like a lot of emojis, capital letters, you know, italics, that kind of thing. And what I'm doing, what I'm writing is I'm trying to figure out, well, how do I trigger nostalgia and memory and emotion in my writing to have this same strong feeling of ownership and connection happen in the conversation?
Speaker 3: (15:53)
So lately, does that its job is to identify what messaging will get you these sort of outcomes. And specifically we, we, we work on social and um, it is a private database and, you know, all those kinds of things. So the analytics piece we were just touching on is so important because when the results from the AI come out, if you can't tell it, if it's done a good job or a bad job, thumbs up or thumbs down, then it can't possibly learn, right? And similarly, I like to think about, um, like a calculator. Look, I have one, I I actually don't actually use a phone. I have a calculator. Um, 'cause I like the buttons. But when you type something into a calculator and the result comes out, you need to be able to know how to do long division, essentially, so you can make sure that the answer is right.
Speaker 3: (16:43)
So you didn't, you know, your big thumb didn't hit the wrong button kind of thing, right? So it's the same idea if you're, if you don't know how to write, if you're a poor writer and you're using AI for content creation and writing for example, you're not gonna be able to do a good job analyzing what results come out, you know, perpetuating that laziness. One more thing that comes into play here, which is this. So globally, there's a massive skills vacuum of analytic skills across every industry. They can't find people who can identify problems because for almost three decades now, we've all been told, don't bring me problems. Bring me solutions. Right? So you can think, like, I, I have some friends who are teenagers. Um, my, her kids are teenagers and they know that they can ask chat sheet bt or they know that they can ask Google a question, but they don't know what to ask.
Speaker 1: (17:39)
Yeah. I mean that's, that's, those are such great points. And I, I think that people forget that, um, it's not a finished answer once it's done. And my, my experience, especially with chat pt in coming up with numbers is that it misses, it's horrible at calculating numbers. And even, uh, you know, I I would say, how many, how many dad jokes did I post? And it would say nine. And I'm like, okay, I can see that I've only given you seven. Uh, can you just tell, can you recount it? Oh, I'm so sorry. Yes, you're right. I'm upon further reflection, I can see that I've only posted seven . I'm just giving, you know, that it's, and that's truth, right? That's exactly what happens. So I think the importance of having a human at the back end to make sure that that content is what you actually want to say to the world is the most important part. And whether, whether or not, um, you're using it to, to write an article or just to come up with a solution, I, you're right to asking the question is, is the hardest part. And that's what people have the hardest part. The prompt. Yeah, the prompt part is hard. But then the, the next part is that I think that the next piece is that laziness that is involved with not actually looking at the content that was produced and just sending it to the world.
Speaker 3: (19:03)
Yeah. And that's the, what we find is, you know, we, we work with sort of two kinds of people. We work with the buyer who's usually the boss and the boss cares about making money. And then we work with the user, which is usually a, a minion of sorts. Not always the case, but, but mostly the case and the minion cares about going to lunch, right? And, and what they wanna do is just be done with it, get it off my plate check. Right? So nothing about efficacy, that's not part of the priority. And this is really frustrating for us time and time again, because these two people are not aligned in, in what their objective is. And these, this is small companies and very large companies. So it can, it's often shocking. You're like, you know, wow. Um, and so we're constantly educating around marketing best practices, AI best practices, , you know, like I didn't think I would have to be in the, I didn't think that selling marketing software would also include educating marketers on how to do their job, but it sure does.
Speaker 1: (20:14)
Yeah. Um, so let's jump into lately a little bit and tell us a little bit about how, um, how your solution learns from the users and learns from the interaction between users.
Speaker 3: (20:28)
Yeah. So we are not a large language model. We're not a language model. We're, we're the algorithm. That's what we do. And the data that we are looking at, first and foremost is your data. So we're studying your social media analytics specifically. You log in and you authorize us to go look at your Facebook, Instagram, TikTok, or whatever it is. And we're able to study everything that you're publishing over the last year and create a, a benchmark here and ra rank it, right? And then we're looking inside that ranking of the patterns within. So when, when you write, well, why, what, what makes it good? You know, what are the words, the phrases, the sentence structures, um, the grammar, even the content itself. Like is it a link in text or video and text, et cetera. And then the second pattern we're looking at is, well, what makes your unique audience emote?
Speaker 3: (21:26)
What makes them take action? Click like common and share, which, what makes them do something? And so now that we can see this, we're testing it with you all the time. We're checking in, checking back in to make sure it's true. And we're asking you the human to make it sure it's true. 'cause life changes and trends change, right? And every result that we generate for you, based on this model where we're learning who you are, every time you, you thumbs up or thumbs down it or make any change, any kind of edit, for example, the AI is like, oh, thank you for teaching me. Right? And it gets smarter and smarter. Um, and um, gosh, you asked me another question and I forgot it. I started to answer it this way, .
Speaker 1: (22:06)
Yeah. It was just about how, how your, how learning is so important
Speaker 3: (22:11)
And taking from different data sets, right? Yes. Yeah. Yeah. Okay. So the second place, so you are the first, you're my customer. So you're the primary source of energy, of, of learning. If, if you're not so good at what you do, then we go to a couple other places to help you out. Um, the second place is me. I, I got Walmart 130% ROI year over year for three years on social media. I write LinkedIn posts that get 86,000 views. So we taught lately a series of my generic best practices that worked for all the clients I used to have. And it will bolster your content with my best practices. And again, you have the option of being like, I don't like that, or I do like this. Right? Um, and then the third place it goes is to my brand. We've been marketing ourselves for 10 years through our own ai. And so the brand has picked up a bunch of best practices that I can, you know, ask you about as well. And the fourth place we look at is, I'll never share your data with anybody, but I can see the patterns of all the customers that we've ever worked with and get a general idea of what works well for other people at large, or people in your industry, et cetera. And, um, make recommendations based on that.
Speaker 1: (23:22)
You said earlier you're not a large language model. Tell us the difference in how, or tell us how you are different from chat GPT.
Speaker 3: (23:30)
Yeah, so my friend David Meerman Scott, who I, if you don't know him, I gotta introduce you to him. He'll be a great guest for you. Um, David was, I think the eighth person hired at HubSpot and he's on their board and he's on my board. So, and he wrote this book, you probably know, called cracy, which is all about how the grateful Dead turn listeners into fans or, or customers into evangelists, right? Um, so David has a great way of answering that question, which he says, there's only two questions that matter, which are who's data and who's math. And if you think of AI that way, it becomes very easy. So with lately, it's your data and my math with chat, GBT and anybody else, it's public data and generic math, right? So there can, there, it's impossible for there to be any customization at all. Now even with private data sets like those, we have a continuous performance learning loop. 'cause we have access to your analytics, but no other, um, AI works that way. So it can't check in to create a baseline and rank content, right?
Speaker 1: (24:45)
Yeah, yeah. No, yes.
Speaker 3: (24:47)
So that's the difference is like, um, and this is in part some of the laziness is because people, it's so, it's such a shame, Brent, people communication is who we are as humans. It, this is a basic function like clothing and eating and shelter, right? We must be able to do it. And the whole world just raised their hand over the last two years and said, we hate writing, we hate it. Which I already knew, , I've been doing this for 10 years, but I mean, it's overwhelming. And so I'm, I am concerned that we're going to be like going back to, I always think of, um, Ringo Star and caveman, like, are we just gonna start grunting at some point? You know, , I mean Jesus. But, but this, I, the difference is you have to ask yourself, what do I want to get done? How effective do I want to be?
Speaker 3: (25:40)
Now the basics are take out the trash, do your homework, , don't forget to pick me up a beer, buy my stuff, right? These are, every, all communication is about getting people to do what you want them to do. And when we, when we forget that, then it's a non, it's a non thing, right? So, so the, when marketing becomes creating, not even becomes, this is already the case, it is creating content for content's sake. Those who succeed know that's not the case. They're out for communication, you know what I mean? So anybody's doing it for the, so if you're typing into chat or whatever, like write me 80 social posts on, you know, great or bad dad jokes, I dunno. And analyze it for me, great. I mean, you'll, you'll get some content and it might be really interesting, but you'll never know if it's gonna be working for your audience. 'cause there's no way for anyone to know.
Speaker 1: (26:38)
Yeah, no, that's, that's a good, that's really good. And it's sort of like Groundhog Day, unless you're, unless you're giving it some instructions upfront in your prompt every single time, um, you are essentially starting over or starting with a newer model. And I think a lot of people have seen that, especially with, with chat g pt, that it te it does change a little bit over time. It sometimes it gets a lot worse and sometimes it's better and it's never consistent. I think that's the key, right? It's not consistent. Yeah. Um,
Speaker 3: (27:08)
Well there's all these dummies using it. Sorry, , but like, you know, a lot of people are putting bad information or bad prompts into it and making it stupider.
Speaker 1: (27:18)
Yeah. That, I mean, that's another really good point is, is that it, because it is such a large language model that it, uh, it is so dependent on what people are putting in and as they keep updating it and, and there's, there's a, you know, there's some public forms where you can actually see when that cutoff dates are for some of these large ones. Um, that, that input is now also connected to something that happened three, four months ago, I think, right? Maybe chat GPT is March of 22 or, or 24 or something. Or maybe it's 23, I don't know. But the point is that it, it is, um, it is essentially old data that somebody's been putting in and it, it's, it's not gonna learn what you're giving it. And that, I think the difference what you're, what you've been talking about is that you're continuously learning and you're starting with your base set of, of information that, that lately has fed it as opposed to the whole world feeding it.
Speaker 3: (28:16)
Yeah. Because I think we're, what we're all doing is we're making an assumption that, and we've already made this as assumption that crowdsourcing knows best, right? That's the assumption. Crowdsourcing versus curation and the pendulum swings both ways, right? So yeah, crowdsourcing is, can be great. It's can be very democratic. You can get lots of different kinds of points of views, all, you know, very, very valuable. But then curation has its own set of values as well. I mean, you get the, usually it's much higher, um, quality someone who's become an expert on this thing, there's a trust factor there, and you're getting a real opinion. Um, you're getting that, you're getting that j seis, you're getting the human sort of taste there, that individual thing. The, the difference I see in radio, for example, is a good, is a good one. Radio used to be a thing people by the way.
Speaker 3: (29:07)
Um, which is you develop a relationship with a curator. And when you have a relationship by curator, I mean anybody who's you are, Ben Brent, if you're wielding the mic or wielding the pen, um, someone who is designing an art show, somebody who has, um, a boutique and like the different clothes they choose to sell out of it, right? So some, some human is behind this taste, this brand, this idea, this vibe or whatever. And you go there because you ident you start to identify, it's your favorite restaurant is a gay, is a great, there's a million Italian restaurants, right? I go to, to the Red Onion, I don't know if I could call them specifically Italian, but I'm, I don't know, maybe American bistro cuisine. But like, I like the way they make mussels and they shuck oysters better than honestly anyone in the world personally.
Speaker 3: (29:59)
Um, and I'll go there for the rest of my life, , right? And I'm glad to, to pay a lot of money to be there because I, I trust that it's always gonna be wonderful and they're always gonna be nice and make me feel special and all the things I want to feel. And so you can't get that from Spotify, right? And I use Spotify, but I, I'll never get that. And so the question is, is how much, you know, again, food is a good thing because we went to fast food and then slow food, right? So it was like, oh, fast food is great, it's fast, but it's not very good for you . And so we're back to slow food, even though it can be a pain in the, a very expensive, but people want this feeling, you know, and Covid showed us, we're all willing to pay quite a lot for that now, right?
Speaker 3: (30:42)
Um, so, so AI will have the same, I mean, everything goes this way, right? Everything will have the same kind of idea. I think people need to be, um, they just need to be educated. We have this thing that we're fighting against Brendan, you know, it already, which is people think that AI is like a human. They can talk to it like a human. They just think it's a super smart human. Um, it doesn't have an understanding of those nuances to ever answer you in a way a human would. It can't possibly do that, you know? Um, 'cause it can't, it can't understand the variances, um, of the questions you're asking and that there might be some other, you know, perspective or et et cetera. And then the second thing is AI is not like AI , it's not really artificial intelligence. It's the, it's really automation. It's the ability to, um, analyze huge swaths of data much, much faster, enormously faster than a human possibly could, right? I mean, I can't go through that many dad jokes, . There's no possible way. Um, so I think when you know that, right, like that you can, a remember what your expectations are with ai, understand what it actually really is. It's not R 2D two or C3 po or ET or Arnold Schwarzenegger . It's just a computer.
Speaker 1: (32:19)
Yeah. And I, I, I want to talk a little bit about the solution and, and, and we, we also started a new company that is, you know, built around ai but involves humans. And one thing that it doesn't do or can't do is, well, a, you need that human to edit it and read it to make sure it's consumable by another human, not by Google. Because nobody, no matter what, Google isn't going to buy something from you, or they're, and they're not gonna stop into your bakery and get a donut. But, okay, so the second part is that, is that deliverability and I think, you know, lately is great at that is is great also at the deliverability part because you need to make sure you have a queue of content out there. Um, talk
Speaker 3: (33:03)
A little, we have to feed the beast. That's right.
Speaker 1: (33:04)
Right. Feed the Beast, right? And how, how people have to be in tune with the cadence of content. And you can't just, I think the other thing that people are doing now is they're dumping, you know, 20 blog posts in an afternoon, all marked, you know, June 17th , and they're, they're five minutes apart and they're not thinking about how that deliverability should go. So, you know, we, we have a few minutes left here. Why don't you just tell, tell us a little bit about the importance of that deliverability part of the content, whatever it is. It could be a social post or a video, and how that has to be, uh, it has to deliver, be delivered at the right time,
Speaker 3: (33:45)
Right? So, so, exactly. And I'm so glad you brought this up. So the old marketing adage was seven times, that was the magic number that a consumer had to see, hear, watch, understand your message before it sunk in. The brand sunk in that number is 24 now, so it's quite a bit higher. So that quantity, the feed the beast must be there, right? You do have to have a lot and a lot, a lot, a lot. Um, but you also have to have the variance because humans are multifaceted. We can smell spam and fakeness, you know, miles and miles away. We don't tolerate it anymore. So you can't just say to me, have a coke and a smile 50,000 times over. I'm not gonna respond. You have to appeal to, you know, all my different sensibilities. There's 54 sensibilities that move people to do what they do according to my friend David Allison at Value Graphics.
Speaker 3: (34:38)
And the sensibilities can be like, um, community productivity, environment, family, things like that, and just point, point in proof. Like you have to alter your message in that way. So what lately is doing is it's actually analyzing the content that you're bringing to it, and it's looking for those multifaceted messages and then lining them up for you so that you can, excuse me, you know, hit people, you know, in, in all those different ways. But the second thing is it's doing is it's taking a, a page right out of, um, Chris Anderson's book, the Long Tail, um, and helping you find the right cadence and drip feed as you were pointing out of when is the best time to, and, and be able to adjust that time, you know, to, to hit people kind of with your, with your message. So it's the quantity we talked about as well.
Speaker 3: (35:30)
And as you pointed out, the sensitivity of like, when do you unleash that quantity and how can you, you can't, you, I mean, I know it's so people want to just automate everything, but like, let me tell you one story if I haven't told you before this. Um, so Betty Crocker releases cake in a box, which we've all had before. I like the, um, the Olten lava chocolate one when it's Devil's food cake. I love that. Um, and with some powdered sugar frosting, oh, heaven. Anyways, when they released the cake in a box, all you had to do was add water. And the housewives, who they were marketing it to thought, this is too weird. I didn't bake anything. And so they didn't buy it. And so Betty Crocker took the powdered eggs out of the box and their slogan became just at an egg. And it sold like hotcakes hotcakes because the housewives now felt they had ownership in the automation in the ai, right? And so there's a, the reason, and we're talking about collaborative AI right now, is what's what this is technically called when humans and AI produce together collaborative AI outperforms AI alone, seven x there's a reason why we want you to make an effort as opposed to make no effort because the results will be, will skyrocket like cake in a box.
Speaker 1: (36:54)
Yeah. That's, um, well that's, that's fantastic. Um, and I live right across the street from the General Mills headquarters. Do you? Not across the street, but very close. Um, amazing. Yeah. Um,
Speaker 3: (37:06)
Can you smell, smell all the
Speaker 1: (37:08)
No foods? It's, it's just a giant office building. Oh, . They have a test center. Like they have a, um, you know, skunk, skunk works or whatever they call it. They don't call it, it's cool. They don't call it skunk words in the food business. I wonder why, right. No . Yeah. Um, . Um, alright, so before we close out, I want you, I want to give you a chance to do a short little pitch about lately. Tell us how we find you, tell us why we should be using it. Uh, tell us, you know, your differentiator. Uh, thank you. Yeah,
Speaker 3: (37:39)
That's so nice. Um, so we're lately.ai, we are again in the business of finding the words that will make your customers do what you want them to do. And our customers enjoy on average, 12000% increased engagement, 245% more clicks, 80% cost savings, on and on because it's ai, right? And, and we're really sort of proud of that. Um, we are a full service social media management platform. So the content that the AI creates, you can publish it wherever you want. You can automatically do the drip feed, ala Chris Anderson, like all this stuff that we're talking about. Um, and those sensitivities, you know, are, are built into it. We are, you know, really, we released a self-service product recently, Brent, I don't know if you've ever seen it, but, um, we're really trying to learn with our customers and figure out, you know, how can you make AI more accessible? How can you make marketing more accessible? These things that feel so, um, hard to do for so many people. So we're pretty excited about that and really just like doubling down on, on the neuroscience, behind the artificial intelligence that we've, we've produced. So it's gonna be a fun summer, that's for sure. And anybody can reach out to me. I'm in all the places. Kate Bradley, Kate Bradley, turn, Kate Lee. Just tell him, tell him that you, you met with me with Brent, and then I'll be nice to you.
Speaker 1: (39:00)
That's awesome. And, um, uh, will you be at the HubSpot, uh, summit again this fall?
Speaker 3: (39:06)
Oh, I can't believe we didn't meet at that. And we, we did. You came and hugged me, didn't you? Yeah, yeah. Yes. Yep. Oh, thank God. I was shell shocked. I was like so tired. Um, who, there was someone, one of my investors was there and I didn't meet him. That's who it was. Yikes. Um, I will not be at inbound this year. Um, I'm going to France on a vacation that I, I haven't taken a vacation where I took a vacation on the vacation in five years, . And I'm going to do that this year. And I'm, I've, I just started Duolingo to, uh, resurrect the French that I learned in school and I'm taking a lot of notes about how I can steal their software ideas and put it into my own. 'cause it's so great.
Speaker 1: (39:45)
Yeah. I have Duolingo Max and
Speaker 3: (39:47)
Oh yeah, me too.
Speaker 1: (39:49)
I keep thinking it's gonna, and I've been, I have a seven year streak and I speak, I can still barely speak Spanish. I have good vocabulary 'cause I do it every day, obviously. But, uh, I don't do, I think the, the next key, it, it is part of that collaboration. You need to actually practice it with humans. Not, uh, not a phone. But anyways, so
Speaker 3: (40:07)
You're on Spanish, so we can't practice together. it. Yeah.
Speaker 1: (40:10)
Sorry. Well next time . Um,
Speaker 3: (40:13)
That's my next language. I'll work on that.
Speaker 1: (40:15)
Perfect. Uh, Kate Bradley from lately, lead.ai, it's been such a pleasure. Thank you so much for being here.
Speaker 3: (40:21)
I love you. Thank you.
