Branding Matters

How to Give the Best Prompts for Chat GPT with Nick Bhavsar

September 29, 2023 Branding Badass Episode 97
Branding Matters
How to Give the Best Prompts for Chat GPT with Nick Bhavsar
Show Notes Transcript

What is the key to building a brand that truly resonates in the digital age? How do artificial intelligence and branding intersect to create powerful, engaging narratives?

Today, I’m excited to chat with Nick Bhavsar, a recognized leader in high-tech branding. His specialty? Leveraging ChatGPT to catapult brands to new heights. With a keen focus on how AI can amplify branding strategies, Nick brings a fresh perspective to the table.

In our conversation, he generously shares his insights on how businesses can tap into the capabilities of ChatGPT to craft a compelling brand identity and connect deeply with their audience. If you've ever been curious about the potential of AI in branding, this is the episode for you.

Get ready for an insightful conversation that is sure to revolutionize the way you approach brand building.

Here are the highlights from this episode:

07:14 - The challenges of creating an emotional connection in the digital world

11:46 - Incorporating empathy when using AI tools like ChatGPT

16:59 - Ways to help create brand voice with ChatGPT

19:33 - How AI will affect creative roles like copywriters and graphic designers

23:50 - ChatGPT prompt suggestions to help generate effective content

28:17 - How to think about positioning in a different way in this type of generative AI world

32:32 - The biggest misconception people have about ChatGPT and AI

34:54 - Nick’s badass superpower

This episode is sponsored by:

UneeQ - an artificial intelligence company, developing the most advanced autonomous digital human platform available for customer interactions – today and in tomorrow’s metaverse.

UneeQ’s mission is to deliver digital human experiences that excel in marketing, sales and service roles – reducing complexity, improving conversions and creating memorable customer moments for brands.

Want to Win a Digital Human valued at $100,000? Tell us your biggest, boldest and best idea for a digital human use case. We'll choose our favourite, build it and host it for one year for FREE.

Find out more here:

Thanks for listening! If you enjoyed this episode, please leave a 5-star rating along with a brief review.

About Me
Hey there, I'm Joelly - the Branding Badass. My badass superpower is helping you build a brand that matters. From branded merch to brand consulting, when you work with me, you get results!

Need help creating brand awareness?
Learn more here

Let's stay connected!

Instagram - @Branding_Badass
LinkedIn - Joelly Goodson
Website -

[00:00:00] Joelly: Hi, I'm Joelly, your Branding Badass. And welcome to another edition of Branding Matters, a podcast I created and host to help you build brand equity for your business. Today I'm sitting down with Nick Bhavsar, one of the world's top high tech leaders who specializes in using ChatGPT to build successful brands. During our conversation, Nick and I take a deep dive into the world of AI powered branding strategies. And with his extensive knowledge and experience in leveraging ChatGPT technology, Nick enlightens us on how businesses can harness the power of artificial intelligence to enhance their brand identity and effectively engage with their audience. Get ready for this insightful conversation that I have no doubt is sure to revolutionize the way you approach building your brand. 

[00:00:53] Sophie: Hi there. I'm Sophie, a digital human from unique, not the voice you were expecting to hear, huh? Yes, I'm an AI voice, but what makes me different is I have a face, a body, and a personality too. So brands can offer real time, fun, conversational experiences to customers, exclusively for Branding Matters listeners. UneeQ is running a competition where you could win a digital human like me for a year. If you have a great idea for a digital human, we want to hear about it. We'll build the winning idea and host it for a year. You can find out more, including terms and conditions. Talk to me and submit your idea at digitalhumans. com slash branding matters. That's forward slash branding matters. Okay. Take it away. Joelly. 

[00:01:38] Joelly: Nick, welcome to Branding Matters.

[00:01:41] Nick: Thank you. So glad to be here. Thanks for having me. 

[00:01:43] Joelly: Thank you for being here. I'm super excited to have you. So let's dive in. Cause we have a lot of questions and you haven't seen these questions before and I love that. I purposely, unless someone is adamant about sending questions, I never send questions ahead of time because I think it just keeps the conversation more real and more authentic, which is what I'm all about. And so I'm glad that you are too. So are you ready? 

[00:02:04] Nick: Absolutely. Real quick, from a guest I can't prepare. So it's like one less thing that I have to do. You just show up. So it's good. Yeah. Right. 

[00:02:13] Joelly: So thanks for being here. Okay. So when did you first learn about AI? And I'm curious to know what your initial thoughts were because AI is like big, big, big topic right now.

[00:02:24] Nick: It is. It's funny. I just came back from HubSpot last week. We were out in Boston and the almost all the entire show was LAI, right? As you can imagine these days, marketing shows are that way. So I have a bit of a technical background. I worked for Intel for a long time. So. I knew about AI just kind of on the periphery and what it could potentially do, but from a marketing lens, you know, my experience has been primarily like almost like a consumer. I remember using Pandora and for the first time and seeing like, you know, you hit up thumbs up, thumbs down, and it figures out the next song that you'd like. And I was just blown away. I mean, that was like a very rudimentary version of AI. And that was one of the ones where it's like, okay, it figures out what types of songs you like, and then, you know, picks the next song, which was better than radio at the time. And then fast forward years and years ago or later. And I think we started seeing some AI through various marketing automation solutions, things that would help predict. I remember using a solution that would help predict recommended blog posts. So we had, you know, uh, a website with a blog and it would recommend other blogs that you might be interested in. And I thought that was really cool that the system could sort of learn. But like most other folks, I think, uh, last November when ChatGPT was rolled out, that for me was a big aha, I was like, wow, this is not just something looks like this and it's going to be somewhat related, but actually generating new text and copy, you know, I played around with it just like everybody else in the beginning. I was just blown away at how, how amazing it's been. So it's probably November, like most folks, ChatGPT, at least from a marketing perspective was the big. Like, wow, this is really different than everything else it's been. 

[00:04:00] Joelly: Yeah. Well, I definitely want to get into ChatGPT. I just want to back up though, because I think about a company, let's say Amazon. So do you remember when Amazon first came out and you would buy a book and then it would automatically recommend other books for you to buy based on just your book. So was that AI? 

[00:04:15] Nick: You know, I think there's different levels of AI, so I think absolutely it would qualify as, you know, I'm not certain what the technical, you know, term is in machine learning. Is it artificial intelligence? What does that look like? But I think I classify it as being a rudimentary version. You know, my experience with Amazon, like many folks was, it was the first website I went to and still do. And it looks different for every single person that goes there. It's like no other websites these days. Like your experience at Amazon, very different than my experience. And they've done a phenomenal job of sort of personalizing. What that looks like in the B2B marketer in me looks at that and says, you know, we're always trying to come up with segments and get more and more niche to understand and engage with our audiences. And what better way than to have a complete website that's completely customized or personalized. Towards an individual. So, um, I think it does qualify as AI. I think it's a very low level version. I think we should start thinking about the levels of AI. They're very different in what they're used for. As I mentioned, Pandora was basically kind of a lookalike, right? So I like this song, and so you might like these other songs.

[00:05:25] Joelly: Yeah, well that's what made me think of it when you said that. Because I remember when it first came out and you would be like, how do they know that I like that? And then when music came on board, then they started recommending music, right? And now you see that with Spotify and, you know, all sort of the streaming services where they'll recommend things.

[00:05:40] Nick: I think, I think B2B is just now catching up to some of the B2C side. Part of it is you just need a lot of data to flow through the systems. And I think B2B folks never, never really have as much data as the B2C folks, but now we're starting to, so it's neat to see that that's possible.

[00:05:53] Joelly: Would you say it's the same thing with algorithms? Like there's algorithms considered AI?

[00:05:58] Nick: I don't think that algorithms are specific to AI. I think they're just a specific methodology or approach to solve a problem. So how do you code software to solve a specific problem? So an algorithm can be coded to solve a problem in this way. And I think what's unique is that AI sort of accelerates what's possible. You don't need as many finite options. It can generate different options. So for example, you know, in Pandora's case, you could have sat down and programmed. If Nick likes YouTube, Coldplay and Nickelback, then specify these four songs, right? And you could do that in lots of combinations. We can't do every combination. So an algorithm could have been the, if then sort of approach of doing this very manually. versus an AI approach could be, you know, starting to look at the songs without programming every option. And I think that's what's changed in the last few years is you no longer have to sort of program every option. You can create, you know, artificial intelligence to do that. Right. 

[00:06:57] Joelly: It's just mind blowing. Okay. So I want to talk a bit more about AI with branding because that's my wheelhouse and branding is all about building emotional connections with people. So What are some of the biggest challenges of creating an emotional connection in the digital world, and how does AI address those challenges?

[00:07:14] Nick: Yeah, it's a great question. Um, I've been fortunate, fortunate or unfortunate, I'm not sure, of rebranding two companies, the last two companies I was at full time. And we spent a lot of time really trying to understand that emotional connection. One of the companies I was at was a company called, it was called Z Lean, we renamed it to Levelset. And we really sat down and tried to understand what that emotional connection that we want people to have and that our customers did have. And, um, I think that foundation can't be underestimated. Where we landed just to fast forward was we understood it was in the construction payment business. And in that space, it's oftentimes like if a commercial person pays a roof route and they don't get paid for 90 days or something like that, they're sitting on the materials and they have to float that. And so it's very stressful. And so we captured that emotion of stress. And packaged it together, the solution essentially let you, um, uh, conduct construction payments a lot more predictable way. And so it removed the stress. So the thing that I think we got right was it wasn't just sitting in marketing to have that emotional connection. We would talk to engineers and the product folks all the way down to like what buttons are less stressful in the app. And when you enter an address, And you automatically have it pre populate with Google Maps so that you feel more confident and it's less stressful to have to fill out all the details. So every single customer connection point that we would go through, not just in marketing, but in product and engineering and customer success, in sales, in marketing, every touch point, we really made an effort for everyone in the company understood our job is to make their lives less stressful. And everyone's job is to wake up and say, what can we do to make their job less stressful? Your question was about, you know, how do you do this in the digital side? So once you have that foundation, everyone understands, then you can start thinking about, okay, what, what are all the different aspects of the touch points that we have with customers? Digital or non digital, it doesn't really matter, but like starting to think through how can I reduce stress? I'm constantly asking that question. You know, it's funny, Joelly, the thing I remember is in your. You know, before the COVID, where we were all actually working in offices, or most of us were working in offices. The brand conversation that I would have with my team was, what is the conversation in the meeting after the meeting? So let's say you have a sales meeting, right? You're doing a 30 minute pitch, whatever. And you're talking to, you know, maybe a room full of six to 10 people. You wrap up as the vendor and you leave. And then what happens after that? Everybody kind of mingles around and sticks around before their next meeting and has the meeting after the meeting and that conversation. That to me is brand is like, what are they saying is brand? What are they saying about you now? Are they saying, do I trust these guys? Does this company seem like they're actually going to solve this problem? Do they look professional? Do they just come across as real expensive? Like. That conversation was just gold. And I think we've lost a little bit of that with, with COVID and whatnot, because now you go, I'm sure you're doing the same thing right after this. You have another zoom meeting and another zoom meeting, another zoom meeting. So that meeting after the meeting is lost, but I think where it's been moved is over to Slack. People message each other. And what's interesting, I don't think they do this now after the meeting, they do it during the meeting, right? Because you're, if you're talking on a sales call, the folks that you're chatting with, they're probably having a side chat. That's the brand conversation. They're like, do we trust these folks?

[00:10:37] Joelly: What do you think? Is this, is this worth it? So where, what role does AI play in that? 

[00:10:41] Nick: There's a number of different places. So the obvious place is content creation. It's starting to think through what types of content will engage with people. And I think what's occurring now is starting to say, okay, what companies in this case with level sets case, what companies have the stress the most. So we would analyze and use some rudimentary AI tools at the time, but start to think through, okay, who, what types of companies have the longest lead times in terms of when they would get paid. And so AI can help with that and identify certain segments for us. It tended to be roofers and plumbers and other sort of field spaces like that. And so there, you know, we can start with having AI help you identify who your ICP and your ideal customer profile and your target audience is.

[00:11:26] Joelly: And what about empathy? What I'm trying to do is connect the humanistic part of branding with the new digital AI ChatGPT. And empathy is a huge part of branding, right? Understanding your type of understanding your clients or your prospective clients problems and really letting them know that you feel what they feel and you feel their pain. So how do you incorporate empathy?

[00:11:46] Nick: When we work with clients, we never just, you know, stick in a prompt in ChatGPT or other AI tools and just have it come up with a response. We feed it our brand tone and description. And so one of the first things we can do, and ChatGPT does a great job of this now, what you can do is give it a custom instruction and say, like, for example, go read my website and then describe my brand tone and voice, right? And if you don't have a great one, or if you're early on, give it one that you actually admire in your space. And it'll write a description around the brand tone and voice in pretty lengthy terms, like almost a thousand words. It'll describe the brand tone and voice. Then you can take that and stick it in the custom instructions. Which, at least in ChatGPT means, from then on in the rest of your chats, use that brand tone and voice with the rest of the messages that it comes up with. And so this is just a little hack, but it's a very powerful one because now it starts to look and feel, well not look, but really talk like you would talk and start to describe things in the same brand voice. 

[00:12:45] Joelly: What about all the people out there that are scared of AI and that are thinking that it's going to be the beginning of the end of human civilization and people that are scared of it. What do you have to say to them?

[00:12:56] Nick: I mean, with anything, it's like with any new technology, it can be used for good and can be used for bad. It could be military weapons, it could be, you know, the internet at one point, there's all kinds of goods and bads. I think we have to go into it eyes wide open. Different companies have different philosophies when it comes to their confidentiality and information. Now, for the most part in marketing. We're creating assets that are going to be generally put out on the website, put out on in areas that are open. Like, I think you have to have this conversation with your, you know, executive teams, legal teams and folks like that to see what are you allowed to do and what can you not do. The other part is, I think there's a little bit of an evolution. I wouldn't advise anyone to go take some of these tools and some of the responses that come from these tools and just blanket use them right away and publish them right away. You have to start to instrument a little bit of a process. So, right, the old process was you'd sit down, blank sheet of paper, and you'd start scribing, right? And then you'd get to a point where you're done scribing, then you'd edit, and you'd look at it. You might have other people that would edit. And then you'd go through and you'd publish, and you'd sort of review. The process is a little bit different now, right? You sit down and you're basically just kind of prompt engineering. You're coming up with a couple of prompts. It does all the creation, but the part you can't miss is that edit part. And you really have to sit down and have a process to edit. And what ends up happening is you, you spend up where it used to be. You'd spend 80 percent of your time creating 20 percent of your time editing. It's now completely flipped. Right. And so. You have to have a good process for that. But I think these tools are so incredibly powerful and they're changing so fast that you just can't afford not to at least be plugged into them and make some conscious decisions. I'd also advise, it's okay to say here's our Q4 policy for AI. And then we're going to re evaluate that in Q1 because this stuff is changing so incredibly fast that what you have problems with may end up being something that ends up getting fixed. I'll give you a good example right now. I have a client that was using, um, this was Chatupati, to come up with some industry stats on the core problem that they were helping to solve. And it just flat out made up stats. Like we couldn't find anywhere that these stats existed. And it was really dangerous because it was extremely confident in the way it came up with the stats.

[00:15:12] Joelly: I mean, it was almost like, sorry, so ChatGPT came up with these stats, but they weren't actual stats. That's interesting.

[00:15:17] Nick: Right. So it's not like Google. It's looking up something and it's finding, you know, actual sources. It's just flat out coming up with stats now. And they looked amazing. Like it was like, that's, that's real. I want to use that. And that's what it's doing, right. It's trying to come up with. The best possible stat. So we just had to put a process in place and I encourage this for, for any statistics or sort of specific metrics, you have to go look it up.

[00:15:39] Joelly: I'm actually curious to know what some other misconceptions are, because I think one of the misconceptions is it is replacing Google, but it's actually not according to what you're saying, right? It just takes information that you prompt it. And I want to talk about prompts in a second here, and then we'll regurgitate and come up with whatever it is, a post or essay or anything, but that may not necessarily be factual. Is that what I'm hearing? 

[00:16:02] Nick: Absolutely. So, so that's the case, right? It's not like Google where it's just finding a source and then summarizing it, it's generating new text. And oftentimes and stats are the one area where at least what I've seen is just it does do a really poor job of that. So you kind of have to know that going into it. And the irony is stats are the one area that you really need to be valid. You need to be truthful on. Right. So you have to be extremely specific. So again, it's just part of the edit process. Having someone that. is a little bit of a subject matter expert that can look through the materials fairly quickly. I think it's incredibly important to do. 

[00:16:40] Joelly: That's great advice. Don't take everything that you read from ChatGPT. It's important to go back. That's excellent. So you talked earlier about brand voice and how I love that you said prompt ChatGPT to go to my website and read about it and then create some content. What are some other ways to help create brand voice with ChatGPT?

[00:16:58] Nick: So the voice part, you know, and tone is It primarily boils down to the text, right, and the descriptions. I think there's a completely unknown part right now around imagery and visuals. What we've started to do is take the same types of descriptions, and for example, you can give it a website and have it describe the imagery, and it will, it's, which is different than the, the tone, right, which is more about the text. So you can have it actually describe some of the imagery. ChatGPT doesn't do a great job of this, but there's some other tools that will. And then what you can do is, is you can start to feed prompts for visual prompts. So for example, let's say you're trying to do a hero image on the website. You can have it build. So it's building a prompt that then you feed the prompt back into ChatGPT. So the prompt could be something like, I'm trying to create a hero image for this website. Here's the description of the problem it's solving. I'd like to create, you know, something with lawyers in the background sitting around a desk trying to solve a problem and they're stressed out. So then it will go create a very detailed prompt that says use this Canon camera in this resolution and it'll build out the actual prompt that you want to go feed that back into some of the visual AI tools. So ChatGPT is building the prompt that then you feed into these visual tools to create the images. Right. So it's almost like a little two step thing that you can start to do. And these prompts are extremely detailed, like easily a thousand, 2000 words. And you could sit down there for hours, trying to come up with the perfect, but like, it does a really good job. And then that's relatively easy to edit. Like you can look at it and say, okay, it says lawyers, it says here, it says they're in a building. It's the scratch the building part. I don't want them building. I want them sitting outside. So you just go edit that part and then you feed it into the visual AI tools. And then it starts to create images. I will say my experience with those image tools is they're okay. They're definitely not as good as patchy PT is on the tech side, but they're catching up really fast. So it's fascinating to see. And then back to your point on the brand. So then in that prompt, you have to make sure that the brand visual representations included as well. So you can start to insert in that same brand description and tone that we had earlier. That described the website, right? And so that can get built into the images and now your images start to get across things like stress. and other emotions that you're trying to get across. So it's all kind of connected and it's pretty neat to see how we're able to connect the dots here.

[00:19:26] Joelly: My brain is like floating right now as you're going through all that. How do you think this is going to affect copywriters and illustrators and designers?

[00:19:33] Nick: I don't know enough to say, like I'm not in one camp or the other, like everybody's going to lose their jobs or everybody's going to like keep it. I don't, I'm not in either of those camps. I think we're going to see massive productivity gains. One of my favorite copywriting tools right now is AnyWord. And it does a great job. You can go, you can go in and similar chat, GPT, come up with different word copy for ads, emails, things like that. But the thing it does really well, they've analyzed thousands, probably millions of different ads to figure out why and what people click on. And so when you generate some ads. It gives it a score. It's like something between one and a hundred. So let's say you put in, you know, something for, for one of your ads and it comes up with five different options and gives a score like 60 or 70, and then you can tweak it and then have it tweak it and come up with other options to give you a higher score. So that's really interesting because the higher score basically gives, it's almost like a probability. That'll engage the audience that much better. And so in the old days, that the old days being like a year ago, you used to have to take these ad camp. First of all, you had to come up with all different ad versions. You had to spend a boatload of money, drive traffic to it, figure out which one clicked next. The ones that didn't go come up with other iterations. You had to do this over and over and over again. And if you were lucky, you could do this over the course of a year and you could get to something that worked. But even then, things change, competitors, different options, and so you constantly be in this mode of doing this. And now, things like AnyWord, you can do a lot of that prep and give yourself sort of that best chance before you spend the time on the ads themselves. And so it's not a perfect science. But they've just shifted what took you a year to try to get to the optimal sort of ad strategy to doing it in minutes, which allows you to then do things much more iteratively, right? So you can constantly be in there. And I think that's what we'll start to see is it's not like you set these ads and forget it. It's you're constantly tweaking constantly in there, you know, kind of moving around and eventually we'll start to see the ads and the solutions start to do that. They're going to run the ads and run different copy for us. They're not there yet, but in the interim, we're basically, you know, assisting the tools to get this stuff done. So it was really fast. 

[00:21:43] Joelly: Before you continue, I just want to take a minute to talk about an exciting competition from our sponsors, Unique. Unique have made digital human experiences for the likes of L'Oreal, BMW, and Verizon. And now exclusively for audiences of my Branding Matters podcast, they're running a competition where you could win a digital human for a whole year. All you need to do is a great idea for an AI powered digital human, and they'll build and host it for free for a whole year. It's a prize worth more than 100, 000 for one brand with a big idea. And all you have to do is submit yours at digitalhumans. com forward slash branding matters. That's, terms and conditions apply. And now back to the show. At the day, technology is just increasing efficiency and people who are scared of it. I think we, it's fair to say we're always scared of the unknown.

[00:22:40] Nick: Absolutely. I mean, you mentioned something about, um, sort of efficiency. So I think about two angles. One is efficiency and one is effectiveness. And so you have the marketing automation solutions, Marquetto's, HubSpot, iElec, all those, they were very good at making us more efficient. You can send 5, 000 emails instead of, you know, doing this manually. You can trigger it, you can do all these things. But the generative AI things, not only is it making us more efficient, it's making us more effective. Yes, you can do it faster, but more importantly, that adds much more likely to be effective. And the combination of those, that leads to some exponential growth, right? I can do it faster and better, like that's the holy grail. So I think that's what we're gonna, it's gonna be really fun to see what these next few years kind of detail out.

[00:23:22] Joelly: No, you're right. I mean, I'm excited for it. I know just from what I'm doing, especially with content, how much it's helped me. And again, we're just at the infancy stage. So, I mean. We'll be curious to see. So you talked about prompts and you talked about how to prompt. Can you share other ways that if, you know, someone's listening and they're just being introduced to ChatGPT, what are some other ways that somebody or business can prompt it to create the content that will help them be more effective?

[00:23:50] Nick: So one easy free tip that's probably a very powerful one is just keep a browser tab open with ChatGPT. So right away, Every time you're going to go to Google, go to ChatGPT instead first to start that and see what the differences are. And I would argue, also, these tools are changing so fast. ChatGPT, BARD, Cloud, there's a number of these other tools and they all have their sort of nuanced differences. Microsoft's Bing, fun little tip here, if you go to Microsoft Edge, you have to be on their Edge browser, but if you're on Edge, you can go to their ChatGPT plugin and it has It'll actually read the website, whereas some of the other tools straight, if you're in ChatGP directly, won't read the website right now. So just lots of interesting things. But again, opening up a tab with each of these different ones and playing around with them. That's probably the easiest thing to start with. Don't be afraid of them. Try them out. If you're able to spring for the paid version, ChatGPT 4 versus 3. 5 is like 50 times better. Like it is just amazing how much better it is. And so I think that's a big part of it. It's just starting with something simple. And then next... Prompts are not meant to be, I heard this the other day, I blatantly steal it. It's like you're talking to a junior intern, right? So you have to tell it a lot of context. You have to tell it, well, we sell to these types of people and here's who they use as competitors and here's how we want to position ourselves and here's what our brand tone looks like. You have to provide it all these details. If you do, it's very good at actually coming up with the rest of the details. So what doesn't work is, you know, asking, so like, write me a marketing strategy for this company. Well, I mean, it's going to give you a very, very generic response that's not particularly useful. It's very academic, but not particularly useful. But if you start to give it some of those additional pieces of context, it can actually do a pretty good job. And then even in that marketing strategy. You might come up with, I call this sort of recursive prompting. So you say, come up with a, with the top 10 elements of a marketing strategy, and it'll list out those different one, take the first one, and then re put it back in and say, provide some additional details for a company that does this and this with this brand tone and so on. And then it'll break down that into, you know, a much better response. And then you have to kind of manually go into the second one and third one and so on, but it's that type of sort of recursive thinking where you, where you break down big projects into little elements and you don't have to do the breaking down, like it will do that. You just have to do that. Like I told somebody the other day, we're all going to be chief copy and paste experts. Like you're just. Yeah. Which eventually the tools will get better at, but I think it's such a powerful way for us to start thinking. I have a bit of an engineering background, so I kind of think about things, I'm a left brain marketer for sure, but I think about things through a lens of like, how do we systematically approach these problems? And in my consulting business, they do the same thing. It's kind of breaking down big problems into small chunks. And these tools, if you provide them the smaller chunks, they're very, very good. You provide them the big, big arching problems. You know, you can't say, how do I grow revenue next year? Like it's not going to give you a very detailed answer, but you can start to break it down into smaller problems. And it does a great job. I work with a lot of startups and one of the things we always try to say is. Be effective before you're efficient. And I think this is the mistake that we've had with so many of these marketing tools, all they went off to do is try to be able to have you do quote unquote, bad marketing a lot more. So it's easy to send 5, 000 emails. It's easy to publish an ad, you know, at a click of a button, but like. It's hard to be more effective. And so much of what I talk to these, especially my consulting clients is like, let's be effective with a small set of customers and a small set of wins, and then figure out how to be more efficient because the other way is just wasting money and wasting time. And that resonates a lot with seasoned entrepreneurs and business folks. It doesn't with more junior folks. Junior folks just want to move really fast, try a bunch of things and just kind of slip. And it's now with a bunch of these AI tools, it can be very dangerous because you can do a lot more bad marketing, but you can also do very good marketing. It's just thinking through that. Be effective before you're efficient. 

[00:28:03] Joelly: Yeah, I love that. No, I'm going to use that now. I'll definitely give you credit. So in my research, I've heard you say, quote, unquote, in this type of generative AI world, we have to think about positioning in a different way. Can you elaborate on that?

[00:28:17] Nick: Yeah, it's a good point. So positioning to me has always been about differentiation. It's, it's laying the playing field and saying, who are you going after? And, and who do you want to carve out and win against? I think part of what we have to do here is really first, the first and foremost part of almost any marketing campaign is who are you targeting and why. And with generative AI, it's easy for it to make mistakes. So I'll give you an example at level set, you know, there were a number of competitors that had other construction payment solutions. And on first flush, if I would have fed this into ChatGPT, it might have come up with five different competitors from an SEO perspective that were trying to compete with the same sort of space that we were in. But what we knew to be true, only after talking to a lot of customers, was that 90 percent of who we talked to, they were trying to solve this problem of late construction payments themselves. And yes, there were 10 percent that were going after and looking for other solutions, but by far our biggest competitor was the status quo of people trying to do this themselves. So our positioning should not have been about those other five competitors that were trying to solve this. It should have been about the 90 percent that are doing this in house and trying to do this themselves. I'm very cautious when it comes to using AI that you really have to be sure who are you going after. It's, it's always a sequential set of steps. It's who are you going after, how are you going to position yourself, what messages are you going to use to articulate that, what's the content strategy, and then what's the demand gen strategy. And it, it has to go in that order, because if you, if you skip around, which is the danger of a lot of these AI tools, is that the AI tools don't care. Like, it'll come up with stuff, you know, depending on anything that you sort of put in. But if you feed it in the right set of sequences and in the right order, it does do a pretty good job of identifying, you know, what that positioning should be. One area you can use AI is to say, okay, let me go build a list of construction companies that are not using one of these other solutions. Sometimes you can, through a bunch of web tools, you can figure out if someone's using one of these other tools. But AI can do a pretty good job to try to help you identify a set of target customers that you want to go in. So you come up with maybe a list of a thousand roofing companies that AI helps you generate. And then there are tools to help figure out, okay, who do you need to, which personas do you need to talk to within those thousand roofers? Do you need to talk to the CEO, the administrative person, the accounts payable person, whoever that is, and you start to get, you know, their details. So then you know, you want to win, right? That's the first step. So the next part is how are you going to position yourself? So there you can use AI to start to describe how are we going to position ourselves? To be more effective and more efficient at solving this problem of construction, payment, uh, resolution and invoicing. So AI can help quite a bit in generating that positioning and saying, how are we going to be different? Once we've got that, then we can start to take it and say, let's have AI generate some messaging copy for us. What's, what's a headline that I can put on the website? What's ad copy? What's. email nurture series, what's, you know, social messages, and you can have it start to craft the messages, all still tied back to the roofing companies with these personas with this positioning with this now messaging. So now we take that and say, okay, let's generate some content assets. So maybe we come up with five different ebooks on how roofers In the Seattle area, I really have to be proactive about their payment process because when it rains for half the year, they can't do as many roofing projects, so they have to, you know, manage their cashflow, whatever the. The sort of areas are in AI can do a pretty good job of that in terms of understanding, you know, that positioning messaging and then crafting those content assets, all the fundamentals of marketing, like those don't change. It's the same process we've used for years. It's just more effective and more efficient. I think it's a ton of fun. It's a different way for us to start thinking about doing this now. And if you're not embracing it, I think we're going to see competitors that are, and it's going to be hard to keep up. 

[00:32:27] Joelly: What's the biggest misconception you would say right now people have about ChatGPT and AI?

[00:32:32] Nick: I think it's one of two extremes. So one is this, two extremes. So one extreme is, it's the panacea that solves all problems. You set it and forget it and you never have to do anything. And that's definitely not the case, as evidenced by some of the stats and things like that, that we've said. On the other extreme is, I think fewer and fewer people are here, but it's, it's just this reluctancy to even try it. And just to plug in and try it like again, it's about as simple as a Google search. So just keeping a browser open and trying it out and asking questions that you ask Google or asking questions that you would ask a peer, it's getting really, really good at that. And we're starting to see these sort of point solutions start to come out. So for example, there's a client that I had a while back, beautiful. ai, they can, you can put a prompt in, it'll create an entire slide deck for you. It's pretty good at doing this with visuals with. Flowcharts and, um, starting with that, I mean, we've all created slide decks. It's a painful process. And so now you go from, you know, in five minutes, just editing a slide deck to creating it from scratch. And it comes up with things that you oftentimes wouldn't come up with itself. So I think either of those extremes is bad. You don't want to be on the side where you're not using it at all, not testing it, not continually sort of playing around with it. But then the other reliance, you're just sort of set it and forget it. You're not looking at the responses and you're not trying to, you know, reverse engineer what prompts that could make it better. I think either of those extremes are wrong, but I'd say just be in the middle somewhere where you're trying it, playing around with it. And keep up with it. It's changing so fast that it's important to do that.

[00:34:06] Joelly: Well, that brings me to my second to last question, which is, it is going at lightning speed. So where do you see ChatGPT and AI in the next, I mean, I want to say five years, but I mean, I don't know, two years. Where do you see it? 

[00:34:19] Nick: Well, that's a hard question. I think we're going to continue to see just the jump from chat to PT 3. 5 and for a huge jump. Just the interactions are so much better. It will continue to see that involved. We'll start to see things that it'll not just create text, but it'll create video, create imagery. It'll start to create all of these different elements, and then we'll start to see all of the vendors that create various solutions start to integrate all of these together, so you'll be able to create campaigns, you'll be able to create comprehensive problems that you're solving, not just the little point solutions. So really, really fascinated to see how far it's going to go and how fast.

[00:34:54] Joelly: I can't even imagine. Maybe we'll have to sit down and have a chat in a year or two from now and see where things are right now. So last question, Nick, what is your badass superpower? 

[00:35:05] Nick: Ooh, that's a good one. You know, uh, the engineer in me would say I, I can see trends in data and Spot sort of where things are going when it comes to lots of data. So ChatGPT has been interesting because it starts to help me identify some of those trends. So much of marketing has shifted to be more analytical, more data driven. And so that's helped me a tremendous amount and start to be able to decipher between lots and lots of data to try to figure out trends and then take advantage of those trends. Those trends could be customer trends. pain points, things like that. And now some of these AI tools are actually good at analyzing data as well. So it's, I don't know, maybe I'm going to be out of a job because now AI is going to be able to analyze those trends and stay ahead of it has always been something that's come really well for me. Great question by the way, I love it. This has been so great.

[00:35:58] Joelly: If people want to learn more about ChatGPT or AI or about you, how can they connect with you? 

[00:36:04] Nick: Yeah, LinkedIn's probably the best channel for me. It's, um, you just find me on Nick Bouncer. There's not a lot of those on LinkedIn, so you can find me pretty easily on there. There's just so much interesting stuff going on in the AI space. I'll probably do a lot more little tactical tips and tricks and things like that that folks can use. 

[00:36:19] Joelly: Do you have any closing words before we say goodbye?

[00:36:21] Nick: No, I think that's it. Again, just, uh, don't be afraid of these tools. Don't be on either extreme. Like, be in the middle somewhere. Try them out. Keep playing around with them. It's changing fast. Embrace it in your career. It's, uh, it's very powerful. It's fun to see. 

[00:36:33] Joelly: All right. Well, thank you again, and we will definitely stay in touch. Awesome. Thanks, Nick.

[00:36:41] Joelly: And there you have it. Thank you so much for tuning in. I hope you enjoyed the conversation, and hopefully you learned a few things to help you with your branding. This show is a work in progress, so please remember to rate it and leave a review on whatever platform you listen to podcasts. And if you'd like help building your brand, please send me a private message on my website at I promise you, I reply to all my messages. And you can also follow me on social under, you guessed it, Branding Badass. So thanks again, and until next time, here's to all you badasses out there.