← Marketing Against The Grain
Loading summary
Transcript102 lines
- [00:02]
Kip
Hey, everyone, if you are like me, you might get stuck sometimes. And today's show is about getting completely unstuck. We're joined by Jason Keith, who's got a great new book out called the Case for More Bad Ideas. And we're giving you five free and amazing prompts that are going to transform your creativity with whatever AI tool you use. They're going to help you write better ads, they're going to help you get better feedback. It can help you come up with amazing and out of the box content ideas. If you're stuck trying to figure out the right name for something, we got a prompt for you. They're going to change how you do any type of creative work. This is a banger of a show. Let's get into today's show. Hey, everyone. On today's show, we have an incredibly important topic that nobody talks about when it comes to marketing and AI nearly enough, and that is creativity. And what's interesting is that most people don't think of themselves as creative people. Most people don't have creative process. And both of those things are really fundamentally untrue. And I get the pleasure today to be joined by my good friend Jason Keith, who just wrote a book. And that book is the Case for More Bad Ideas. And it's a book all about creativity and the creative process. And one of the things that Jason and I did for the day show is we took a lot of the fundamental concepts and process from the book and created prompts so that even if you are a solopreneur or you're a team of one, you can take these ideas and have ChatGPT, have Claude, have Gemini, have whomever be your creative partner and help you iterate and build a creative process for any of the business or marketing problems that you might have. Jason, welcome to the show. Thanks for being here, man.
- [01:54]
Jason Keith
Thank you so much. Kip, excited about today's show and talking creativity with all these great people.
- [01:59]
Kip
So one of the things you and I talked about before we started recording is you're like, I've spent a lot of time on this topic and most people actually just don't have a creative process. And that sounds true to me. But what have you found? Why do people actually need a process to be creativity? I think we were taught growing up that just like creative ideas come to us.
- [02:20]
Jason Keith
Yeah, I mean, and that's true to an extent. But to get better at something, you really have to define a process for it. And that could be a personal process. There's lots of ways to Be creative. It's not like A plus B equals C, but we just don't teach creativity. You go to art school, you go to ad agencies. People have like the same brainstorming process that was taught, you know, in the 30s, but there's nothing past that. So I really wanted to take all these things that really creative people do naturally and reverse engineer them and really dial them back to core concepts so people can test and repeat and kind of grow their own personal way to be more creative.
- [02:51]
Kip
I think what's interesting is that on the show we talk about AI a ton. It's most of what we talk about. And when people talk about AI, they talk about it in terms of like, oh, I'm more productive, or I can automate this thing that used to take me a bunch of time. What they don't actually talk about is that AI can be a really amazing creative partner to help me just get a better quality of output, work, product, idea out in the world. And that's what the show today is all about. And so before we kind of get into the prompting, we're going to give you all some amazing prompts. They'll be down in the description below. Talk to me about like the big concepts from the book that kind of were the impetus for these prompts. Right. Like, these prompts were kind of structured off of a lot of work you did. Reverse engineering. What makes great creatives actually great at creativity.
- [03:39]
Jason Keith
Yeah, I mean, I think the core concept of the book is Bad Ideas. It's like the name of the book, it's one of the first key chapters, but it's just a small piece. And that counterintuitive, I mean, bad ideas, I say are the shortcut to creativity. They're one of those true things that are a shortcut that actually works and saves you time because it broadens the way you're thinking about a problem. When you start with creative thinking, you typically start in a small box and bad ideas and a lot of other things in this book will break you out of that. So using things like constraints, assumptions, collisions, collisions are, you know, just doing more interesting things in your life. So you're brushing up against life a little more, you're getting more inputs. And I think understanding that like the solo creator concept is a myth, like collaboration on some point is required. Whether that's with AI, whether that's with, you know, if you're writing a book, it's with your readers and your publisher. If you're selling a product, your. Your customer is kind of a collaborator, right? So there's always a collaboration point in front or behind you and just realizing when you need to break out of solo work and when solo work is really helpful.
- [04:36]
Kip
Yeah. And I think what's interesting about what you just said is, like, I found the collision part of what you just said to be very true in my life that, like, it's human nature to be like, oh, I'm stuck. I'm just gonna keep my head down and I'm gonna, like, just keep grinding through this, where it's like, I've always gotten unstuck when I'm, like, on a walk or, like, doing an activity with friends or what have you, where it's like, oh, this life experience, this different way of seeing the situation has now taught me what I need to know to, like, actually break through and solve that hard problem that I was trying to solve. Right.
- [05:08]
Jason Keith
Yeah. I mean, you've got incubation happening there, too. Meditative task is how I define that. Like taking a walk, taking a shower, doing the dishes. There's also just, like, finding richer media and sources of information, whether that's, you know, certain people or traveling. Like the guy who invented the first rolling suitcase. It happened to him when he was on vacation and carrying heavy suitcases. Right. It's him out in the world experiencing pain. And he was also. He worked at a luggage company. So he had this expertise that matches up against the life experience. And that's where we make those collisions or connections.
- [05:40]
Kip
Yeah, it's having that. Right. Like, overlap and Venn diagram of, like, life and knowledge, which is super interesting. Okay, so obviously we talk a ton about AI on this show. This is our first show on AI dedicated for creativity, which I'm super pumped about. And we're going to go through five amazing prompts for creativity today. And maybe you could pull up just the prompts and just talk at a high level. Kind of the five things we're going to be kind of talking about and working on. I've done some of them for the first time. And so I'll give everybody watching some of, kind of my feedback of what I thought and how they've worked and how I use them. But I'd love to kind of hear your take of, like, hey, this is why you made these prompts.
- [06:20]
Jason Keith
Yeah. I mean, I think not everything in the book can be turned into an AI prompt. Right. But I wanted to take the things that could be turned into a process and. And really put them to marketing use cases. So the first one that we have is collaboration Council. So this you were speaking already to the fact of sometimes you don't have a team that can work with you, sometimes you're working solo, but you really would like some feedback. So the point of this Collaboration Council prompt is to give you multiple Personas to give you feedback. I designed it kind of for website feedback, which it works really well for. Like, when I first used it, I was shocked at how well the feedback matched. Kind of what I knew needed to be fixed on my agency website, but also things that I had not thought of. But it also just gives, in this case, three specific Personas, a really strong pov, like they're not going to equivocate and be nice like a normal AI prompt would. So we have a tastemaker inspired by Steve Jobs, we have the artist Persona inspired by Maya Angelou, and we have the Skeptic inspired by Anthony Bourdain and really defined a little bit about what they feel deeply and what they react to. And those are three wildly different perspectives that. And I think getting that broader set of perspectives is always very helpful.
- [07:28]
Kip
I think what's interesting for people is, like, even when you as a human are like, I think this is a great for, like the landing page, website kind of feedback use case and a lot of user experience folks. Design folks. You went to design school, Jason. It's like they'll go do user interviews, but it's actually very hard to find, really difficult, different sets of humans to talk to. Right. Like, what you've done here is you're like, I have three vastly different perspectives. It's like, I might go talk to 10 people, but the variance between those 10 people can often be very, very small. And it's nice to have a breadth of perspective. And that's one of the things that I'm hearing so far from you, is that you have to change the aperture of both how you look at a problem, how you get feedback on a problem, to actually kind of break through and have enough change to be successful.
- [08:15]
Jason Keith
Yeah, I'm going to steal that. I haven't thought of aperture, but that's the perfect word for it. And, you know, a lot of marketers have kind of the customer Persona saved as maybe a knowledge block that they use with ChatGPT or other AI. These are other Personas that you can bring in. These are, you know, experts or mentors or creative feedback sources that can be important for. I mean, I've used it for ad creative. I used it for several use cases. It's a great. When you need that initial feedback from a broad Audience.
- [08:38]
Kip
I've done something similar as, like, a personal board of directors, like, if I'm looking for advice or whatever, like, where you pick a set of people to do it that way. But I haven't done it for, like, real creative feedback, and I'm totally gonna steal this. And we're gonna show everybody how to do all of these. So we're talking about this first one, which I really love. Let me show the second one real quick because I've been iterating on prompt number two, and I'll show you where I'm at, and you can tell me if I'm doing a good job in terms of how you thought it should work versus how I'm actually doing it. The second prompt you have is all around brainstorming. Like, how do you get a lot of ideas? And what's interesting, Jason, is I've been thinking a lot recently about this, like, next era of marketing that we're in. You know, this is a marketing podcast. I do marketing to marketers. So it's like this post AI world. What are we going to call marketing? You know, in the last 20 years, inbound marketing, content marketing, we had a few big trends in marketing. Like, what's the. The new trend in marketing actually gonna be called? And so I wrote, like, a big memo on this topic with all my ideas, with what's happening in the market, all those things, but, like, I couldn't come up with, like, a decent name at all. And you wrote this prompt that's one of my favorite prompts I think I've ever seen, which is a very detailed prompt to get naming ideas and get naming ideas in lists and rounds, you know, so you're not just, like, getting a few ideas. It's like you're getting 20 ideas and you're picking a few winners from each round, and you're kind of continuing to try to train and get closer to an example, right? Yeah. And you gave, like, kind of clear naming methods, which I thought were really good. This is, like, one of the best creativity prompts I've seen, just because of how it's both trying to get to a creative output, but it's doing it in a very, like, regimented way. You know what I mean? And most people wouldn't have put this much rigor into a prompt about naming, which is why we'll share it with you all. So I ran the prompt, and it came up with a bunch of kind of, like, portmanteau words or, like, smushed together words. And I was like, nope, I don't want any made up or combined words, you know. And so I was like, I see what's happening now. And I was like, oh, I don't like any of these. And I'm like, this is exactly how it should be working. Yeah, I should just be like very harsh with it. And then it's like, cool, I need a checkpoint that's like very nice AI speak for like you're not giving me enough information and feedback to like actually change my output. And like the feedback I gave it is like none of these words connect enough with what's happening in the market. And it's like if these were on a book cover or if somebody saw one of these in a vacuum, how would they know this was about like marketing versus like some other like obscure business concept? Right?
- [11:22]
Jason Keith
Yeah, Context.
- [11:22]
Kip
Yeah. Like some of these are like cool words like pivot, cascade. But it's like that could be about anything. Like pivot's probably more associated with like startups than anything else. Right. And so you're like, well, how do I actually make this about marketing? And so then it started giving me some like, interesting marketing related context as you're talking about it. And then I was like, well I was figured we'd pick one word and kind of follow it by marketing. Because for like the average person who's not a marketing dork, they're not going to just infer that this thing is about marketing. If you have like funnel, though, I kind of like signal to sale. That wasn't bad. And so then this is where I'm at now. I'm still iterating on it. We've changed dramatically from the first set of 20 to like the sixth set of 20 or whatever we're on right now. And I can actually see like continue to go through this and I will probably go through like a thousand names, but I think I will get to like a very interesting point where I at least have a couple of good options. Am I doing it how you thought it should be done?
- [12:22]
Jason Keith
No, I mean, I think it's exactly right. I mean in a perfect world you would have something earlier. Right. That you feel good about. But I think naming is a very hard challenge, one of the hardest because I've done tons of rebrands or naming new products and the decision maker always has a very specific lens of what they want. I mean, I think I have 10 or 11 kind of naming methods in here from compound names like you mentioned, blended names, foreign names. And people have very specific opinions both about what they like in those categories and about what's perfect for the job at hand. So this is designed to give you a variety of those types. It's designed not to fall into certain traps. Like, there is the compound word trap. Like, if you just ask AI to name things, it will use kind of compound or blended words more often than not, because that's the accepted assumption of what works in naming. So I actually reduce that down in here. Like, I try to limit it to 40% or less of the naming by default. And if you keep not picking it, it'll go away. And also, I mean, another thing that's happening here is we're trying to give you, like you said, a thousand names you might go through. That's pretty standard. Like, if you're doing a rebrand or naming a new company, you really want, like, a thousand names in the spreadsheet that you're considering, because that gives you the breadth of consideration that you need to find the right connections. And the last thing I'll mention here, I think that's very useful is you want quick iteration. So we're specifically doing 20 at a time. I formatted it so that it'll put three on a row so you can see them all at once. You can process them all more quickly.
- [13:44]
Kip
I really liked that.
- [13:45]
Jason Keith
And then once you start seeing the ones you like, you can just hit they're numbered specifically, so you can just type the number back so you don't have to type names. If you do want to give more context, like you're doing with quick sentence responses, that's fine. But if you want to just go through 20 at a time, 20 at a time, 20 at A time, all you have to do is type. You know, I liked one, write the number one, write the number four, write the number 12, and it'll just take that and run with it, because it understands that's the input that it's expecting.
- [14:07]
Kip
So, like you said, you've done this a lot. You run an agency called Social Fresh. You've helped a lot of companies do marketing. My question for you is without, like, ChatGPT, which is what I'm using in this moment, like, how long would have this process taken a few years ago to go through, like, a thousand names? Seriously?
- [14:25]
Jason Keith
Yeah. I mean, the last rebrand I did, we probably spent, I would say, 40 hours to get to a thousand names.
- [14:32]
Kip
Yeah, yeah.
- [14:33]
Jason Keith
Maybe a little less, but, like, 30 to 40 was time spent. And that's not even, you know, you have meetings and you have reviews and you. It's a whole Ordeal. But just the naming process alone can be pretty intense.
- [14:44]
Kip
Yeah. And that's over weeks if you're moving fast. Right?
- [14:46]
Jason Keith
Yes. That was over two months, so.
- [14:49]
Kip
Exactly.
- [14:49]
Jason Keith
That was multiple feedback rounds.
- [14:51]
Kip
Right. And so what do you think it would take us to get to it here? 30 to 45 minutes.
- [14:55]
Jason Keith
Yeah. I mean, I've used this a few times now, and I've gotten, I would say a core set of like five to ten names that I took and then did some actual customer research on to see if, you know, I was getting the response I wanted. It probably took me an hour max with each of the times I've used this. Yeah, it's much, much quicker.
- [15:13]
Kip
Yeah. So you're talking about one day versus eight weeks, and you're talking about an hour versus 40 hours. That's a massive time savings for, I think, what is probably as good or better of an output. It's not like you're sacrificing the output. I don't think you're just doing it faster.
- [15:29]
Jason Keith
No, I mean, I think as long as you're giving the right feedback, like one of the rebrands we did, the person got fixated on it being an actual name or like a foreign word. That's much different than the use case you're giving where, you know, you want the second word to be marketing. You know, you want contextual to the industry and the customer. So as long as you're clear about you give that feedback at some point, I think you can still get to the same quality.
- [15:50]
Kip
I agree with that. I mean, look, like you said, you're still going to want to do customer reviews and customer feedback and have some meetings on it, but it's definitely cutting out a lot of that early stage work, which is really good. Okay. The third prompt you came up with is another one that I did. And then I'm going to hand it back over to you to talk about prompts four and five. But the third one is something that I think most people are going to want to use. Right. Like, this could be the most popular prompt from today's show, which is all about ad writing, and you call it the messy ad writer. Can you give me a little bit of context on the messy ad writing kind of nomenclature? And then I'll kind of walk through what I did and see if you think I got it right.
- [16:30]
Jason Keith
Yeah, there's a couple things there. I mean, I'm trying to name things that are memorable, but I also am leaning into breaking expectations. You want to surprise people, you want to include absurdity contrarian thinking. And there's a concept in the book called fast and ugly. You know, people have heard of write ugly first drafts, but you also want to move with speed and speed isn't as important as momentum. So I think the messy perspective gives you permission to just not accept these as a final product. Understand this is like a working process and you're really just looking for inspiration to get you to the next steps to, you know, get 80%, 90% there and then refine it as a human would. So yeah, I think being a messy ad writer is a required part of the process.
- [17:09]
Kip
I think that's interesting because I think one of the best writers in the world today is Jason Isbell and he writes amazing songs. And I was watching an interview with him a couple weeks ago and he was like, when I write a song, I just write it. I don't edit it. I put it all out and then I go back and I change and I edit it. Which is basically exactly what you're saying. It's like for that initial version, you don't want the constraints cause you might edit out like a nugget. He talked about like, I might write it all out and then find one nugget and then write a new song based off of that. Right. And it's like if you're editing yourself down too much in the beginning, you lose. You lose the opportunity to do that. And that's kind of what you're saying. It's like, let's open again the kind of the aperture possibility and then you can kind of edit it down. Once you have a direction that you like that you're going in. Am I getting that right?
- [17:58]
Jason Keith
Yeah, I mean, I think that's. Songwriters are a great example of this because they break so many rules all the time and do well. They're great tastemakers. And I think songwriters will take songs and produce them without the actual words. You know, they'll just fill in with nonsense or with humming.
- [18:13]
Kip
Totally.
- [18:13]
Jason Keith
So like sometimes you just need the shell of the idea and you fill in the details. So yeah, I think again, the aperture is a spot on metaphor there. And I think just being open to a wider range of kind of risk taking ideas is the goal here.
- [18:27]
Kip
I love that in the prompt that you wrote here it's like we've got hooks and headline options for a product. And so you're basically giving the AI ad hook writing methods. So these are the ways to approach writing this. If you're an AI, right? And you're giving it basically methods that I'm sure that you cover in the book and that some of the best ad writers use. But what's interesting to me is that you have a step where it's like asking the user for feedback, which we just saw in the naming prompt, like, how do you get that feedback loop? How do you refine and repeat? And then how do you like, figure out who the winners are? And then what the person using this prompt needs to do is put in the product the benefits and the tone personality. And so I tried to do this for HubSpot's customer agent product, which is our AI agent that does customer support and handles website chat and calls and WhatsApp customer support inquiries. And I wanted to see if the AI could research and determine what it thought the benefits should be, because I always think it's fun to see what they can do. And then I said the tone should be helpful and accessible. And the one thing that is interesting that I did with this, I ran it with two different models first. So it gave me these 10 sets of options and I started with GPT. Model 0303 is the model I use most of the time when I'm in ChatGPT. And there are some good options, right? Like the 37 second fix that clears a thousand ticket backlog. I like that. That was good. Right? Goodbye. Hold music. Hello, HubSpot agent. That's not as good, but you're like.
- [20:01]
Jason Keith
Oh, wow, I like our robot. Beat you to the phone. That's pretty.
- [20:06]
Kip
I kind of like that one too. That is good. And I was like, oh, these are interesting. But I want to just be like, just run it back again. But this time I use GPT4.5, which is I'm in ChatGPT. I could be doing this in Claude. I think probably the best writing model of the ChatGPT models. And so then it's like. Holdmusic that's so 2004. You see a little bit more. Press one to speak with nobody. Because this agent never sleeps.
- [20:30]
Jason Keith
Yeah.
- [20:31]
Kip
Your support team just got infinite coffee breaks. Like these are noticeably better with a different model, I think.
- [20:36]
Jason Keith
More personality.
- [20:37]
Kip
Yeah, yeah. Like you can tell there is. It's capturing like the tone and it's. I feel like I need to give it less feedback. Right. To actually get it right and to have something that you'd be like, no, I would go take some of these and run like a test to see if any of them, like outperform what we're currently doing. Right?
- [20:56]
Jason Keith
Yeah. And I think, I mean, a couple things we did here with this prompt, I mean, number one.
- [21:00]
Kip
Yeah, please.
- [21:00]
Jason Keith
For all of these, I'm keeping the user input very short and quick so you don't have to add a lot. And part of the reason that's possible is because I'm defining a lot of the work it's going to do with those inputs and the rest of the prompt. Right. So you giving it a website and like the tone might be enough for most of these. And secondly, you know, you're at a point here where you've gotten something you like. I think some of these are very usable and you could keep going. But we've also designed that it's again, just like with the naming prompt, it's wanting you to judge these to go to a next stage.
- [21:30]
Kip
That's what I just did while you're talking.
- [21:32]
Jason Keith
Okay, cool. Something I see with a lot of AI prompts is it's just a question and answer. They don't really define a step. Two, like, even if there's steps in the prompt, you're not telling the AI what to do after that first output. And I think whether it's a feedback loop or questions asked, I think that's an important step and it really helps when building these kind of creative process opportunities.
- [21:51]
Kip
And I think, Jason, to build on what you just said, I picked three that I like from that last round that we were just talking about. And I think this feedback loop just gives you better output. Right. Like, please hold said no one ever again. Thanks to HubSpot. It's pretty good.
- [22:06]
Jason Keith
Yeah, that's good.
- [22:07]
Kip
Some of these are pretty good. Press 0 to experience instant happiness. What's the beauty of the prompts you made is you can really rev through them like a lot.
- [22:17]
Jason Keith
And iteration is so important.
- [22:18]
Kip
Yes, talk about that.
- [22:20]
Jason Keith
It seems like we're both going the same direction. I think, you know, there's two ways to scale ideas. There's ideation and iteration. So ideation is just let's put 100 or a thousand ideas on the page to start with a group of people, that's easier and that's helpful. I mean, both of these are very helpful. The more ideas you can get on the page or out into the wilderness, the better. But the iteration piece, less work typically goes into iteration for most people, even marketers. Like, we may do a second round of the blog post editing. We may do, you know, a couple revisions on the logo, but going like 10, 20, 30 rounds of iteration, that's where like creators and artists really live. And I think AI helps in both ends of the scale ideation iteration, and these feedback loops allowing you to type numbers to quickly iterate, allowing you to see more of the answers on the screen at the same time. These are designed so that people can do more rounds of iteration in less time, you know, to help our monkey brains get through these steps a little easier without getting overwhelmed. And I think that's a huge leap for again, widening the aperture and opening yourself up to a larger range of possibilities, which is how these kind of collisions and connections happen more organically.
- [23:31]
Kip
Hey everyone, if you want the prompts that Jason and I have covered in today's show, click the link in the description below so that you can go and put them into action yourself. I just want to double click really quickly for everybody. You know, if you're listening on rss, if you're on Spotify, Apple, podcasts, whatever. You definitely want to check out the YouTube version of this because we are showing a lot on video. But at the same time you just said something really important. Like the two inputs here are ideation and iteration. And you're like, most people spend a lot of time on ideation and less time on iteration. I will just tell you every great piece of content art, anything I have been closely associated with or seen somebody do. The thing that made it great was the iteration. Like the iteration and just the meticulous grind out iteration over a long period of time has what has made that in product so remarkable. And the great thing with AI is that iteration is way less laborious and way faster. Those are some of the reasons it was hard to do historically, right? It's like, oh, gosh, am I just going to manually write like 10 versions of this article and then I'm going to be happy with it like that? That seems daunting. And now you're like, oh, with AI, I can do that much faster and in a much different kind of way.
- [24:48]
Jason Keith
Yeah, yeah, exactly. Okay, so these last two, I mean, these are kind of more far afield, meaning these are going to give you ideas. They're more in the bad idea direction. They might not make approvals with the cmo. Right, Kip, you never know. They're intended to be genuinely new directions, get you thinking, thought starters. The first one we'll take a look at here is a square peg content engine. And the concept of square peg is, you know, square pegs don't go in around holes. There's a whole part of the book talking about this. It's basically trying to fight back against assumptions because the original meaning of the square peg round hole quote was actually intended to be the opposite, which is you're typically not going to show up in your ideal job or your ideal place in the world. You're going to be a square peg in a round hole. But you might be closer than you think. So the goal here is to get you to use constraints, assumptions, and really counteract those again by a little bit of contrarian thinking, a little bit of breaking expectations. Okay, so this prompt, we're trying to come up with content marketing ideas, really broadening our horizon on what that looks like. Get some inspiration. The user inputs are. I'm defining the customer as the New Jersey Botanical Gardens. I gave the website the brand voice, fun source of activities, community education, interesting events and target audiences. Families, active adults and event seekers. These are kind of standard audiences for botanical garden.
- [26:06]
Kip
Yeah.
- [26:07]
Jason Keith
So let's look at our results. Welcome to the plantasia Rebellion. Inside tips for care of plants. Things you can't do at a regular park.
- [26:15]
Kip
Oh, I actually like that a lot.
- [26:18]
Jason Keith
Mundane parks compared to botanical gardens. I love that. Your horoscope. But it's a tree. I tell you. One thing I've noticed when you ask ChatGPT for content ideas, I like these because I think they get people thinking. But it'll go to tarot cards and horoscopes quite a bit, which is fascinating.
- [26:33]
Kip
Considering it's trained on the Internet. Tells us how much of that content's on the Internet again.
- [26:36]
Jason Keith
Exactly. We asked three toddlers and a golden retriever to review our botanical garden.
- [26:42]
Kip
That's pretty good. But I like that too.
- [26:43]
Jason Keith
Secret nightlife of plants, rejected event ideas that almost made it 100% honest. Garden FAQs. Will I get lost?
- [26:51]
Kip
Yes.
- [26:52]
Jason Keith
And it's glorious. Yeah. So, I mean, these are, you know, like, these are pretty. Pretty good, I think I was going.
- [26:57]
Kip
To say in all seriousness, Jason, you and I consume a lot of content on the Internet. These are better than what most people are actually putting out there. Yes, like, they are. I mean, I will say the ideas are more interesting. You still have to execute on them. You still have to, like, have the follow through, make the video or the article or whatever you're making good. But the actual, like, hook and angle is way better than what the average person is thinking of right now.
- [27:22]
Jason Keith
Yeah. And again, I think sometimes when you just put a couple sentences out there to ChatGPT and ask it to brainstorm ideas for you, I think the ideas come back a little middle, like more often than I would like.
- [27:33]
Kip
Not bad.
- [27:34]
Jason Keith
Sometimes I get winners. Sometimes I get things that shine through. But More often than not. And you can even see. Recently I saw a study of AI use cases from 2024 to 2025. And brainstorming used to be number one and now it's dropped down to like number seven. And I think that's probably why. I think the results just aren't quite up to what people are expecting from other quality use cases.
- [27:55]
Kip
Correct. I mean, I think what I have found is that you either have to give it a great prompt or great context to get a solid answer. And if you can give it a great prompt and great context, like if we redid this and we gave it a bunch of deep research on the New Jersey Botanical Garden. These are the 10 best botanical gardens in the world. This is what they talk about. Here's a really in depth audience profile. If we did that, it would be even better. Right. It might only be 10, 20% better, but it would be better. These are still very, very good and very usable as it is. Yeah.
- [28:24]
Jason Keith
I mean, if you have other knowledge blocks like customer Persona, brand guide, examples of content you've already produced, that's. Yeah. It's huge context. And I think, I mean, these work without that context and they're much quicker than amassing all of that context. But together I would agree it's. It's going to be much better. And I can, I know that because in my personal chatgpt, it has my whole history of all my clients and all my personal work and my book work.
- [28:46]
Kip
Yeah. The context of memory, I've tried it.
- [28:48]
Jason Keith
In other chatgpts and it's. Mine is very much dialed into my personal output. Right. What I prefer my taste and I think that's a larger piece of that kind of context block.
- [29:00]
Kip
Yeah. The memory in ChatGPT, this total aside for people watching is the unwritten game changer of ChatGPT, that context and everything there. Like Jason, Kieran and I recorded like, I think like two days ago we recorded a show and one of the things is we were asking for like raw and honest feedback on ourselves. And like, Kieran kept pushing, you know, don't be nice, don't be nice, keep. And he. And finally it was like, you told Kieran, you're 10 steps ahead and 10 steps emotionally removed from everybody else. And it was like very like finally got. And Kieran was like, that's so true. I've never thought about explaining it that clearly, you know, and it's like that it's a good example of like how valuable the context of the interaction over time really is.
- [29:43]
Jason Keith
Harsh. But True. That's one of the things I did with the collaborative council prompts that we showed at the beginning is asking it to be a little harsher. Because if you do that prompt without that instruction, it equivocates. It's like, I've done the same thing with like a meal planning project in ChatGPT that I use for myself to kind of pick the right food I want to eat. And it's like a coach and I, at first it was a little soft. It was a little like, yeah, you could do that, but maybe you should do this with it. And it was equivocating a little bit. And then I asked it to be a little harsher and it's so much better. It's like, no, you cannot eat that food today. It's like, stay away from that. Eat an apple instead. And it's much, it's, you know, it makes these decisions for me. It's much more helpful with. And it has that strong opinion.
- [30:21]
Kip
Dude, I did this whole, like in depth, physical MRI whole thing because I'd like to not die. And I uploaded everything because I don't care about my privacy, uploaded everything into chat GPT and I was like, make me my ideal daily schedule for a week, like diet, exercise, sleep, everything. And it was awesome. And it was just like, it was just to accomplish the goals outlined by all the data and everything that I shared. And it was incredible. I don't know how we lived. In some ways, I'm far better at life over the last couple years than I was before, for sure.
- [31:02]
Jason Keith
Well, yeah. And to bring it back to what we're talking about, we don't even have the full context and full wide aperture that we could. Because we're not tapping into, like, data yet with these tools. We're not tapping into history in a way outside of ChatGPT that exists. Like whether you have past emails sent or blog post metrics or sales, all these numbers and pieces of context that are missing. So just imagine how much better this can get. Much less. These tools are also going to get better along the way.
- [31:30]
Kip
Yeah. So I think one of the things that's really, really important is that context you just set. Jason, this is really step one. But like you could go and take some of these ideas and put them out in the world and then once you get the data and everything come back and the next round is just going to be even better. Right. Like, this is just like a starting point. But the more you can keep the data and feedback loops in this process, the better you're going to kind of continue to make the output.
- [31:56]
Jason Keith
Yeah. And you'll build them into your own process. You'll find out how they work for you. You customize them. All of these can be customized for different use cases, so run with them.
- [32:04]
Kip
Perfect. All right, we got one more we're going to show before we close out today's show. Walk us through it, man.
- [32:10]
Jason Keith
Yeah. So this fifth one, it's a little similar to the content idea one, but it's meant to be focused on campaign ideas. So higher level marketing brain here also, this is based on kind of the bad idea methodology. This is the core methodology of the book, which is simply to start your brainstorming with bad ideas. So these are intended to be bad ideas. And how do we define bad ideas? The two core ways are obvious ideas and absurd ideas. So it's kind of a spectrum. Kind of boring and exciting even. Requiring so much in this prompt to say at least one of these needs to be dangerous.
- [32:39]
Kip
Oh, I like that.
- [32:40]
Jason Keith
So again, these are not likely to get CMO approval necessarily. They may kind of challenge your brand tone in many ways, but it's intended to get you to think. So for this, I chose UNC Chapel Hill, one of my favorite colleges. Go Tar Heels and their business school as our kind of customer and product. Brand voice, aspirational, inspirational. I just pulled this from the their actual brand voice target audience for business leaders who want to stand out more in their careers.
- [33:06]
Kip
Love it.
- [33:06]
Jason Keith
So let's run this bad boy degree in a box. We ship you an MBA in a cardboard box filled with branded pens, buzzword flashcards, hype house. Recruit top faculty to live in a mansion and livestream.
- [33:18]
Kip
Dude, that's a dope idea. I would totally do that.
- [33:20]
Jason Keith
Turn faculty into creators. Survive the simulation. Students are dropped into a remote wilderness. Only a syllabus.
- [33:27]
Kip
With only a syllabus. That's fantastic. Starlink and a piece of paper.
- [33:34]
Jason Keith
Table for one. Please run a full campaign about how lonely, isolated and self made success truly is. Wisdom for night owls. Make every class and event happen between midnight and 5am there's some pearls of wisdom in these ideas. They're challenging our assumptions. But all these take these into interesting. They all make me think of the next like how could this work? Or what's a real version of this? That could work. Build your own campus. Instead of coming to campus, applicants must build a scale model out of Legos top hat. Tuesday is mandatory. Vintage business attire every Tuesday.
- [34:03]
Kip
That's awesome.
- [34:05]
Jason Keith
And then I think this is probably the violent one. Campus bonfire, the vanities students burn an item representing their former self on day one. Yeah. So again, you can pick winners from this. You can continue to do iterations, but all of these are meant, again, to widen your thinking and to take a few more risks within your kind of initial ideation process.
- [34:26]
Kip
I also like that you have remix as an option in there where it's like, oh, I kind of like to flavor this, but remix this into a slightly different version is a cool concept.
- [34:35]
Jason Keith
Yeah. Again, you can write out that feedback and say, I like this and that, but it needs to be a little safer or it needs to be something I can get approved. Or just give it numbers and it'll keep running and keep giving you kind of a wild variation.
- [34:46]
Kip
Perfect. I love that. So what we've done here today is we've talked about creativity a lot, but we've given five really amazing prompts that I think literally anybody can use. I mean, I don't know in which somebody who cares about marketing or cares about business wouldn't use at least a few of these, if not all of these. I know that I have the Google Docs open and will be using them a bunch because I love the rigor and the kind of formatting of them in addition to, like, the frameworks that you base them on. I guess my last question for you is, these prompts are great. They're going to help people, but people still need to kind of have a couple of core, like, concepts and ideas in their head to help, like, guide their own creativity and creative choices. Like, what are the couple of things that you're like, if you at least have these in perspective and in mind, you're going to get a better creative result.
- [35:39]
Jason Keith
I think building your personal taste is an important piece of this. I don't think enough people have confidence in their ability to make these creative decisions.
- [35:46]
Kip
No.
- [35:47]
Jason Keith
So a part of that is honestly just making sure that you have that brushing up against life part dialed in. And that's so different for every person. It could be like adding hobbies into your life, like learning another language, taking up pottery classes. It could be painting. It could be, you know, vibe marketing, vibe coding, whatever. It is like something. There's a whole chapter on play in the book. And one important thing about play is to remember that play is not the opposite of work. It's the opposite of depression. Like, play is required for you to feel better about life and have inputs, and play is the practice of creativity. You know, without any judgment.
- [36:25]
Kip
Right.
- [36:25]
Jason Keith
Without any results. You have to worry about and play looks different for everyone. Like sitting down and writing a well researched paper on like the history of dinosaur discovery is something that might be really boring for some person and not get them into that play zone, but could be very exciting for the next person, right? So play is very personal. So finding those moments in life where you're. It doesn't have to be adding a hobby. It could be traveling more, it could be hanging out with more interesting people more often, really consuming things that are outside of your professional or career expertise. So if you love sneakers, just really dialing in and geeking out on those, like finding aesthetic ways in your life where you're building confidence in your taste making ability, your confidence in making creative decisions, play is a part of that. It could be board games. There's all kinds of things. There's a lot in the book about it. I think that's a missing piece and I think everyone has seeds of that, but they don't really look at it as an investment. They look at it as a nice to have if my life allows for it. So I think that's something that really is a worthwhile investment and will dial up your ability to make these decisions. And also on the more kind of regimented side of this, ask people what they like. So if you're doing a newsletter, let's say if HubSpot is launching a new newsletter based on your concept of where that marketing term comes along. Or maybe it's like, I know you have a sales newsletter. Maybe you're launching a customer service newsletter, right? Coming up with 10 ideas and literally just running it by customer service leadership people out there in the wild that either are customers or could be customers, and saying, which of these newsletters would you be most interested in reading based on the title? And you do that for 10, 20, 50 people. You know, you do the type of audience research that you can. This could be LinkedIn DMs, this could be emailing out to your list. It could do whatever is like scrappy and quick because those iterative feedback pieces are really important. And doing multiple rounds of it is even better. I think those two things. So you have, you know, brushing up against life more, doing more kind of feedback loops and whatever way you can.
- [38:22]
Kip
I love it. All right, the case for more bad ideas, it's out. We'll put the Amazon link in the comments below. So go check out the book to get the full deep dive on everything. The prompts we went over on today's show will also be in the links below. And Jason, thanks so much for coming on the show. This is a blast. I think everybody learned a lot about how to be creative in this kind of new AI Era. So thank you so much.
- [38:46]
Jason Keith
Yeah, I enjoyed it. Thanks for me having be good.