![01 | Rory Flynn on Operationalizing AI for Business Efficiency Artwork](https://www.buzzsprout.com/rails/active_storage/representations/redirect/eyJfcmFpbHMiOnsibWVzc2FnZSI6IkJBaHBCR21Ma3djPSIsImV4cCI6bnVsbCwicHVyIjoiYmxvYl9pZCJ9fQ==--cfe01beacf4f0a5b0769c1e74fb102192f9a6ad1/eyJfcmFpbHMiOnsibWVzc2FnZSI6IkJBaDdDVG9MWm05eWJXRjBPZ2hxY0djNkUzSmxjMmw2WlY5MGIxOW1hV3hzV3docEFsZ0NhUUpZQW5zR09nbGpjbTl3T2d0alpXNTBjbVU2Q25OaGRtVnlld1k2REhGMVlXeHBkSGxwUVRvUVkyOXNiM1Z5YzNCaFkyVkpJZ2x6Y21kaUJqb0dSVlE9IiwiZXhwIjpudWxsLCJwdXIiOiJ2YXJpYXRpb24ifX0=--1924d851274c06c8fa0acdfeffb43489fc4a7fcc/Rory%20Flynn.png)
The AI Marketer's Playbook
The AI Marketer's Playbook is an actionable podcast focusing on AI and marketing. Each episode covers AI strategies, tools, and trends that are changing marketing. Listen to interviews with industry experts, analyze case studies, and get practical tips. This podcast is for anyone looking to leverage AI in marketing to improve results.
Tune in to stay updated on AI marketing.
The AI Marketer's Playbook
01 | Rory Flynn on Operationalizing AI for Business Efficiency
In this episode of the AI Marketers Playbook, host Audrey Chia chats with Rory Flynn, founder of Systematic AI. Rory shares his journey from e-commerce and email marketing expert to AI specialist, focusing on how AI can operationalize and amplify business productivity. He discusses the practical applications of AI in streamlining workflows, from creating brand assets to automating repetitive tasks like note-taking and CRM updates. Rory offers insights into his favorite AI tools, including Midjourney and ChatGPT, and provides a detailed walkthrough of his ad generation process.
Tune in for valuable insights and practical advice on operationalizing AI in your business.
Join my weekly Newsletter: https://lp.closewithcopy.co/welcome
welcome back to the AI Marketers Playbook, where we cover actionable frameworks to actually leverage AI in business. I'm Audrey Chia, your host, and today I have with me someone whom I deeply admire and respect, Rory Flynn. Now, Rory is the founder of Systematic AI, where he specializes in operationalizing common AI tools for businesses with the goal of amplifying productivity, impact, And efficiency. And prior to systematic AI, Rory was an e commerce and email marketing expert, helping over 100 clients increase their sales by over a hundred million dollars. Rory, I am so glad to have you with us.
Speaker 4:Audrey, it's awesome to be here. There's a long time in the making, I think, but, no, really excited. I'm glad to, glad to be here and hope we can dive into some really cool stuff today.
Speaker 3:Awesome. I think for our listeners, it would be really great if you could share a bit more about your background and how you started with Systematic.
Speaker 4:Totally. Yeah. So, I come from the email marketing world, the agency world, a lot of it focused in the e comm space. you know, this was something that I wasn't expecting to be in AI. It never really was like, Oh, I saw, you know, this tool online. I have to be really good at it. It was more so a necessity. So back when we were doing, you know, A lot more, you know, or when I was with the agency, we just had needs. There's, there's no other way to put it. Like we had a very small team, we had a big client base. We had a lot of demands. And at the time we were kind of drowning to be quite honest. So it was something that we utilize. We, you know, happened to just see something like Midjourney pop up on Twitter, you know, if everyone's not familiar with that, it's an AI image generator, but it was like, Oh, this thing can do a lot. Let's just go play around with it. And, you know, First, it started with me creating, you know, like aliens and spaceships like everyone else does, but, that quickly turned into, wait, can we do this for brand assets? Cause we realized that something like email, right? Like was a perfect medium for this to begin with, because number one, it was small format because at that point you couldn't really upscale any images in AI 16, you know, 18 months ago, almost, and I don't know if anyone here goes and checks their. Old emails in their promotional inbox. I don't think anyone does. So it's a really short lifespan, about like eight to 10 hours at max. So there's a great opportunity to test and it could give the hell has tell better visual stories within email. And we just didn't have to repurpose old assets over and over and over and over again. So once we. Sort of nailed that down. It was like, where else can we use AI? And then that kind of opened Pandora's box to, you know, how we started. Can, you know, can we plug this into operations? Can we plug this into, you know, account management? Can we plug this into, you know, the onboarding process? So it kind of just snowballed, but it started with one little problem. And then before you know it, you have a ton of problems. Then there's AI for that. So
Speaker 3:I think it's all about like figuring out where can AI help you in your workflows, right? Because a lot of people, like they get so overwhelmed by all the different tools, but actually it's like starting with the problems and then figuring out how can I solve this tiny problem? And then putting things together to, to form, you know, a much bigger system.
Speaker 4:Totally. Yeah, I think it's I think there's this big misconception, right? I think a lot of people are just looking for AI to do 100 percent of the work and it all be like finished product. Like, oh, I need to create my ads 100 percent with AI need to create my copy all 100 percent with AI. But like most of our problems that kind of Right. Don't allow us to create that stuff in general, start on the operational side. And it's all little things that you don't even think about. You know, sometimes it's if in the agency world, I'm sure there's some people in the agency world listening, you have account managers, you have, you know, sales reps, things like that. They're on calls all the time, right? Like all day, every day. And you know, you have 10, 15, 20 calls a week. Right. I think about certain things that used to frustrate the hell out of me, which was, we have to take notes on these calls in hand, you know, in handwritten form and then transport or transfer them to a digital medium, it's like such a simple process that no one probably thinks about too much, but how do we, cause that, you know, if you're going to take your notes from a 30 minute call, it's going to take you time to organize them, then put them in a CRM or somewhere else that everyone else can have access to them. Part takes 10 minutes. You're doing 20 calls a week. It's 200 minutes a week, 800 minutes a month, 9, 600 minutes a year. It's like that across 10 account managers. That's, you know, we're spending almost 10, 000 minutes a year, moving something from a pad to a, to like just a computer, which is crazy. So you look at things like fireflies or any of these other, you know, automated note takers. You can build a simple zap from Fireflies to your CRM platform, Monday, Salesforce, whatever that might be. And you just saved maybe 10, 000, you know, like 10, 000 minutes per person. Right? Like, it's just crazy that that's, that's available to us now. And I, that I always have to plug Fireflies cause it's like changed my life because I can't take notes and do calls at the same time. So simple things like that, you don't even think about, but that's just one little issue. That's one little task in a business. There's so many. So it's just starts small. Yes, definitely start small.
Speaker 3:You know, I think it's all the, the little things right, that add up that people don't realize that takes so much time from our day-to-Day workflows and a lot of creatives. And because we are both from an agency background, some of them might be afraid that AI is gonna take over their jobs, but at the same time, if they. See in a different perspective, it frees up so much more time for you to do more. Like you just have to shave, like shift your perspective. Right. And then you realize how much more you can do, but I would love to also ask you like, how exactly are you perhaps using AI in your own business? give us examples. Like how can someone get started with it?
Speaker 4:Totally. Yeah. So of course, you know, from, from my perspective, a lot of this is doing a lot of organic content. So naturally having a partner to help create images for me is great. I don't have to use stock sites anymore. I can really tailor my content to what I want it to be. I can give it a very unique and distinctive look. So it stands out, especially on a, you know, platform like LinkedIn, right. Where, you know, there's a lot of corporate material on there and then I'll just throw a picture of, you know, a face with neon glowing tattoos on there. It's a, Definitely something that most people haven't seen today. I would assume maybe, you know, maybe a little bit more, more so, you know, five, six months ago than now, but you know, that's, that's one way. I mean, that's really simple. Even just using, is it a cover for a carousel or, you know, an asset for a website or something you're going to put on organic social, that's something we've been doing, but you know, we've really tried to systematize some of these processes, especially in, you know, I come from this e commerce world, we work a lot with media buyers. You know, they have a lot of issues, you know, just like every other role. But one of them is always just a lack of assets. You know, their, their whole opportunity and their whole basic function, right? Is they have to, they have to put ads in the right place and make money off of them. So, you know, what really happens there is oftentimes they work with fatigued assets, stuff that's just been utilized over and over and over again. They don't have enough variety to test and then to optimize so that they Then, you know, again, the ads don't perform as well, so they really just need new creative to cycle through. So we've been focusing on that a lot, how we can amplify their creative load. A lot of the time it's in static still, you know, fairly decent split in terms of, you know, what, produces on some, on a platform like Meta, you know, that's really a big part of their ad sets. So yeah, that's what we've been, we've been focusing on really just finding again, like these little tasks. Systematize, you know, like building a process around it, building S O P S around it and then systematizing it so that it's easy and it can help, you know, variety of teams throughout an organization. So it's, you know, that's one of, I don't know, many things. There's, there's a lot of other ways you can go. I really like, you know, outside of the outside of just like the image creation world is in the analytics side. That's really where I find some tools like ChatGPT to be insanely helpful. you know, finding different insights and, you know, key trends and patterns that, you know, something like Shopify or something like Gumroad or LinkedIn, even itself, their analytics are very, I mean, face value, you can't really find really intricate trends without paying for a really high end software or a consulting firm. So, you know, having something like that at your disposal is insane. And what it can pick out is insane. So yeah, I love, I love it for that. I love it for helping me build strategy and just being a thought partner there.
Speaker 3:So much value to unpack in there. I don't know if everyone is catching this, but you mentioned so many different use cases for AI in just one brand. It's true. Like most people would be like, Oh, I use it to, you know, Right, one headline, but for yourself, you look at the way you, you, you break it down, right? There is a system for thinking, a system for analysis, and a system for creating work. And there are so many buckets in which we can use AI to really enhance our day to day and also like how we can really streamline certain workflows. Let's talk about Analytics, right? I think this is something nobody really talks about unless you have a more data driven mind. How would you use AI to help a business break down certain metrics? Like what kind of metrics are you looking for and what tools do you use for that?
Speaker 4:Totally. So I'll give you, I'll give you a great example. It's just something that I do on my own, right? Like I, I sell digital products on Gumroad and Gumroad is a content creation or a curation platform. Basically it's just sells, you know, you can buy and sell digital content on there. Now, I like to think about this in terms of my, My profession or my experience, which is in the email marketing side, it's really hard for a lot of people to understand the email marketing. If you haven't been in it, cause there's so many moving parts, but yeah, I I'm basically taking what I want. What is the end goal that I want? And then reverse engineering it, right? So for me being an email marketer, like automation is gold for when it comes to when it comes to marketing communication, but that automation has to be good, right? It can't just be, let's just throw out this random stuff left and right and see if it works. Now. The way I've utilized analytics there was like, for example, I'll break this down granularly, right? I essentially, when I look at something like gumroad, I offer two different types of products. One is free. One is paid, right? So naturally, you're going to have a lot of people that download for free, and there's going to be people that buy stuff. So. To me, I'm looking at analytics because I need to, you know, Gumroad doesn't have the best analytics for this stuff. So, yeah, I need another thought partner to actually go through it. So I'll plug in, you know, look at my different sales trends. I'll obviously not put anyone's information into, into something like JADGBD. I make sure to go and create unique identifiers for these things, but essentially I'll take a look at what's the percentage breakdown of people who. Download free products versus by paid products. Then from there, you know, as chat GPT, like what's the average or what's like the typical conversion rate between someone who buys a something for or something who downloads something for free for something down, who pays for something right now, I get like a conversion rate. So I know that that works. Then it's okay. How often, how long does that typically take them to go from a free download to a paid download? Right now I get a timeframe where that's, you know, three to four days. It's typically the. The sort of situation there that, okay, what typically like in terms of like the buying sequence, what do they download for free? And then what are they purchased? What is the common trend there? Now, this all sounds like it might be just random thought process, but now I've just built an automation. I know how long it takes for someone to go from free to paid. I know what products I like to buy in sequence. Now I can build an automation around that. Right? So now it's okay. Whenever they purchase X, now I send an email at Y talking about this, right? So that's something that wouldn't be available to me to synthesize through 5, 000 purchases over the last, you know, couple months and be like, yeah, I can do this with a Google spreadsheet, right? That would be that would take a Google, I'm like a, you know, Excel expert or something like that. so now with tools like, you know, chat GPT, which is completely objective, it's just looking at the numbers and computing. I tend to think it's better at that than almost anything else. Cause it's not really, you're not really asking it too much of it. It's to do like it's baseline functions, but it's almost like having a 360 conversation with a, with a spreadsheet, right? It's like a holistic four dimensional conversation. You're not just looking at a spreadsheet, which is 2d. And it's like, Oh, what does formula in? It's like, well, how can I leverage this? Where else do you see opportunities? What would someone, you know, what would be some insights that a typical human wouldn't catch? Or what's some advanced analytics that we think we should look at? It's like your partner. So, you know, that's a really long answer, but I think it's a good way to think about it because that's not just for email, right? You can do that for a lot of different situations, whatever you're trying to reverse engineer the outcome and then utilizing the tool to get those analytics that you need. So that's a, yeah, I think that's my favorite usage of it. I'm also just not really great at getting it to write like me. So maybe I'm biased.
Speaker 3:Well, you just don't know the tricks for that. But yeah, I love that you said that. Sometimes like people think that chat GPT is only great for the output, but actually it can help in like the higher level thinking and strategic part of things. So like, even for me, what I get it to do is to. Analyze a lot of my audience transcripts whenever I have customer testimonials or in the mix And then you can really pull out the key and common insights that you find Across the different kinds of chats right because each chat is one hour long And as a human it will take you so much time to process all the information, right? So even that bit of like Using it as a partner to help you to think of ideas or concepts has been very effective for me. And speaking of like, you know, workflows, I was thinking, how could, you know, an ad agency streamline their workflow? Like, do you have an example of how you would break down their processes and make it like 10 times faster with AI?
Speaker 4:Totally. Like I said, look, there's, there's. So many different functions within a team, right? Like everyone has their own tasks, every person. And, you know, if it's a five person unit, everyone's got a lot of tasks on their plate, which ones can we look at to eliminate time? Cause that's all, you know, essentially this promise of AI, right. Is that it's going to reduce time and amplify productivity. Probably save us some costs if that's, you know, again, depending on what tools you're using. But, you know, again, we're looking at functions, right? So I look at media buyers. In a sense, like they're, if they're doing static ads, number one, they have to create the ads, right? You have to, either one, you're using some canned platform, which might not get the best, you know, results. You want to be more, you want to have more control over the output. So, you know, we look at simple things, like even copy pasting, you know, copy and images into an ad template. Things like that, that might take a really long time for a designer or a media buyer. Whoever's doing it, simple things like that, that could take forever just to format this stuff. so, you know, we've tried to attack that process. Number one, you know, teaching people how to create different unique visuals with They're, you know, with the AI tools, you know, utilizing brand imagery, to sort of hack their own identity and create unique either lifestyle or product photos from that. And then, you know, from there, it's once you have the assets, how do you amplify it? How do you get the ad creative, right? Like, that's cool to have the assets, but how do you get more out of them? And how do you do it quicker so you can get this stuff uploaded? Cause you know, media buyers, it's almost like day trading back there. you know, when you're in the platform like meta, oftentimes, you know, you're, you're killing ads throughout the day and you need to add different ads back into the, you know, into your ad sets and things like that. So just how do we make that process more efficient? So you don't have to worry about just wasting money essentially. So yeah, it's, you know, I can, I can definitely show you the little bit of that, if that would be, if that would be useful for the audience, I can, show you a little bit how we break that down.
Speaker 3:That'd be amazing. So if you could share with us, you know, your breakdown, and of course you have some listeners who can't see the slides, so we can just talk them through your process. I think we will gain so much value from it.
Speaker 4:Totally. All right. Let me, let me share my screen real quick. This is okay. Screen. Let's do this. Okay. Can we see that now? Good. All right, cool. So this is typically how we're breaking this down. Like this is just one process, right? Like this isn't the Holy grail. This is just how we do it for some of the, you know, especially on the static side, right? So I think, you know, one thing to be cognizant of is being creative within your own workflows, right? Like your workflow might look different than my workflow, but this is typically how we do it. So when we're starting this. Essentially what we're doing is we're building templates out. So we build these templates out in Figma. Oftentimes we're not going too crazy on them because we're in such a, depending on what phase you're in, in the media buying standpoint, right? Like you could be in testing, you could be an optimization, whatever. oftentimes we're just picking winners like previous templates that have worked. We know have good data. On them. So we're either one building them out to, to mimic that, or number two, we're utilizing the same ones, just maybe changing some different elements. So we'll build out these templates and Figma. And I think I should go back one thing here just to show you how like, actually this breaks down. So if let's just say you wanted to create a thousand ads, do you ever, do you need a thousand ads? No, but I'll show you kind of how that looks. So it's, you know, basically. If you wanted a thousand ads and a lot of times we're just talking about variety and small changes in ad sets. You don't need to have 1000 unique ads. Maybe it's an image with a couple different sets of copy. Maybe it's a couple, you know, one set of copy with a bunch of different images, right? Like, or you want to test CTA buttons, or you want to test different, you know, across different aspect ratios, things like that. So when we're doing this, if we were to do basically only five images. Five different headlines or sub headlines and five unique call to actions. And if let's just say, you know, again, outside of meta, if we were doing this across eight aspect ratios, so you think YouTube, Tik TOK, Instagram, Google, whatever, those are different mediums that you're advertising on. Basically, if you just made a, like every combo, every possible combination of those, you'd have a thousand different. Ads, right? So again, this could be 200 ads. This could be 20. It doesn't have to be 1000. I'm just using that number because it sounds cool. but really, we've done this for, you know, again, at that at that level. So the five tools that we typically utilize for this number one is we use Midjourney or stable diffusion for images. Those are, you know, again, to create assets. You can also use your own assets does not have to be AI generated. A lot of times we're using chat GPT for copy to either one, get us really close and make small edits, or two, just to generate in full. Then we use Figma to build the ad templates. Then we use this, this Figma plugin. It's called copy doc. I would suggest for anyone that is, you know, moving a lot of data and trying to create a lot of things at the same time. It's a really good. plugin, it's free and it's awesome. And we'll go through what that looks like a little bit and then any sort of like Excel or spreadsheet software, right? So that that's typically that'll all work. Now, the process that we're looking at, right? Is like I said, we create these templates in Figma. Then we go and we either generate our images in Midjourney slash stable diffusion slash any other image platform. and then we'll utilize, you know, real images as well. We generate our copy in chat GPT. Then we organize all of that into a spreadsheet, which is probably going to take you the longest of all this process, which might be all of 15 minutes. And then we will generate the ads in Figma while running the copy doc. Plug in, so you look at this, it seems pretty straightforward, right? It's not crazy when you go into it. Basically, like when we're generating templates, like I said, we're using high performing templates with the structured aspect ratios. You know, we add placeholders in here for headlines for sub headlines and then the CTA. So basically, if you're thinking about how you can link between a spreadsheet. And, Figma template, right? Like you want to have the city, they need to connect, like they need to talk to each other. So you have to have, you know, different places that, you know, the, the data can be shared across. So when we do that, typically then we'll generate images, right? Like if we don't have images to utilize, we'll go and generate the images. Like we'll talk, we'll, we'll hack our little brand identity. So this can be very translatable across images. you know, pick winners that we want to utilize, get them loaded up. Then when we generate copy. Right. Like everyone's process is different here, really. You doesn't have to be, you know, you can go through your chain of thought prompting, you go through your mega prompts, whatever. I'm not, I'm not an expert on, you know, chat, GPT copy generation, but I know how to get what I want out of it. So it's, basically, you know, we'll, we'll feed and prime the, the system with a number of different things. A lot of it is utilizing previous data, brand guidelines, you know, different prompts, That we've utilized in the past to get it there. But I think some of the most important things, especially in the ad realm is focusing on copywriting guidelines and restrictions, because we all know chat GPT likes to get a little bit of verbose at times. So, you know, again, like a, a CTA that's six words, isn't going to work. Right. Like it needs to be, you know, CTA is less than two words. A headline with 13 words. Isn't going to work either. So maybe, you know, headline in less than five words, but we're utilizing all these different things to kind of narrow down that focus, like, you know, brand guidelines, tone of voice, you know, different examples from brand marketing, target audience, product focus, which is also important. Like what the benefits are, the products, things like that. So we'll get copy. I think, like I said, everyone has a different process for this, but you know, this is how we we've been able to generate pretty consistently. Now, this is, like I said, is the most boring thing that you will probably ever see or do in this process. And it's just organizing everything into a spreadsheet. You can also have ChatGPT do this for you. But, like I said before, we're just setting up these, these different headers up here with whatever they're linking to in the template. So, if I go back. Right. You know, we have our headline or sub headline, we have our CTA, right? So those are all going to be linked together. And then we're just putting the different copy sets in across however many different ads we're doing. We also put the image name for whatever image we want to link into those ads. So that, you know, again, and then also the template, wherever, which template it's going to the name of it. So now when we do this, right, basically what you're, what you do is you take. Spreadsheet. You export it right into a CSV file, and then essentially from there, you zip it up with the images, into a, into a zip file with the images that you're going to utilize and you head over to Figma, you highlight the templates that you have, and then you just press, you know, you run the copy doc plugin and press, press sync. Just like that, everything will go from the spreadsheet to the ads. Right. And then from there, they're in there. So no copy pasting, no dropping images in, resizing them, everything like that. They will go into it. So it's really a time saver when it comes to this. I think, you know, it's not like we're, we're stealing creativity from anyone here. It's, I don't think any designer wants to go and copy, paste a bunch of texts and reformat it a bunch of times. So now it's all in here. Right. And it's, you know, you can do a thousand ads. In less than an hour, a thousand different variations, because variations are again, super important, especially when it comes to copy. But, you know, simply just like that. Now they're all, they're all in Figma. You can go in there and you can edit. You can edit the, you know, you can trim the size down. You can maneuver the images. You don't have to worry that these are just like done and you can't actually maneuver them. So it's just a way, like I said, if we're going to do, you know, Basically like a thousand ads, right? Think about how much time that would take to just copy paste every single thing. Resize stuff, add the photos, add the, you know, add the copy. It's great for testing. It's great for getting these different variables, you know, out into the marketplace. It's also great for localization, right? Like I think this is something that is pretty, you know, it's not something that would have been accessible, you know, a while back, really, we have the ability to start talking about, again, even from an image perspective, right. If you're a multinational corporation, you know, if you have an American model eating a hamburger, you might not want to run that in Asia, right? Like maybe that's not, this doesn't hit the same as it does with AI. We can slightly alter that photo to be more local to the market or copy wise to, right. You look at something like, you know, how we speak in New York might sound totally different than how we speak down in Charleston, South Carolina, utilizing different vernacular and things like that. That's where chat GPT can help, you know, utilizing common words, things like that. And it's just, you know, it's major from a time saving perspective, but like the, the thing about utilizing a process like this, it's not, it's not. All AI, which I love, you know, it's just how to take your AI stuff and then amplify it. Right. So really like that Figma process, you know, that's, that's replicable, not just with any AI generated images. You could take one winning ad and turn it into 50 different variations, right? You could take winning copy and test 10 different photos or 20 different photos. You could take one winning, one winning image and test it. And test 50 different pieces of copy, right? So it's a great way to do it. It's not, this isn't just, you know, really limited to ads. You can think about this for YouTube thumbnails, social content, you know, anything that requires just mass production in volume, and that's how you can really just think about. Utilizing the tools at your disposal, things that your designers already know, things that you already know and just, you know, really take it to the next level. And we're saving time on little things. Copy pasting, right? Like that, who would think that that would take hours in a day, but it does. So, yeah, long explanation, but I think it's,
Speaker 3:Beautiful flow. I think as someone who runs ads myself, right, and because I'm also creating ad creators, I know how much time it takes to do like the small tweaks and sometimes when you're doing A B testing, all you need is a tiny headline tweak or you just need a second visual that maybe pops a bit more and you'll see your conversion rates just drastically change. With these simple tweaks so for you to be able to piece together different parts of a process to make content that can be tested and experimented with and optimized in such a short period of time I think that is actually the real magic in ai And I also wanted to talk about the copy part of things because I have been testing a few different tools For copywriting for ad creatives and interestingly chat GPT is the worst It's true, we have come a long way in fact right now, Gemini and Meta AIR HD I, for now, two of the best tools I have seen for ad copy. And you should check them out.
Speaker 4:Interesting. I definitely will. Now, Meta makes sense to me now that you say that, because I mean, where's their data coming from? I think we, I think we understand that, but it's a, that makes a lot of sense. You know, it's, I've always found ChatGPT for whatever reason. Anything that's longer than a couple words, to me, it starts to sound generic. I mean, obviously you can manipulate it if you get good. I've done that. But when you get it to short and punchy, to me, it's, it tends to, it tends to do very well because it's just like, it can't go off on its own little adventure. It's like, here's the, I need it. And he did it with five words. All right. There's no fluff, so no delving, no transforming, you know? so yeah, I'm going to have to tell, I'm going to have to go check out meta now. Cause I didn't even think of, that's a great point. I didn't even put that together, but I think that's really good.
Speaker 3:Ad assets, right? So when I tested it, I just use a very simple prompt, a very basic one to take a look at the output and the copy was formatted in the style of top performing ad copy, which was why I was so impressed because the prompt was so short, whereas for me to get there with ChatGPT, I have to give it so many, like what you said, parameters and restrictions, which then makes me like guiding the AI to get to where I want to be. Already knows. What you want so I think that could help to close up that gap for a copy for you guys
Speaker 4:That's amazing. I'm going to go, you know, now, thanks, Audrey. I'm going to go down a rabbit hole once we jump off here and really, you know, this will be the next six hours of my life. Yeah.
Speaker 3:But I think one, one thing that I would also like to ask you about is actually starting an AI first business, question like for someone who is going or trying to go into the AI space, how would you suggest that they build a business around it? Because again, it's new and some clients may not be as receptive to AI. And of course we are evolving. So more and more people are open to it, but I think there is still some resistance out there. So how do you overcome it? And what is the model that has been working for you?
Speaker 4:Do what you're already good at. So I think, you know, if you're a if you're a copywriter and you're trying to open an AI design business, it's not the same thing. So, you know, that's why it's, that's why I don't wanna say it's, it is worked for me, but it's been, it's told a better story if you want to,'cause I'm looking at it through the terms of what I've already done. So I've done email marketing, I've done paid media, right? Like, I know those industries in and out. I know the common problems. I think it's a big thing, right? Especially if you, you know, you're a copywriter, you know where the problems are with copy, you know where the problems come with, you know, your clients and, and the copy the that you guys write for them and things like that. So it's way easier to translate cause you're just making yourself almost bionic, right? Like you're, you're adding an extra layer to your already, you know, already really decent skillset. So that's how I look at it. I mean, from a really niching down to start, I think that's actually. Better than going more mass in general, you know, looking at again, something like where we started was email marketing. It made a lot of sense. We were only doing images for email, but that's because we knew what we needed from the images out of email. We knew what we were trying to do before that started to get into, Oh, like we could actually trim the process of working on email. so, you know, it started with just getting to know one tool, one tool that was really fun, that was really helpful for. your own individual workflow and expertise, getting really good at that and then realizing there's a lot more out there. So, you know, I always tell people ask me this, you know, why, just like why basically it's like, well, number one, it was a necessity, but two, I knew, I just knew the pains. Like, I know what it's like working in an agency. I know, you know, the back and forth and the revisions between clients. I know the never ending demand for creative. That's just always been there. So it's just like, how do we just take that problem And turn it and utilize something to, to solve it either faster or with higher quality. So, you know, I think those are, if you're looking to start something, just look at yourself first. You know, there's, I think a lot of people go out there and search for market opportunities and it's like, it's a great opportunity, but you're not really interested in it. And you don't know a lot about it and the time to get to that point of expertise, you've already have a lot of experience. I've already done a lot of things, probably worked with a bunch of bad clients and know what it's like. And, you know, I've been burned a ton of times from different things. So that's, there's any piece of advice I have is like, how do you just take your arm and like inject like a, you know, basically build a robot arm. Like how, like what, what's going to make you better. What's going to amplify you. if that makes sense.
Speaker 3:Didn't take away at least from what I got. Right. So the first thing is five problems that you want to solve for yourself. And I think that's how you create a product that is. Painkiller and not so much of a vitamin because if you have that pressing pain point, There will be other people who are like yourself trying to solve the exact same problem and the second thing that I really liked about what you said was You already are an expert in a certain field or niche, right? So AI should be the thing that you use to supercharge that because you still need that foundations of you know, being an expert in your field and knowing your craft You and learning like how to use AI to maximize that craft. So a lot of people think that AI is the solution, but I think for people like us, we realize that the sweet spot is the human AI synergy. And that's where you really see like the magic happen.
Speaker 4:Totally. I think it's, I think it's also too, right? Like when you're, when you're selling, let's just say you're selling. If you're talking to an audience that is totally foreign to you, like if you're like, I'm going to be in FinTech and utilize AI, but I've only been in fashion design, right? Like when you're talking to the audience, it's not going to sound the same as if, you know, you're coming from a copywriter talking to a copywriter. It's like, there's a leveling of the, of the playing field there. And, you know, And they know that, you know, so it's like, okay. Cause it talk about these little things. Like we know that, you know, for on the ad side of things, people just, they have less creative and it kills them because they're just like clients are like, just run it again. It's like, well, that doesn't make money. Like just run it again. Things like that. And it's just that that's a pain point that kills them. And like, so, you know, there's, we know that we can build products around that. We can build services. Around that. So to me, it's just really hard to learn an industry, learn a tool, learn a workflow. If you're good, if you're trying to get something started quick, that's the wrong way to do it. If you like, you don't realize how many things bother you until you start looking for ways to solve them, just like arbitrarily, and then you're like, damn, I do. This I hate this little task. This really annoys me. That communication annoys me. So it's, you know, okay, I can fix those. So, yeah, that's, that's the other thing. I really just what you said. 100 percent just amplify yourself solve problems for yourself. because if 1 person has that problem more than 1 person has that problem. it's rarely, you're, you're one in 8 billion, I think it's the best way to put it.
Speaker 3:It's true. And usually when you find a solution, right, other people who are in the same shoes, we will really appreciate that. And you'll be like, Oh my gosh, that's so smart. Why didn't I think of that? And that's where I think like, you see a lot of people starting to slowly. Understand the potential of AI, especially when people like yourself are sharing your knowledge. So they see what is possible and what can be done instead of like, oh, it's just hype. Now they can see that. Oh, it can actually help my agency to get more business, close more clients and actually you can generate revenue from that, right? And I think that is a beautiful part about learning how to really leverage AI in your business. So wrapping up Rory, What do you think, are some of the key tools that anyone needs to master when it comes to image generation?
Speaker 4:I mean, there's my favorite. I'm biased though. It's it, it is Midjourney because I think it's, if you're a beginner, right? There's something running something like stable diffusion is gonna be really hard. Dolly's really easy to me. The image quality isn't there. It doesn't depends on what you need though. If you need something that's more art focused, you know, Dolly can do the trick and it's very easy to, you know, the one thing I love about chat GPT with Dolly is having the conversational availability. So being able to like talk to. ChatGPT with that, you know, make subtle adjustments to things, but Midjourney can do just about everything. So it's a great, I call it the Swiss army knife. You can get just about anything you want. If you want to get really precise, you know, you're going to go into different things like, you know, stable diffusion and, and, you know, Different. I love the tool Korea just to throw this one out there. It's so much fun to play around with. If anyone is looking for, you know, a way to kill seven hours of your day or night, you know, you put, basically put an image in there and you can adjust an existing image and just, you know, draw on it. It's really like fun, but yeah, Midjourney to me. Is obviously if, if anyone goes to my profile, you'll know that I have a subtle obsession with
Speaker 3:little subtle, obvious at all.
Speaker 4:And,
Speaker 3:and with that, like, where can our audience find you and who should reach out to you, Rory?
Speaker 4:Totally. Yeah. I'm on, I'm on LinkedIn, Rory Flynn, R O R Y F L Y N N. But I'm also on Twitter, to be quite candid, I go a little bit more experimental on Twitter than I do on LinkedIn. So if anyone's looking for a good time, over there, I also have a newsletter, called Systematic, which break down a lot of this stuff more so than just the Midjourney piece of it, go a little bit more into the operation side. So that's, you know, that's, that's where everything's at. it's, it's been crazy. To say the least, but I'm, I'm really excited about all this stuff. And I appreciate you two, Audrey, for bringing me on and also help. You know, like I look at your stuff and I look at a few other creators stuff to, to give me. The info that I need to, to actually like work on Midjourney. Cause I know I have to keep up with other things. So when I see everyone else's stuff, that's going further and further into these other tools, it's really helpful. So I would like to say thank you for that too.
Speaker 3:I appreciate that. I also put a lot of. Thought and work into creating content like yourself because we want to provide real actionable value for our audience and with that Thank you so much for your time today. Rory. It was an absolute pleasure and thank you guys for tuning in So don't forget forget to subscribe to our podcast and hit the bell button for more actionable insights We will see you in the next show. Thank you. Rory
Speaker 4:Awesome. Thanks. Audrey. Take care.