Build & deploy an AI-powered digital marketing strategy

Resources: Slides, Prompt cheat sheet, Persona template, Keyword/content plan

[00:00:00] Will: It's so great to see you all. And we are fellow community members. I did this, um, really, for no other reason that I joined the Ben's Bites community. And Ben put something out about doing these expert workshops and just Had this compulsion that I wanted to do something.

[00:00:15] Will: So, um, I'm just really happy that he's given me the floor to do this really. Um, and great to see my, uh, yeah, fellow community members. I love what I love about this community. Percy's it. Ben's not telling me to say this, but I just want to say, I love how, um, it's just a bunch of people trying to work this stuff out together and there's not a lot of hype and BS it's incredibly helpful.

[00:00:35] Will: And I just, I love that about it. So yeah, that's why, that's why I'm here. Um, so I'm not a. You know, I suppose a, a, a tech entrepreneur or, or a maker or that kind of person at all? My background is advertising and marketing. Um, I got into it weirdly, 20, 25 years. Well, yeah, 20, 20 years ago I started working at MySpace, which was like this social network.

[00:00:58] Will: You, you have to explain that these days what that even was. Um, and it was massive with musicians. I'm a musician. I just got myself a job there when they started expanding globally. I won't go into it, but Rupert Murdoch bought the site and basically it stopped innovating. It was based on NET. It got older and older as every month went by.

[00:01:18] Will: And of course, Facebook came along with a much shinier, nice interface and killed it. So me and a lot of my mates went to work in ad agencies. Uh, I worked at the big agencies thought this moves a bit slowly. We need to start our own agency lads. So we did, uh, working with big brands, but on small, innovative, social and digital projects.

[00:01:39] Will: So yeah, marketing's my thing, but I'm very much interested in the technical side of it. So AI in lots of different ways and generative AI has just taken a lot of my interest. And these days I exited out of that agency a few years ago. These days, all I do is courses, workshops, talks, podcasts, all Digital marketing and how to do [00:02:00] it better and how to use these tools.

[00:02:01] Will: So that's kind of what I'm going to show you today. It's like, you know, this session is, is about AI and how to use largely chat GPT with a few other tools. It's also about digital marketing strategy, you know, and I think maybe I would say this, but I think everybody could do with a bit of a better handle on what goes into an effective strategy.

[00:02:22] Will: Cause you know what? So many people don't know that and it's actually very, very simple. So I've got a lot to kind of squeeze into the time. As Ben said, you can ask me anything. I do this every day of every week. I've been lecturing all day today. Like I'm used to having, I'm just being bombarded with questions.

[00:02:39] Will: So chuck any questions you can in the chat. I will do my best to get to them either as they're asked or at the end. I'm going to share my screen and, uh, strap in. Here we go. So, um, yeah, strategy. What's, uh, what is in a strategy? Well, it's, uh, it's pretty simple. It's these things that go into a digital marketing strategy.

[00:03:06] Will: Objectives. What do you want to achieve? Turns out, by the way, it should say People often ask me about like, ChatGBT, copywriting. My response is often, ChatGBT is not your copywriter, but it is really good at strategy. It's your co thinker and it's really good at doing those kind of things. Cerebral cognitive tasks around what is ultimately language.

[00:03:29] Will: It's a language model. It just turns out that it's not maybe the best copywriter. So that's why I love using it for this stuff, by the way. So yeah, your strategy is objectives. What you want to achieve. It's lots of research into the industry, the competition, the audience, et cetera. And out of that, you're going to build these personas, which is a lot of fun.

[00:03:48] Will: We'll do that in a bit with a chat GPT. Then you build your overarching strategy. Then you work out what channels are we going to be in? And you're very selective about that. I would hope because you can't be everywhere. [00:04:00] Despite what people like Gary Vaynerchuk say, you can't be in 23 different places.

[00:04:04] Will: And no one's got the time to do that. Then, uh, you've got to have a content plan. What are we going to talk about? What are we going to be famous for? And you can, again, only pick a handful of things. Then how are we going to use money, uh, for paid media? And finally, how are we going to get this stuff done?

[00:04:20] Will: Who's going to do it internal or external? What training do people need? What resources, equipment? Etc. Do people need to get this stuff done? That's it. That's what's in your strategy. So if that's all you came for, then you're good. You can, you can log off now. Um, but that's unlikely. So, um, I do actually have a bit of a cheat sheet of prompts, uh, here, which let's just see if that opens.

[00:04:46] Will: Uh, if I click that, I'm going to pop, I'm going to pop this in the, you In the chat. Um, let's see. There we go. Uh, there you go. So this might be useful to you. Now, I want to say, I don't know how some of this is going to go down with you guys, but the first thing I want to say is could split the room is I don't believe in prompt engineering fully.

[00:05:11] Will: Okay. I'm not a big, I'm not a big subscriber to the idea that there's some like perfect prompt that will solve all of your problems. Okay. There's a reason why it's a chat interface and you kind of are managing it sort of like an intern or like a dog's body really to go and do things and think about things and bounce things off.

[00:05:31] Will: Right. So Yeah, I want to be really clear, like, these prompts, you need to put some stuff in here too, and you need to then further those and continue that conversation. It's not like a one shot prompt that's gonna, um, solve all of your problems. Uh, looks like link sharing has turned off in that dark. Oh, you're joking, is it?

[00:05:51] Will: I'm so sorry. Yeah, there we go.

[00:05:56] Will: I've just done that. Um, thank you. So, [00:06:00] um, yeah, but they are good starting points, um, for everyone. And, uh, I think, look, we were all busy. We could all do with saving a few seconds. So there you are. You're welcome. So, right. The first. thing that I want to do to put together my strategy is to work out some objectives.

[00:06:18] Will: Um, and honestly, in your work or in your personal life, it's a really good idea to make sure that your objectives meet this five point checklist. Now this is, this may not be news to you because this is really well established in the world of marketing that you can't just say like in real life, you can't just say I'm going to get fit by Christmas.

[00:06:39] Will: Cause you're going to fail guaranteed. So what does that mean? Are you going to be able to run around your local park in two minutes? Are you going to be able to bench press 30 KGs? I don't know, but you've got to really make that incredibly detailed as an objective so that everybody's much clearer why they're, you know, what, what they're working on.

[00:06:56] Will: Right. So you can't just say we want to blow up on Tik TOK. It's not going to work. I want to become a thought leader on LinkedIn. It's probably more along your lines, right? Not going to work. You're going to fail. What does that actually mean? What will it look like if someone, a third party came in to observe whether or not that had happened or not, would they be able to see very clearly whether or not you had become a thought leader?

[00:07:21] Will: So an easy way to make that an observable, tangible, real objective is to put it through this, make it specific, incredibly specific. Like, do you want a certain amount of. Followers, do you want to get a certain amount of clients or clicks or offers of work through DMs or whatever? I don't know. Something very specific as an outcome that's happened or not.

[00:07:42] Will: Um, how are you measuring that? Uh, is it achievable? Don't make stupid, you know, unachievable and overly ambitious goals. Is it relevant? Is it what you need? Sense check it. Right. What do you actually want in your business? And are you creating marketing goals that drive towards that? Are you just creating ones that sound good?

[00:07:59] Will: Because everyone [00:08:00] says you should be a thought leader on LinkedIn these days. Um, and also what? What's the deadline? What's the time frame of that? Now? It turns out that, um, chat GPT. pretty good at working stuff like this out. And what we're going to do here is we're going to start a chat conversation with ChatGPT that's going to get really long.

[00:08:20] Will: So, unless it's changed since I last checked, there's 128, 000 token context window in ChatGPT Which I assume a lot of you are using. Not quite sure what it is in the free version, but um, let's just say in the paid version of this or in Claude or Gemini, it's pretty bloody long, right? It's a very, very long conversation that you can have.

[00:08:42] Will: And what I've, I don't know, again, this might be so obvious to you guys, but I only sort of realized a few months ago that actually the value of this is in having these long conversations and making Chat GPT really think about things and put them on the page and then use that as, um, part of any subsequent prompt.

[00:09:04] Will: So you kind of build this kind of, uh, column of knowledge that gets better and more well informed as it goes on. Is that just really abundantly obvious to everyone or not? But it did take me a while to kind of work that out as opposed to having new chats for every task. So you'll see our strategy can build quite nicely in that way.

[00:09:24] Will: So I might say, um, um, create a smart and it knows what that means by the way. So you don't have to explain objective for my. which sells cupcakes and I want to grow on Instagram. Okay. Um, so it knows exactly what I mean. It's going to put that together. It knows I mean specific, measurable, achievable, et cetera.

[00:09:51] Will: Now I would query that if I wasn't in a rush and didn't have loads of stuff to get through, I'd have a back and forth conversation with chat GPT [00:10:00] on, you know, why, Um, et cetera, but something that you really definitely need to do is to benchmark and, uh, chat. GBT is just pulling this stuff out of its training data.

[00:10:12] Will: So you need to get it to actively research. So you need to, because what is a good engagement rate? You know, what is, um, what is a good click through rate on a link in bio? What is a good amount of followers? So you need to explicitly, verbosely tell it to search Google, um, for benchmark engagement rates in this industry.

[00:10:37] Will: Uh, and it will go off and do that. Then you can get it to revise. And again, it might sound obvious, but pushing chat GPT to go and do research and bring the fruits of that back, it will make all subsequent conversation and output better. I don't know why that is. You would think that it would be able to sort of just do that with its own training data, but it doesn't seem to be able to.

[00:11:04] Will: It seems to be really helped by you making it do the thinking and show it's working. Um, but anyway, Ben, is that really obvious? Does everybody know that? Or am I just an idiot or what?

[00:11:15] Ben: No, I think people, people want the quick version. So people want the right prompt to do the right thing. But

[00:11:21] Will: yeah, that's how I feel.

[00:11:22] Will: Yeah.

[00:11:22] Ben: If it's when you see people teaching this, how to write the perfect prompt, it's like a mini essay anyway, because you're trying to stuff. All that context in that one prompt. And that's why we teach the way we teach is that the conversation as it goes on, it's getting all that relevant data and it's getting more specific on the thing that you're trying to do.

[00:11:41] Ben: So I don't know if it's obvious either, but that's how we think about it. So yeah, we agree.

[00:11:46] Will: Oh, good. Just checking. Um, um, yeah. So, uh, that, that's, that's important. So we'd work that part of it through. Um, now we might want [00:12:00] to start getting a bit of understanding as well of what other people are doing. Now there's probably some, you know, there are some fancy tools for doing this for, for, you know, Getting data and doing like competitive analysis.

[00:12:14] Will: Um, I mean, here's one example. This is one I love. You may well know about this, but, um, if I ever want to know what, um, a certain YouTube account, TikTok account, Instagram account. is doing and how they're doing that. I do this, which is a fairly replicable, um, easy to replicate a tactic where, uh, let's say that I search Le Creuset, right?

[00:12:42] Will: I don't know why that came to mind, the cookware brand. And I'm, I'm using this tool called Toc Backup, by the way, which is linked to in the, um, in the slides. And it, what it's going to do in a minute is produce a CSV. Of all the accounts videos with all of the like counts, comment counts, a full transcript of what was in the video.

[00:13:06] Will: You know, uh, the text description, hashtags, the time of day it was posted, all that kind of stuff. Um, I, I, maybe that this is my one gremlin, is it, today? I don't know. Oh no, I think that was just a login in bit that foxed it. Um, so, I'll probably need to give this a minute, but I'm gonna download this, like, mammoth Excel spreadsheet, basically.

[00:13:27] Will: Of the, the, the, the account's entire history, right? I can look at it visually on the page, but um, it's, um, yeah, it's gonna be, um, giving me this, this great spreadsheet. And again, probably obvious to you guys, but I am definitely of the mind that with chat GBT, it's generic in, generic out and I like to really load it with data.

[00:13:50] Will: So, uh, something I always say to my. Uh, my students, my delegates that, to be honest, aren't probably as wise to this stuff as you, [00:14:00] is every time you see an export CSV or export data button anywhere, whether it's Google Analytics, these kind of competitive analysis tools, anything, just grab it, grab it, and just chuck it into chatGBT.

[00:14:13] Will: To see what it thinks because yeah, you know, it loves to chew on data. It can analyze that data in ways you've just not got time or the brain power or headspace to do. So this is another way that you might sort of start to hunt down what your likely objectives might be, right? And I think you can do this in really Quite basic ways, by the way, um, you know, taking screenshots of someone's linkedin page, um, taking screenshots of, um, their ads, other results from other tools.

[00:14:46] Will: I'm not adverse to just literally chucking screenshots into these things, uh, well, into ChatGPT because it's really good at uh, grabbing screen reading, passing that data just from a bunch of screenshots. So it doesn't have to be sophisticated. You just have to put the time into actually getting it the information, um, and feeding it in.

[00:15:05] Will: Uh, we'll come back to that anyway. Um, another tool I like to use all the time, I'm sure you know about this, but I'm, I'm guessing you might not have used it recently, is SimilarWeb. SimilarWeb is very cool because, yes, I know that if you've been around for as long as me, you know that SimilarWeb tells you the estimated traffic for any website, and it gives you a bit of information about the analytics.

[00:15:27] Will: Now it's not as reliable as actually getting access to their Google Analytics, but it's a good steer on What kind of traffic they get from what kind of demographics, countries, et cetera. But in recent years, uh, similar web has grown into showing you a company's Google ads, showing you their social media activity, even like, um, some of the messaging they send out, you know, all kinds of stuff.

[00:15:53] Will: And the good thing is, unlike a lot of these tools, There's a seven day trial that's totally free. And I [00:16:00] like, I know this sounds really skanky, but I just use a different email address at willfrancis. com and I'll just get the seven day free trial. Um, take loads of screenshots of a competitor's, uh, analytics and their ads and the stuff they're doing and feed it into chat GPT.

[00:16:17] Will: So that can all inform, um, you know, what I'm doing, um, and how I'm trying to scope out my competition, how well they're doing and ultimately arrive at how I'm doing. Well, I should be aspiring to what, what my ambition could and should be right with my objectives. Okay. So that's some of the stuff you could do now.

[00:16:39] Will: Again, I know in the interest of time, I have to sort of scoot on. So. Um, let's get onto the next, uh, bit, which, uh, well, let's say we've done our objectives. Now we need a persona. Now, uh, I think there's been some quite good tutorials or as I remember, probably in, in Ben's bite, but, um, you can definitely create.

[00:16:59] Will: Uh, imaginary focus groups in chat. GBT again, I think it's something that's quite good at. Um, so I could say, create a persona for the, um, and you would give it again, there's, there is a prompt in my cheat sheet that goes into way more detail here, but I'm just going to do it the quick way to save time, uh, for the older.

[00:17:20] Will: Uh, Moms, I'll say, um, who, who could shop here, right? Shop here just to get a starter. Again, I might prompt chat GPT, go and search the internet and find out about certain components of that persona. Now you're, you might not be marketers, I'm guessing. So you might not be so bought into the reason why we even use personas.

[00:17:43] Will: It might seem a bit academic. I'm not an academic. I'm a practical marketer. I like to get things done, like to get results. And honestly, personas are so useful because they are a generalized, fictionalized representation of your target audience. And when you create them and you make your [00:18:00] audience real, you start to realize that there's a lot more you could and should know about them.

[00:18:05] Will: And it kind of leads you towards working out things like their pain points, which is the most powerful thing you can know about people when you're marketing to them. But also their goals, their ambitions, their fears, what keeps them up at night, what they hate and love and all that kind of stuff, right?

[00:18:20] Will: And it also, hopefully, helps you avoid dog food in your audience, i. e. feeding them crappy emails and LinkedIn posts that actually you yourself would never consume. So I think there's a really good case to be made for making your audience real. And it doesn't have to be seven or eight different personas, all the like granular slices of your audience, just a handful, two or three stereotyped personas that kind of personify your most valuable potential customers.

[00:18:51] Will: And just forget everyone else, and you're gonna superserve those people. But you need to get inside their heads first. So, you start out, perhaps by asking ChatGPT to, uh, outline that person, what they're into. ChatGPT actually understands humanity in quite a scarily good way, I think. But we're gonna go a bit deeper.

[00:19:10] Will: And we're gonna do something weird here. Don't worry, there's no ethical You know, fallout from this is cool. You can do what you want. I mean, I'm not going to do anything too sordid, but I am going to inject poor old Sarah with a truth serum because I think that's quite fun. So it was it Sarah. Yeah. I'm going to say inject Sarah with a, should we make a potent truth serum just to be like, you know, extra, you know, Um, because it's always good to be extra, I think, and extra kind of explicit and verbose and pushy with um, ChatGPT, a potent truth serum.

[00:19:40] Will: Um, so she reveals her innermost emotions about, um, yeah, our business and her needs. Right, so she's like, the, the truth serum's flowing through her veins. And she's going to start [00:20:00] saying all this kind of deeper stuff. Now, this might seem silly, but we're just trying to get beyond the generic in the same way.

[00:20:06] Will: I might bring some other research in here, some stuff I'd found and that's all useful. I'm trying to get to the, I'm trying to get deeper. I'm trying to get beyond just like, yeah, Sarah's a mom. She lives in a town, you know, it's like you're not going to hang a winning marketing campaign off that stuff.

[00:20:23] Will: But if you can find out what bugs her, what she hates, what she objects to, you know, it might be, um, yeah, really kind of mad thing. So I might say like, um, uh, does she, uh, does she worry about hygiene of. cakes or something. Um, you know, does she worry that we're like, I don't know. Um, and we're going to go deep into like neuroses and like quite kind of deep psychological stuff.

[00:20:52] Will: And that's all fun to do. But again, remember that that's going to exist in the conversation. And some of that stuff might come out later it when, when we start prompting for kind of ideas and Okay, cool. So we've got our personas again, we might have, um, two or three of those. Um, I can, uh, give you a nice template for a persona, which would be like a one pager.

[00:21:18] Will: Um, I will pop that over to Ben and he can put it with the recording, um, if that's all right. Okay. Next up in your strategy, you need a channel plan. You need a platform. Plan or channel plan, whatever you want to call it, it's the choice of where you're going to put your efforts. I mean, this is pretty simple, right?

[00:21:36] Will: If we're targeting, um, Sarah, a mom, and we've quizzed her at length about her life. One thing that's good to think about when you choose platforms. Don't be too simplistic about it. Don't go. Oh, yeah, she's 18 tick tock. He's a businessman in a suit linked in. It's not as simple as that. So what we're gonna do with Sarah is we're gonna say map out Sarah's [00:22:00] typical week in terms of media consumption.

[00:22:06] Will: This is something we always used to do. In agencies that I worked at, we, we'd put this on the whiteboard. We'd map out that persona's typical week. Does he listen to a podcast on the commute? Does he look at LinkedIn on his lunch break? Does he watch longer YouTube videos on the afternoon on Sundays? Do you know, does he, I don't know, check Instagram during the ad breaks 8pm.

[00:22:28] Will: And when you start to do that, you start to realize that everyone's a much more complex consumer of media and marketing than these simplistic things that people Stupid things that people say in meetings, like, Oh, she's 18 tick tock. You know, it's always more complicated than that. Um, so again, we might quiz and question that, but we can see there's quite a lot of information here and it is quite complex.

[00:22:52] Will: Um, but now we're going to say, you know, based on this, um, give me a channel strategy that defines we should reach. Through. Sorry for the typos. Chat GPT doesn't care. Um, so it's going to come up with that. Okay. And again, it's, it's all kind of growing in value, um, over time. Right. Okay, good. So we're getting somewhere.

[00:23:25] Will: This is quite a quick race through building a strategy, but I like it. Um, right. So we know where we're going to be. We know who we're talking to there. We know what we want to achieve there. Because we could actually loop back here at this point and make smart objectives for each of our platform. But keep it small, keep it tight, keep it focused.

[00:23:43] Will: You don't need to be trying to service, you know, 10 different objectives over the next quarter or year. Um, so keep, yeah, keep it tight. Be in fewer places, a bit of advice, you know, don't try and be in loads of places at once, like I said. I'm sure that's quite obvious. And that being [00:24:00] said, now what we need to work out is what do we talk about?

[00:24:03] Will: You know, what are our content pillars that we're going to become known for with Sarah and maybe the other personas that we have? Content pillars are really important. They often get a bit overlooked, but ultimately, You know, you're, uh, you want to become known for X, Y, and Z. You don't just want to be just another cake shop.

[00:24:27] Will: You want to be that cake shop with the Instagram where they do the drippy chocolate reels or something, you know, and the more you can become famous for specific formats for specific types of content, whether or not you are truly famous, objectively, the point is you're more memorable to the people that you address, however big that group is.

[00:24:48] Will: So that's why, by the way, when you see brands that are doing well in social media, for instance, you might have noticed that they often have a thing that they just repeatedly do again and again. And it's, it's because for some reason, it's easier for us to, as humans, to have a relationship with a, with kind of quite one dimensional entities.

[00:25:09] Will: And it's more easy for us to remember and recall those entities, whether they're creators or brands. So kind of pick your thing is what I'm saying. Whether it's a format or a theme. Um, and having done this research into what you want to achieve and to whom you'll, you'll be much clearer because otherwise you just like aimlessly brainstorming themes, or you're just looking at the competition and copying them.

[00:25:32] Will: So we do need to think about that, but also whether it's content marketing that we're doing to try and appear in search, or we're trying to appear in YouTube search results in social. We know that the algorithms that select content in those places consistently reward websites, if it's SEO, Google, or creators in social and video, we know those algorithms reward the producers of content that have [00:26:00] clear track record in specific topics over time, rather than just create like random scattered all over the place bits of content that don't really have any kind of running theme, okay?

[00:26:13] Will: So humans prefer content pillars, uh, algorithms prefer content pillars. Yeah. So we need to come up with them. That's my case I'm making. So, um, now I would be cautious when you get into the creative stuff, mega cautious about leaving this to the intern, which chat GPT is. So we're going to say, um, we're going to just, it's going to be a draft.

[00:26:35] Will: And there is more in the prompt sheet about how to do this a bit more, uh, you know, um, thoughtfully, uh, create some content pillars, um, to reach. I would say, um, not reach. We don't want to reach her. We want to knock her off her chair, um, to create some content pillars to entertain and inform Sarah in social.

[00:27:02] Will: And I don't know, uh, search and video

[00:27:11] Will: fine. So it might create some content pillars, but you might get stuck here. You might be like, Oh, what do we actually want to talk about? We need to do a bit of brainstorming, bit of research into that. Yeah. Um, and yes, chat GPT can't do keyword research despite what some people say. And it's not great at coming up with this kind of stuff.

[00:27:30] Will: So there's a couple of tools you could, um, call on here. Um, I don't know if anybody has anybody used this tool. It's called the globe Explorer. It's so interesting. And there is, um, it's, it's basically, I think just tapping into GPT behind it, but I'll show you how it works. So let's say I'm, um. Yeah, running a cake shop just to keep things simple.

[00:27:56] Will: Um, I'm just going to put cakes, [00:28:00] cakes into it. What it does is it creates this map, this very detailed map of any topic you throw at it. So it creates this kind of like wiki style breakdown down the left hand side. And then it all, it also creates this like, Uh, detailed map of all the subtopics within that, and that can that I find that really useful personally when I'm feeling out the map and the landscape of any particular theme or topic.

[00:28:29] Will: And then if there's something particular in here, you can click that and it will create the map of that. I'm actually looking at this. This looks like this might be a little bit, um, too zoomed in. So I might like step back a bit and go, well, what about baking, right? Let's, you know, make it more generic.

[00:28:44] Will: Like, you know, what subtopics does kind of cakes sit alongside? And maybe baked products, something like that. So it takes a bit of, uh, feeling out, but personally, I love it as a, as a way to navigate anything, any topic. It's great. So, uh, what you want to home in with your content pillars is what do I really care about?

[00:29:09] Will: What do I love or what am I an expert in? And do those things have an overlap with what my audience wants to see? Wants to know, okay? Quite simply. So that's, that's your content pillars or your content clusters. Then what you can do is you can then do some keyword research. Um, there are lots of expensive keyword research tools out there.

[00:29:32] Will: One of the best ones you can use is honestly, uh, Google. Uh, and I'll show you simply how you do that. Let's say, uh, I mean, I'm kind of making this up, but we'll go with it family celebration cakes. Okay, let's go celebration cakes, right? Quite a broad topic. Maybe that's just one of our pillars. We put that into Google [00:30:00] and, um, we go down to People also ask, which is an absolute goldmine.

[00:30:08] Will: If you're coming up with ideas and ways to be useful and valuable to an audience, remember that, and please remember this anytime you're about to publish any content. Anywhere, email, social, website, video, whatever. Unless content has value, people aren't drawn towards it, it's a waste of time. And value only exists in two flavors, uh, entertainment value, informational value, usefulness, let's say.

[00:30:39] Will: So unless your content is entertaining or useful, it's dead on arrival. And I would suspect that a lot of you, like me, uh, probably lean more towards expertise led content, content that's useful. Rather than trying to be, you know, massively entertaining. So, um, the best way to be useful and, and the best way to give useful knowledge is work out what do people want to know.

[00:31:05] Will: And whilst things like SEMrush and Ahrefs and all the expensive SEO tools can do that keyword research. What I find personally is that people also ask, it's a bit more creative. It's a bit more liberal. It's not so literal, you know, it doesn't just return all keywords with, Celebration cake in it thinks about it a bit more and also it's from the horse's mouth.

[00:31:26] Will: So like on some level, maybe I trust this information a bit more. Now you can see what I'm doing here is I'm opening and closing any of the questions as it happens, the top question many times. And when you do that and people also ask, the list grows and grows and grows. So I've now got this like really long list of questions.

[00:31:45] Will: Right. So obviously you can, you can guess what I'm going to do now. I'm going to swipe over that, copy all those questions, and I could do loads of things with that. But again, I'm avoiding a generic prompt. Okay. I'm not just going to chat [00:32:00] GPT and saying, you know, create me content about celebration cakes.

[00:32:05] Will: Cause it's not going to be able to do that in a qualified way. What you're bringing to the table with these questions is this is what people want to know. Now, create me a series of. Instagram posts, Instagram reels, blog articles, whatever, that over the course of let's say a month answers these questions.

[00:32:27] Will: Create a content plan, uh, plan for Insta, uh, that over the course of a month answers these questions about celebration cakes. So maybe we're going to make it celebration cake month, why not? Uh, paste those in. And that's really giving it something to chew on. Now, if you do use a keyword research tool, you might use something like answer the public, something like that, right?

[00:33:02] Will: Which tells you all the questions people are asking Google about, about anything. So, um, you know, I could get fairly similar stuff, but definitely a more kind of literal read on it. When you use a tool like this, again, you're looking for that Export button, download that CSV. There might be 500 questions in here.

[00:33:22] Will: Give it to chat GPT to chew on, ask it stuff. Say what are the recurring themes, um, build a content plan around this. Et cetera, et cetera, right? Um, okay. So let's check in on that. We've got our, um, content plan and to be honest, it looks like pretty good content. I don't quite know why I've gone this down this cake shaped rabbit hole, but I have, and that's where we are.

[00:33:49] Will: Um, so yeah, just remember it's the intern, it's a draft, it's, it's not going to be, um, you know, perfect. If you're [00:34:00] Thinking about more maybe, um, putting this in some sort of plan, um, I can share something with you, you might find quite useful, particularly if you're thinking more along the lines of like, Content on a website or on LinkedIn, more authority driven content.

[00:34:18] Will: Uh, I'm going to share this, uh, link with you because you might find it useful, maybe. Um, let's have a look. Um, so, what is this spreadsheet I'm sharing with you, this Google sheet? So, this is my personal spreadsheet. It's a bit in its early stages, so go with it. But, um, be kind. But it's, uh, this is my personal content plan that I'm just starting to build out.

[00:34:45] Will: So I know my content pillars, yeah? I've got my content pillars. These are the things I want to be known in the world for as an expert on. Okay, debatable, I know, but, um, these are the things I consider myself an expert. Okay. Um, so what I've done then is the kind of process we've talked about. I've gone and mapped out those topics.

[00:35:05] Will: I've gone and, uh, seen what's in keyword research tools, basic ones, like people also ask and, uh, you know, uh, answer the public, stuff like that. And I've come up with my long list of, uh, keywords here. Um, I mean, I don't think we've perhaps got time to talk about keyword research as a thing, but I'm sure you've got a sense of perhaps of what that is just working out what things you want to rank for in Google.

[00:35:32] Will: Um, but then what I've plugged in and something I've found myself calling on in lots of different contexts in my sheets over time is, um, something I'm sure a few of you must use. It's called GPT for work, GPT for work. com. Does anyone else use that? Um, and it's basically an extension for sheets and docs.

[00:35:52] Will: In fact, it comes up in the list here is GPT for sheets and docs. Okay. Um, and you don't have to open the sidebar [00:36:00] for it to work. It's just like there in the background, ready to be, um, activated if you like. So I'll just close the side panel so you can see that I can create a formula. All I have to do is equals GPT, a prompt, and then.

[00:36:17] Will: It spits out whatever you want in that cell. And the beauty about that is it's quite easy to scale. So what I've done in this prompt here is it references my keyword and it comes up with an SEO optimized headline. But now I've done it once, I could literally, like, you know, like, extend that down. Loads of cells.

[00:36:38] Will: It will read the keyword over in the first column. It comes up with it. And now it's spitting out those SEO headlines. And you could take this to, it's like logical, maybe bonkers and silly conclusion. I don't quite know you, you'd be the judge of this, but, um, I've then got this other column here where there's a whole outline for that article.

[00:36:58] Will: Right. So, um, there we go. I'll extend that down a few lines. It's going to write me some outlines. Um, for the article, um, then based on that, so I've got a column that's, um, the meta title and meta description. So how it will appear in search results. Then it's going to come up with some LinkedIn ideas based on that article.

[00:37:19] Will: Uh, so ways I can reuse some of that research and thought as a LinkedIn. Carousel or post, um, then some ideas for a lead magnet that I can include in the article. So something that people have to give their email for, but that's high value, like a template or a cheat sheet or something like that. Again, it's draft work.

[00:37:39] Will: It's just the intern having a bit of a think about what I could do. Uh, but you know, it's, it's pretty good. I find it actually, the results here are pretty good and they get me started. I'm. I'm someone who I really suffer from blank page syndrome, and that's honestly one of the greatest gifts ChatGPT etc.

[00:37:56] Will: has given me, is that sort of like, starting draft. [00:38:00] And I'm sure you're, uh, to some extent the same, right? Okay, so, um, we're turning our plan, uh, into content there. A few more things to say before we Q& A is, uh, you've No doubt come across this idea of, uh, when you have content being created, it's really important that it's in the right tone of voice.

[00:38:22] Will: And I'm sure you've seen people doing that and training chat GPT or custom GPTs, you know, but do you know what? I've actually found that the projects. Functionality in Claude is quite well geared towards this, right? Because you can basically feed it loads of documents, loads of your existing content, and then it can actually create a thorough, uh, tone of voice guideline that you can then use in any future content creation tasks that you feed to a chat GPT or Claude.

[00:38:53] Will: But honestly, however you do it, the simple idea is you feed it loads of your existing content. You get, uh, you tell it to produce a tone of voice guide. You know, you tell it to describe the tone of voice of that content, and then you use that in future prompts around content creation is nothing more complicated than that essentially.

[00:39:13] Will: Um, Claude course with projects from Ben in the chat there. Yes. Um, I found projects to be useful for very specific things. Um, although I definitely do call on chat GPT for most, uh, everyday stuff. Um, and then look, the last thing that you, uh, just to sort of jump to the end of this. Before we go to, uh, chat about it a bit.

[00:39:36] Will: Okay, you've, you've, you've done all of this strategy work. You've got to the point where you're producing content. Um, there's stuff in the cheat, you know, in, in my prompt sheet about, um, coming up with ad copy. Um, and you might even use it for landing page copy and stuff like that. But then you no doubt want to pull it into some sort of report perhaps.

[00:39:59] Will: And [00:40:00] something that I've found really useful for doing that is a tool called gamma. Does anybody use gamma?

[00:40:09] Ben: Um, tutorial, um, that went quite popular and I'm an investor in gamma. So I'm glad you're

[00:40:15] Will: joking. Oh, mate, that's so good. I it's really good. I've used. I mean, look, you know, I, some of my, I would say 5 percent of my time is created, creating decks, but it feels like 55 percent because I hate it so much.

[00:40:33] Will: And I've used some of these tools, and they are really bad, but what are they doing at Gamma? Because it does work. It delivers on the promise. I'm not just saying that because you say you're an investor. Like, it genuinely delivers on the promise of you chuck us a load of information and we'll make a deck out of it.

[00:40:49] Will: So. Yeah, great. I don't know what, what's behind that, but, um, it's, it's really, really good. So, you know, for instance, um, here, what I did, uh, was I created a new deck and I gave it a load of, again, exported data from a social media listing tool and some other keyword research tools, something like that. And it, and it created this, you know, like, uh, actually, I mean, all right, it's a funky image there, but because it actually creates the images with a I, their generative images.

[00:41:24] Will: Um, and it's very good. And there's something about the way that they've, um, chosen the layouts for the slides. I think that works really well. Someone who really understands PowerPoint has done that. I think that's probably what's in there. Whereas The people that like Canva and Octodeck, whatever it's called, I'm not sure they're quite so on top of it.

[00:41:47] Will: Um, so, uh, yes. I I found Gamma to be incredibly good, um, it's very it's very it makes it very easy for you because Because it uses Gen AI to actually [00:42:00] generate the brief for the deck. So you can just feed it written information, but this is, this is a useful for you. But what I found works best is I create a Google doc, blank Google doc.

[00:42:13] Will: I paste all of my kind of findings and things in there. Um, I sometimes run it through chat. GBT first. So I might put a spreadsheet, a massive spreadsheet or CSV file into chat. GBT have Jack GBT. Write a detailed report, paste that into my Google Doc, and then I build up a bit of a brief. Takes me like five minutes in the Google Doc.

[00:42:35] Will: And then I just put, put that into Gamma. And you can connect Gamma to a Google Doc or you can just paste the text in. Um, and I find that works really well for, um, creating decks. I think I've run up against time a bit there, Ben. Sorry if I've gone over a little

[00:42:51] Ben: bit. The problem is, I think, we want to continue hearing you talk and going through stuff.

[00:42:56] Ben: Um, so I don't know if, if people want to get some questions ready, um, then we can get to that. But if, if there's one or two things that you think you haven't touched on and you're a bit gutted that you haven't yet, like, we're happy to, to listen to those, I think. I speak on behalf of everyone here that, yeah, this, um, Um, and yeah, we're happy to listen a bit more if, if you've got more to share.

[00:43:24] Will: I do. Um, I think I do say that. So I do. Uh, let me have a look. Um, so yes, um, Let me, let me switch tabs. I'll do, I'll do another five minutes and then come to the questions. So, um, I suppose all I was going to say, I mean, this could take a bit of time if I actually went into the doing of it. I'm just reiterating here that, you know, uh, ChatGBT does live charts.

[00:43:54] Will: Um, and so, you know, it can present reporting kind of visuals based on things like an [00:44:00] export from meta insights or Google analytics or something like that. Um, but yeah, like I say, I think the magic of it comes in. So take a step back. What no one wants from a marketing person is a load of data in my experience.

[00:44:18] Will: But no one wants you to turn up. I don't, I know, I don't even know why I'm saying this cause you're not marketers, but, you know, No one wants a marketer to turn up and just say, yeah, we have 3000 website visits and we've got 30 more followers on LinkedIn. And this post got 3000 views. It's like, no one wants that.

[00:44:33] Will: People want to know why you're like a detective, right? We know you've got loads of evidence. We don't want that. We just want you to tell us. In forensic detective terms, we want you to tell us what happened on that night? Why was she murdered? Why, why was he there with a gun? And where did he run afterwards?

[00:44:49] Will: And why was the car rolled over? And why is there blood there? You know, it's the why. They want the, they want your interpretation of the evidence, not the evidence itself. It took me a while, you know, in my early career, to just kind of work that out. I thought I was a genius because I was the only one with a login to Google Analytics 20 years ago.

[00:45:07] Will: And then I realized, all right, no, people actually want a bit of a theory on why this has happened and what we can do about it. What we can do more of, what we can do less of, what we can stop doing, what we can start doing. And I think that's where a language model, again, is actually strangely good at that, in that you can feed it this data.

[00:45:25] Will: And ask it to build hypotheses with you in collaboration, because you're there to challenge it again, like the intern, you know, you're like, Are you sure about that? What about this? What about that? And to build those hypotheses, but starts with chucking CSVs, usually, and big data sets, um, at ChatGPT, um, you know, just like we saw with my, um, what's hopefully ready, Uh, my TalkBackup example, right?

[00:45:55] Will: You know, I can download the data on a social account. This could be your own social [00:46:00] account, by the way. It doesn't have to be a competitor's. It could be, uh, technically anyone's social account. And just let me show you how detailed this specific data is. TalkBackup, you can use it for free, but it's about nine dollars.

[00:46:12] Will: a month. So, you know, if you do work in content or social, it's worth it. So you see, it's even got the transcripts of videos. It's got all the data around videos. Now you can, uh, chuck that into chat GPT and just ask it anything like, um, you know, uh, what can I steal from this? Um, you know, Uh, so, uh, what is making this?

[00:46:38] Will: Oh, you can ask really stupid questions like, What is making this content work? Uh, what can I steal? And again, I personally find very brute, uh, Brutally blunt, direct, I think there was a study, wasn't there? I might have seen it in Ben's Bites, I don't know, but In one of the newsletters I follow, I recently saw There was a study that, being rude, Uh, It gets better results on average, but I do personally find that being really blunt and like, you know, using that kind of language.

[00:47:09] Will: Um, so yes, this analysis is, I think, what it's best at, but I think I've probably hammered that home, um, enough. Right. So I answered some questions. I think that would be good. Uh, I'd love to see what you're asking in the chat. Feel free to drop me absolutely. Any questions you want in the time we've got left and I don't mind staying a bit late if they take time.

[00:47:30] Will: I, um, yeah. Um, okay. Yes. Ah, right. Okay. So Alborz, you mentioned about perplexity, uh, automatically disappearing. I think, yeah, you're, I think you're referring to the bit about conversations, um, and maybe web citations or something. Not quite sure. Um, [00:48:00] but yeah.

[00:48:04] Ben: Yeah, I think the questions from Megan had the first one. Um, so a question. Is what's the more of your favorite AI tools to assist with marketing or running a business?

[00:48:19] Will: Well, I am, um, I'm, I'm unusually dependent on chat GPT. I got to say that one thing I really love is obviously the headphones button on mobile. So, you know, if I'm thinking through things and I just want someone to unload onto, if you obviously I'm sure you've tried that, but persevere with it. I think that's great.

[00:48:42] Will: Um, that that person does it for me, uh, look in, in the marketing world, people aren't taking generative video and image AI as seriously as it might seem outside of marketing, right? It's still not the quality we can rely on. So in terms of generative stuff, we are kind of, uh, largely just using the language models, really.

[00:49:05] Will: Um, and. I think to me, it's more about the combination of traditional analytics tools with chat GPT. I know that's a really boring answer. Um, because I often find when people build AI tools that try and do other stuff, they can often fall flat. And I think it's unavoidable. There's some work, you know, required of you to gather that information, um, from the sources, from similar web, from Google analytics, from your social insights.

[00:49:40] Will: and give it to chat GPT to analyze. Sorry, that's a bad answer, but you know. Well, I mean,

[00:49:44] Ben: that's the tools you use. That's the tools you use. Um, there's a question from Sean, which is, is there a piece of content or strategy you've executed that you are particularly proud of? And I think the second part of his question is something I was thinking too, which is [00:50:00] you'd have to see something that's nuanced since so much AI generated content is general and he says bombastic at the same time.

[00:50:07] Ben: So I think, yeah, it's like, this is the intern that you keep saying and How does that then get into something you're proud of, or something that is really good?

[00:50:19] Will: It will, that's the thing, I mean, look, not maybe what you want to wear, and this is only my opinion, but I don't think a lot of this stuff is gonna see the light of day.

[00:50:30] Will: This is all about your, this is your scratch pad, but it's a, you know, whereas three years ago we literally had an actual pad. This is like a, you know, a living, breathing scratch pad. I absolutely agree that. You know, the final product is going to not be something you copy and paste from here. It's going to need a lot of you.

[00:50:52] Will: And you know, I was recently like this week, I've been working on some, uh, LinkedIn ads for a client and I actively did not use chat GBT. I kind of walked away and I, and I just used, I just did the ads with a designer in Figma and a pad and a pen. And, um, you know, that's, that's just how I work. I don't. Um, it's great for research and thinking and strategy and the behind the scenes stuff.

[00:51:20] Will: I don't personally believe in it as a great copywriter, but maybe I'm, I don't know, maybe I'm old school. That's what I always worry about or think, I think.

[00:51:28] Ben: I think, I think this comes up so often because it probably is the truth. Like it is. I think so. This is the kind of baseline of where this content is because it can't be too creative.

[00:51:39] Ben: Because it's based on certain things that it has learned. So it may, it may be the limitation of the tool, but the fact of. The user interface of it and how it responds to anything makes everyone feel like it's going to come up with something that I can copy, paste, use, and it's going to be better than what I've got.

[00:51:58] Ben: But I mean, it's like you said, again, [00:52:00] with the blank page problem, it gets everyone off that very, very quickly.

[00:52:05] Will: I mean, the, the, the, the, the way that I put it to people who haven't really used this stuff a lot, who often are, you know, people saw a chunk of the people on my courses is when you're about to use it, just check.

[00:52:17] Will: Am I trying to save time and phone it in because I've heard there's like a one button solution to what would have been 10 minutes work? Or, am I trying to supercharge my work and make myself a superhuman top performer because I've got this extra tool? And if it's the first one, no. You know, if you're just using ChatGPT to phone it in because you can't be bothered using your brain, no.

[00:52:43] Will: If you're using it. To create all this sort of new research and thinking and synthesize. Um, information from multiple sources to help you think and stimulate your thoughts and get over that blank page, but also get deeper insights. Yeah, that's, that's how you use a tool. That's the definition of any tool.

[00:53:08] Will: Um, so that, I suppose.

[00:53:11] Ben: Yeah. Um, Xenia, I think I'm saying your name correctly, hopefully, has asked, How do you decide on what to ask? She says she's a marketer with, but lots of your questions weren't the questions that that would occur to her.

[00:53:27] Will: Oh, interesting. Um, I'd love to know more about that. Would you unmute yourself and ask me? I can see you there with your Yeah, if you raise your hand, I can unmute

[00:53:36] Ben: you, Xenia, if you're there.

[00:53:48] Ben: I've just asked Oh, hold on, sorry.

[00:53:50] Guest: I'm This is, like, my brain is on fire. Cool. Um, [00:54:00] yeah, there's just certain things, you know, I'm so used to asking the same questions. Again, it's from Tom's engineering. And how do you, I'm like, copy pasting everything I'm doing. So I'm like, wow, I didn't think about asking this way.

[00:54:15] Guest: So how do you come up with that?

[00:54:18] Will: That's a really interesting question that that has actually blown my mind actually in a way that That is that, yeah, that I've not thought about that. I suppose maybe as a product of using it a lot or having to. think objectively about how other people use it and guide them through it.

[00:54:35] Will: I don't know, but somehow I personally have arrived at a point where I now just talk to it like there's a person on the other end of a WhatsApp conversation. Um, and I'm just asking the questions I have. I'm not trying to format them in a special way anymore. And I was doing, I was thinking, how do I engineer this prompt?

[00:54:55] Will: Cause apparently prompt engineering is like the next big thing. That's very

[00:54:58] Guest: interesting. Yeah. Cause I've worked with AI engineers. Part of my job, you know, when I worked at accounting and all of this, so I was trying to do that. Then I found that when I'm just talking, it's different. I mean, you just kind of like ripped it off.

[00:55:09] Guest: You're like, Hey, just talk. I was like,

[00:55:11] Will: wow. It's only my opinion. It's only my opinion, and it's maybe a marketing thing. I don't know. But I think that this technology would be a bit of a failure if it required people to format their prompts in a specific way. I mean, what kind of language model are you?

[00:55:27] Will: Seriously, if you can't just take simple instruction, it's rubbish. And anyway, You know, let's face it, wasn't it John Oliver, I think, said quite wittily on his show that if there's one job that AI is coming for, it's the ask AI questions guy. You know what I mean? It's the prompt engineer. The idea that that is going to be a useful job, it, to me, is just, is just silly, I think, because

[00:55:55] Guest: So you, did you see the As you continue with what you said earlier, I guess, [00:56:00] the longer you converse with it, the better it does.

[00:56:04] Guest: Yeah. And having those long conversations. Yeah. Okay.

[00:56:09] Will: Right. It's all about that for me because there's something tangibly different that happens when you force it to put its thoughts on the page. Um, because those thoughts pervade in its future responses, where if it's constantly just referencing its training data.

[00:56:27] Will: It's completely different, the quality of response. Um, but again, that's the, I mean, that's the reason why they called it chat GPT, right? It's not send your prompt and get the response and that's the end of it. GPT is it?

[00:56:40] Guest: I just said something in the comments. I was, you can ask what other questions you should be asking.

[00:56:45] Guest: I love that. So you can go and say, okay, what's the next question I should ask?

[00:56:48] Will: Well, that's it.

[00:56:50] Guest: Absolutely. Great.

[00:56:51] Will: And that's, funnily enough, when I show people the thing about, take the talkbackup store for your meta insights and put them into chatGBT, often what I say to people is that, don't try and, don't try and be clever, just say, what could I learn from this data?

[00:57:06] Will: What is the most game changing thing I could learn from this data? So, be, it's okay to be an idiot. It's okay to sound stupid. Um, let ChatGBT do the thinking because that's what it's really good at, I think, is thinking. Um, it's not some like copy spitting machine. Uh, it shouldn't be treated like one and you have to manage it.

[00:57:30] Will: I think nowadays I feel like a manager, not a user of chat GPT. That's weird, isn't it? Jesus. That that's proper, like the future, isn't it? I've just realized as I've said it, but I do feel like a manager and not a user. And I think that's a good mindset to approach it with.

[00:57:48] Ben: Yeah, I think there's, there's some very different parts of the prompt engineering debates.

[00:57:56] Ben: There will be a prompt engineer is necessary when there is [00:58:00] something that is built So if you built an AI tool that needs to do something over and over again and get something out of it that way Yeah And that is there's a prompt that needs to happen there so that the responses the outputs are similar in length in Whatever they look like so if you've got an AI note taker There's a prompt behind that because it needs to be there.

[00:58:21] Ben: If you're using Oh yeah,

[00:58:21] Will: true. That's true.

[00:58:23] Ben: If you're using a tool Yeah. as your sort of semi intern who's really super smart and has loads of things, there's a conversation to be had. There's different ways that the prompting works. And I think the basic prompt, like, it all gets mixed up in the same conversation, which is difficult to then say what's prompt engineer and what's just writing a question or a prompt.

[00:58:46] Ben: And I think when having that, this is what a prompt template looks like to have with ChatGBT is different because that's just a way to say, this is just the context I'm quickly stuffing into you. So then when you're looking at your training data, you'll have that context of, Hey, you're an expert marketer.

[00:59:03] Ben: You have these examples that you're going off. We're targeting this, our business is this, my family life is like this. You're just quickly stuffing some of that in there. But the prompt engineering debate, it has like, has multiple sides, I think. So I think that sometimes is where the issue is.

[00:59:18] Will: That is true.

[00:59:19] Will: I'm stuck in my own, uh, kind of bubble a bit on that one. So I'm not really tuned into that other side of it, I suppose. But you're right, absolutely. If there's a specific function, Uh, it needs to perform, then yeah, clearly there's a prompt template there. Uh, because that's all Fireflies are doing, or whatever, isn't it?

[00:59:36] Will: You know, they're just pre loaded to, yeah. Um, oh, thanks Cam, quoted! Wow, it feels like so elevated when you put it like that, like a proper, proper quote there, it feels so big. Um, a banger of a quote, thank you, you're too kind. Um, it's very funny.

[00:59:55] Ben: There's one, one final question we'll go to, which was from Rob.

[00:59:59] Ben: [01:00:00] Um, so you're, you obviously talk about how you like ChatterBT a lot. And then you mentioned Claude briefly. So he's asked, are there any other unique strengths or utilities you found with other models? compared to chatGBT? And like, do you use both? Do you pay for both? Would you switch in different tasks?

[01:00:17] Ben: Like, how would that work for you?

[01:00:19] Will: Um, again, I sound like, so like lazy, maybe and un inquisitive. I use both. I pay for Claude. I pay for chatGBT. I don't use anything else. I often put the same prompt into chatGBT and Claude just to see. I'll often just literally copy and paste my prompts that I've just put into chatGBT, put it into Claude.

[01:00:39] Will: And I, I can't say that one often out, outperforms the other. What, what annoys me about Claude is it won't read the internet. That's, that's a bit of a showstopper for me because as a marketing person, I'm constantly needing, you know, Google this keyword. Take the headers from the top five results and let me copy the best of what they do to optimize for that search term or go and research like I did with benchmarks or, I mean, I can't live without that.

[01:01:11] Will: So, but yeah, with more kind of a strategic or language stuff, then I do compare. I don't use a lot of the bells and whistles like custom GPTs and things like that. I just use chat GPT long conversations. I love the fact you can throw all kinds of documents at it. And yeah, I love the fact, as someone who hates trying to write something out on a blank page, I love the fact you can talk to it in a quite a natural, reasonably natural way on your phone to get something started, like I'll just waffle about a talk I'm gonna give and then it'll create an outline for me or something, you know?

[01:01:46] Will: So, I think ChatGBT have done a really good job of winning through their feature set. Um, but yes, Claude, as you all, I'm sure I've heard, or I've found can be better at some of the creative [01:02:00] writing tasks. It's less cheesy. It's less AI ish when it's writing, um, you know, creatively for sure.

[01:02:07] Ben: Yeah, I feel the same way.

[01:02:08] Ben: Um, well, Will, this has been really, really amazing. Um, I want to thank you for all of this and yeah, we'll definitely make sure that we get. All of the links that you mentioned, the templates, the, um, the keyword thing and the slides and stuff, we'll get to all of that in the recording. Um, yeah, this is awesome.

[01:02:29] Ben: And, uh, we'll get the, I'll get the recording. Edited as soon as I can.

[01:02:33] Will: Thank you. And all I'd say to people is, you know, this is my job. I'm trying to help people through this stuff. So if you've got any questions, it doesn't cost you anything, not selling anything, drop me a LinkedIn message. I'd be really happy to hear from you or go to my website.

[01:02:48] Will: willfrancis. com drop me an email. I'm just happy to help. Uh, if you want to bounce something off me, ask me some really annoying question about your marketing that's, you know, aggravating you go for it. I'm here.

[01:02:59] Ben: Awesome.

[01:03:01] Will: Thanks very well. And thanks everyone. Cheers. Good to meet you guys. Catch you all soon.

[01:03:05] Will: Bye. See ya.

‍