053: Using AI to Generate Copy that Converts

this audience, those might be mistakes. They may want the kind of thing that hasn't worked for us in the past. Or, they may, uh, not want to hear the things that have worked for us. So, um… What we want to do is get outside of our own minds, our own brains.

Every one of us that's working on copy in our teams is… going to suffer from… All… some of these, or all of these sorts of things. And then there's our ICP. So we want to keep them in mind. We spent all this time doing the research to understand who our target audience is, who those ICPs and personas are.

And, you know, when you look at this, this is one of ours, this is Blair, he's an e-commerce business owner, and… His questions are, can I trust you? What are my competitors doing? If I hire you, how much effort and work will you need from me? And as I'm writing copy for my business, I have to keep all of these things in mind, and make sure that I'm addressing the The primary objections, depending on what sort of a page I'm producing.

And add to this that we have for personas that we deal with. humans really can't… properly right for these different kinds of visitors. So we're already behind the eight ball. When we're testing, there's another important thing, too.

Um, we can go in and make small changes to the copy, but if we do that and try to do an A-B test, typically the effect is not big enough for us to detect in an A-B test. You can go change a headline, those are usually very powerful. Um, adding captions to images is also very powerful. Captions get read as often as headlines and sub-ads.

Um, but as you start getting into the exploratory copy, when you're making claims, when you're showing your benefits, when you're showing the proof of those benefits, Um, the features that deliver that, Uh, we… We find it difficult to change the copy enough That we could detect something. So this… the form… the structure that I'm going to give you here is designed to generate what we call corner cases. copy that is significantly different, that appeals to a different kind of research mode.

This is some research that Brian and Jeffrey Eisenberg outlined in their book, Waiting for Your Cat to Bark. This was published in 2009. I found this very freeing and have used it in the agency ever since. And… but they surmise here is that there's a quadrant here.

You have people who are going to make decisions and research in a very logical way. Uh, and then you have people who are going to research in a very emotional way, make decisions that way. Uh, they're also going to be quick decision makers, and they're gonna be deliberate decision makers, so people who are ready to take action versus people who are still gathering information. And this gives us a really nice approach for creating these corner cases.

The Eisenberg brothers gave them names. So, those that are logical decision makers, And quick decision makers, we call them competitives. These are the people whose first question is, what's in it for me? How will you make me better?

How will you make me better at my job? How will your ski products make me a better skier? These folks are generally very brand loyal when they find a brand that has made them better, and they stick with it. But they really need to understand the underlying philosophy of what you're doing, and why you're doing it, how it improves their life.

Now, staying with the logical decision makers, if we look at a deliberate decision maker, We call these methodicals. And these folks need to know the details. They need to understand how you do what you do. They want to see the reports.

They want to see the case studies. They will not reach out to you, they will not take action on a webpage, until they already feel like They know most of the answers. So, um, this is where we'll spend, you know, on our pages, more time explaining why we do what we do. Uh, showing screenshots.

These are the people that will be reading that sort of thing. Then we jump to the emotional quick decision makers. So, these are the spontaneous folks, and they are just looking to take action. So we're not talking about personalities so much here as where they are in the funnel.

So someone early on will be coming in a more deliberate phase, so they may be coming in a methodical mode. researching, they're filling out their evaluation spreadsheets, or they're evaluating, and… Um, favoriting and saving products that they're looking for. And then, when they've made the decision, they're like, yes, I'm gonna do it. All they need to know is where they go to take action.

Where's the form? How do I sign up for a demo? how do I buy this product? How do I find it, put it in my cart, and go through checkout?

So, the same person can come in a very methodical, Or a very spontaneous mode. finally, we've got people who are going to make decisions emotionally and deliberately. And these folks we call humanists, because they need to understand, because that emotional mode, they need to understand how they're going to feel If they take action on your website. They need to understand who you are, who they're dealing with.

These are the people that read about pages. They want to understand the community that they are joining. So, these are four very… different sorts of targets that we can use. to, um, to create these corner cases and generate copy that is significant different.

Now, um, this is, uh, probably something the slides will be sent to you guys and made available, but you might want to take a picture of this, because this is kind of how you handle the page, such as a homepage, where you're dealing with, perhaps, multiple modes of research. Those quick decision makers there at the top. Spontaneous, looking for action. This is why on landing pages, sometimes it benefits us to put the form at the top.

For those people who are coming back, who have made the decision, to sign up for the consultation, or get the demo, They want to see that form there, and they will take action. The competitives also need to see very quickly what's in it for me, and give me a reason to continue reading your page. So they want something that says, here's the benefit, And something that won't tantalize them and tease them to continue scrolling. Our deliberate decision makers, the methodicals, and the humanists will scroll more likely, so lower on the page is where we can continue to build out the case.

either the details of how we do what we do for the methodicals, or the reasons that they should deal with us. testimonials, um… pictures of clients, pictures of our employees, things like that. Those are the things that the humanists are looking for. The beauty is, language models do not have these biases.

We can confidently tell them to act like somebody else, and they will produce copy. without those biases. Now, don't get me wrong, language models have built-in biases that come along with the a body of knowledge. total human knowledge that they were trained on.

Uh, but when you set them on a task to do something specific, they step right in. So we're going to use these LLMs as our sociologists, and what we're gonna do is, uh, we're gonna leverage a, um, leverage a tool called the Myers-Briggs Type Indicator. And Jeffrey and Bryson Eisenberg did the research to tie these two things together. This is a personality test.

It came out in 1962. Uh, it's based on Carl Jung's psychological types, it's been used for career development, team building, personal growth. I am an ENFP. Last time that I took this, and it has been fairly steady.

When I was a younger man and building companies originally, I was a little bit more ENTP. And you don't need to know what the type indicators mean. Uh, the most important thing about this is that all the language models have been trained on Myers-Briggs. They understand it, and there are thousands and thousands of studies that have been published on this, and it's been used… So, they know very well Myers-Briggs.

So we can use this as a common vocabulary to communicate with the language models. For instance, if I say to the language model, um, Rewrite some copy for me, and uh… it is for a kind of person who needs to know what's in it for me, um, who's looking for Um, the underlying philosophy of what we're doing, uh, and, uh, is, is brands efficient. It begins to get kind of sloppy. Um, and one of the things we want to do is be able to consistently rewrite across pages, across sections on an existing page, and if you're an e-commerce business, we want to be able to consistently rewrite those product, um, product descriptions.

So, the competitive is, uh, on Myers-Briggs, will have an N and a T in there. signature. So they might be an ENTP, they might be an INTJ, Uh, but they have this NT, and this indicates that they're of this nature, that their internal locus and their thinking rational. The methodicals.

These are sensing judging. And again, you don't really need to understand what this means, but what this means is these people believe that there are rules. These are the people that you want working at your utilities, because they are the guardians. They make sure that rules are followed so that when you turn on the lights, the lights come on.

When you turn on the faucet, water comes out of it. Uh, our IT people are often SJs. There are certain rules, And they are the guardians, and it's often frustrating for those of us that want to move faster, but thank God for them, because otherwise, Chaos would emerge. Those emotional quick decision makers, those spontaneous guys, they're sensing, perceiving.

So, sensing means that they take cues from the external world, But, uh, rather than believing that there are rules, they also take guidance from you. They get action or energy from outside. Uh, of them. So, they're just looking to take action.

And then the intuiting and feeling humanists, like me, They have an inner locus, but they… they get energy more through how they feel about things, rather than the competitives that are tease, how they think about things. So the research indicates that these are, uh, for the primary ways that people are, um, moving through the world, and they… very, very well onto the modes of research that we're going to be talking about. Um, and this becomes our language. So, what we're gonna do is we're going to let the… let an agent do the rewriting.

And, um… Uh, there is a sign-on… a page on my site where you could go and get the messaging skills and the full prompts. Um, you can pull this very simple prompt and drop that into… drop it into any one of your chats, that is simply like, hey, I separate people coming to my website into groups based on their Myers-Briggs type, Those that have an SJ in their types are called methodical, those with NT are called competitive, And this just allows us to use this language. We can say rewrite for a methodical, rather than saying rewrite for Myers… Briggs MBTI type that it contains an SJ.

Um, now your chat, or your gem, or your project, or your custom GPT, We'll speak this language. When we are, uh, looking at our website, um, we want to understand um… kind of… who is coming to the site? And, um, do they have any proclivities towards any of these research modes? So, one example I pulled here is for Calm.

com. Um, and the prompt I said is review this website, and I pasted in the URL there. based on information available about this industry, describe the primary visitor personas that this website should be addressing with its copy. Now, the skills that I'm going to be providing you on that page, or that previous QR code, has a much more, um, a much richer skill for generating ICPs.

But this is a great place to start, especially for those of you that are primarily using chat. Uh, I put this in, um, I did not have any supporting documents around… for comms, so this is strictly from their website and industry knowledge, but they came up with a stressed professional, They said this person is most likely to come in a competitive or a methodical mode. Um, that thinking layer. And, um, it gives you some clues about who they are.

So they are ambitious and goal-oriented. Um, they are, um, they see mental wellness as a tool for improving their performance and productivity. So you can hear that what's in it for me on the competitive side, right? It also came up with the Anxious Seeker.

This one, it said, is a humanist. Uh, this is somebody who's looking for personal growth and well-being. Um, they may be dealing with anxiety, trouble sleeping, emotional distress. And they're looking for a compassionate guide.

They're looking for… A self-journey in inner peace. So you begin to see how this corner of our model is very different. And the third, uh, ICP that it generated was the sleep-deprived parent. So this person sauna is overwhelmed, They're just looking for… to solve one problem, how do I get to sleep?

They're going to be practical, they're looking for on-demand solutions, and that screams that spontaneous, um, that spontaneous mode. But early on, if they're not up in the middle of the night just trying to get back to sleep, They may be coming in a methodical mode. So this gives us these… well, these personas are based on personalities, it does give us a starting point. for what we might be looking for in our, uh, as we're rewriting copy.

So, what I want to do is I want to give you an ear. I want to give you an ear, like, playing music for you. It will give you an ear for how writing for these different coroner cases, uh, changes, how it's different. Um… This is, uh… I took this fact because I think it did a good job of describing what Calm was, and gave us a good sense of what their brand voice was.

Uh, and it is what is calm. The first thing we want to do, uh, when we are, um, evaluating our copy is ask it So who do I tend to write for? So, you could do this with a copy on your website, point it to your homepage or a reasonable landing page, and say, um, evaluate this copy, and tell me which of the four types it's most likely to appeal to. And then paste that copy in, or give it the URL.

For calm, it said, well, while the copy contains elements that appeal to all four, The strongest appeal is competitive and methodical. And so, if you'll remember, we had one ICP that was pretty strongly humanist. We might be missing the boat on this one. Um, and then it explains, um, you know, how it talks about the number one mental health app, It's very competitive language.

Uh, the Copy Friends of Wellness is a strategic endeavor optimizing and self-improvement. Um, these are the sorts of things that, um, really scream at competitive, uh, speech. And then, uh, there is evidence, reliability, those are very methodical values. So, if we look at it on the research mode, we've got, um… we're doing a good job of talking to the competitives and methodicals, What do we do, though, for that humanist Uh, that humanist ICP, how do we, um, do that?

So I took that same chunk of copy, and I said, okay, now rewrite this copy to appeal to humanist readers and explain your choices. I am still of the belief that our teams of… our writing teams are doing a good job of getting the important points onto the screen. Um, you know, giving a language model a bunch of ICPs and asking it to write copy for you has proven to me not to be Um, that effective. And so, I'm really focusing on using language models to rewrite the message that you have already developed, your marketing team has already developed, Um, rather than… creating from whole cloth.

Now, as you've, uh, if you'll… when you take the prompts and the skills that I'm going to give you, and you Load them up with your brand voice and your personas and your brand guidelines, and examples of your writing, It may get good enough to write copy from whole cloth. For now, we're going to talk about rewriting. So… Rewriting this, um, let's just take a little sample of what it was before. Column is the number one mental health app designed to help you manage stress, sleep better, and live a happier, healthier life.

Our research-based tools help you build life-changing habits to support your mental health. Talks about their 500-plus sleep stories. Um, uh, case studies appeal very much to those thinking types, methodicals in the the, um, uh, competitives. Um, 60-second breath bubble, there's a little bit in here that's, that's, um… spontaneous.

And now let's compare this to the rewrite. Calm is a space for you to discover a more peaceful and fulfilling life. Okay, already out of the gate. We're having this more humanist voice.

We believe that caring for your mental wellness is a beautiful journey, and our purpose is to support you every step of the way. So you're hearing that support? We are there for you, very humanist. Whether you're looking for… looking to find rest, cultivate inner peace, or simply connect with yourself, Our tools are here to help.

And it talks about the gentle sleep stories, but instead of listing that there are 500-plus Um, uh, sleep stories. Um, it talks about the immersive soundscapes and guided meditations that are crafted to help you drift into a state of deep rest. When you feel overwhelmed by a cascade of thoughts, our simple, Breath bubble can lovingly ground you in the present moment. Okay, a couple things are happening.

I know that you're hearing this difference, and I'm also hearing a gag reflex. And if you're not getting a gag reflex from this rewritten copy, then there's something wrong. Because what that gag reflex is, is your confirmation bias trying to drag you back into the way you've done things before. We know we have testable copy when everyone's looking at it going, My gosh, is it… is this really?

And this is why we test. Because is this more effective? Would this be more effective for these? These humanist visitors?

We don't know at this point. Um, our intuition may be right. This is going to be the difficult part for you as you negotiate these changes. The fallback always is, we're gonna test it, we're gonna know if this is too big of a risk, Because the flip side of this is, these are the sorts of changes, big changes, that will substantially step function your conversion rate, When you pick the right voice, or combination of voices on these pages.

Let's do one more. The spontaneous reader. Notice how much shorter… This copy is than the original. Claim is your go-to app for instant calm and better sleep.

Feeling stress? Our 60-second breath. Bubble is the ultimate quick fix. You hear this?

action, I'm impatient, I want to know what to do. Be out in minutes. Get instant access. Interestingly enough, what we find is, uh, calls to action and ads are generally written in this spontaneous voice.

Um, you know, to build urgency, to build immediacy. Um, and so this is, um, a… very helpful mode to write for in those situations. The flip side of that is, you can just keep pounding people over and over again with these high-urgent ads, Um, if you wrote your ads in calls to action in one of these other voices, You might find them more inviting to different… Visitors app. You can go in and you can ask that to combine the two of, uh, the two together, but, um… I find that sometimes this is a little, um, jarring.

Um, I have not tested this, we have not tested this in particular. So, use this with, uh, use this with care. We're talking about the top-ranked and proven research-based methods To achieve a happier, more balanced life. So you're kind of swing back and forth between the two.

So, getting past that gag reflex, Ask the language model to explain its choices. So, uh, we rewrote the motivations of explaining the choices. This rewrite had a blend of two very different motivations, competitives and humanists. And it will go in and say, we chose top-ranked and research-based message for that.

Competitive persona, Um, and the number one mental health app, So this will help you as you're selling this very different copy, as to why these sorts of decisions were made. Um, for that humanist side, uh, achieve a happier, more happy, more balanced life. These are the things that I think your ear is already hearing. Um, and so… We won't dive into that.

Um, resist editing, if you can. You want to remove inaccuracies, you want to make sure that your trademark notations are not… are marked. You want to add links, and you should probably pass it by your SEO person to make sure that you're using the right terms. Um, for testing, though, I would argue that… You don't want to test anything that the SEOs won't pass.

Um, so, uh… Test it, and when you find copy that is working really well for you, Maybe then have the SEO folks, um… do the magic that they do in that. So your ICPs will give you a clue. Let's look very quickly at a B2B example. This is an HR services company, they, um… They take over the HR, um… tasks of, um, small and medium-sized businesses and enterprises.

And they have two primary personas, the chief HR officer and the chief financial officer. Uh, the CHRO, they want to balance strategic initiatives, with genuine care for employees. They are under pressure to demonstrate value to the CFO, though. The CFO, on the other hand, they want to focus on profitability and risk mitigation.

They're kind of skeptical of HR's value, just because those methodicals, you know, And so, uh, there is a natural tendency here, uh, there's a natural, um, tension here, This… from reading this excerpt, I would certainly say that the CHRO is most likely to come in a humanist research mode, And the CFO is most likely to come in a methodical research moves. So this gives us our initial path. Um, just very quickly, humanists value empathy, personal growth, emotional well-being, and harmonious relationships.

Methodicals value stability, security, risk, management, practicality, and efficiency. And at any time you can ask your language model, like, tell me, tell me again, what is a humanist, tell me again what is methodical. Just to kind of level set you. This is a section of a landing page.

Now, this landing page is targeted for, um, chief HR officers. So the ads are really targeting the CHRO, and this is the landing page that brings them in. This is the portion of the page that we call the claims portion, where you're listing the things that you do. Reduce overhead, tangible ROI, risk mitigation, operational efficiency.

Um, so it's an important part of a landing page. Our hypothesis is, um… Well, the first question we want to ask is, who does this appeal to? So, what voice… What research mode does this appeal to? So again, the prompt is, uh… that prompt, uh, you know, telling it about our NTNFSJ.

Spontaneous, etc. Uh, so yeah, I want to evaluate the copy on my website to see which of these writing groups this style appeals to. Please analyze the copy, and you paste that. Copy in there, and here's what you find.

Uh, the website copy is, uh… Most closely aligned to the methodical group. Uh, which is the sensing judging, practical benefits, efficiency, risk reduction, clarity. So, this, the voice for this B2B company, and quite honestly, for many B2B companies, is this very methodical voice. But, this is on a page that's appealing to someone who has tendencies towards humanists, at least that's our hypothesis.

So we don't want to appeal to the CFO on this page. We want to appeal to the CHRO. We're squarely in that methodical mode. We need to move this copy more towards humanist.

So, rewrite the copy to appeal to the humanist group. Here's the original. Minimize risk. Work with a U.

S.-based team that understands federal, state, and local labor laws, reducing the risk of legal issues and penalties. How was this rewritten? Headline was changed to Protect What Matters Most.

Partner with our caring team. Who understands the human impact of compliance, protecting your people, While fostering a workplace where everyone feels secured, secure and valued. This is quite a difference. Are you getting that gag reflex, huh?

Are you? That's a good sign. You've hit the point, so you're talking about compliance. You're talking about, um… making things better for your people.

But you didn't do it in that methodical mode. Let's try this one. Free up resources, spend less time on payroll, benefits, and other HR functions. Focus more on the strategic priorities that drive your business forward.

Unlock true strategic potential. Free yourself from administrative burdens to focus on what truly matters. So we're hitting the same points, just with different words. Supporting your people's growth, building meaningful workplace relationships, and bringing your organization's values to life.

Instead of driving your business forward. So these, you should hear, are quite different. And if we do this with the entire… all of the copy, we have set ourselves up for a test. So, the original… methodical copy is the control.

Uh, we will change this to the variation. run an A-B test. The, uh… the… the goal that we're looking for is having people sign up for a, um… for a call on the calendar. And our testing tool will tell us which variation was a winner.

Um, this variation never actually went to test, so this is not the graph for this test. Um, but I think most importantly for this is that you hear the difference in the music of these different ways of writing. So, layout ideas. Um, we've got multiple personas, multiple rewrites, how do we handle this?

Well, uh, I'm a fan of simply repeating them. So, minimize risk. is in methodical voice. Protect what matters is in humanist voice.

And when you read through here, it doesn't sound like you're repeating yourself. The people reading aren't going to realize that you're making the same points over and over again. Only you're gonna know. But… those humanists will come through and, um, gravitate to the messages that appeal most of them, their own confirmation bias will set in.

So we wouldn't necessarily do this on the… to the CHRO page, but if we had a homepage or a landing page that we weren't sure which persona was coming to, then we could double up on the copy that we're writing. Um, as I showed you, uh, you can lay it out. Where you've got, um, information for a spontaneous and competitives at the top, And then blades down the page that, um, deal with methodical and humanist concerns. You can alternate those, you can put them side by side.

That is the hard work that your designer must do in making sure that the visitors' eyes are getting to the parts of the page. that move your value proposition forward. Um, you can rewrite for both. I showed you an example of this, so I won't labor on this.

That is, um… so that is really what we've got. Next time… actually, this is the wrong, uh, that shouldn't be there. So, um, what I want to do is take some time for some questions. I would welcome you to unmute yourself.

Let me answer those specific questions. For those of you that want to hang around, I will take you through the process of using the skills that I have for you, um, and how we build our messaging agents, we build one of these for every one of our clients. Uh, but was there someone that had a question? No questions?

You could type them into the chat as well if you want. I've got the chat up. How about just a little bit of feedback? Was this… is this model helpful?

Is this working for y'all? Is this something that you feel like you can use? There are… there are people. Remove fear at the copywriters department.

That's… that's very interesting. Um… MR, are you willing to expand on that? Is that Michael? You want to unmute and expand on that?

Because this is, you know, selling this internally is really… I can hear you, yep. Yes, probably I'm not sure if you can hear me, but that's great. So I experimented a lot with Creating copy, providing tone of voice and all that stuff. And in the end even sending already existing landing pages to the same agent in order to understand If it hits a tone of voice, if there's something probably wrong or whatever, and the biggest challenge was and still is making my colleagues accept the results from some artificial Whatever.

So they are still denying it, ignoring it. That's still the biggest task from my point of view. Maybe you have some advice. Yeah, well, so the advice we have is, uh, and this… This might be a little rogue, but, you know… run an A-B test and bring them the data.

Um, and some of your businesses, you're going to have the brand police, and they're… very specific about consistency across brands. Um, they don't tend to think in terms of multiple kinds of buyers. And you… you very well may be shackled. But those areas where you might have a little bit more flexibility, uh, maybe on the ad side, Uh, in emails, in blog posts, where you have a little bit more flexibility.

It's difficult to do testing on blog posts, they typically don't get enough traffic to run an A-B test. Um, social media posts, If you can, um… Test out these… these different, um… modes of research in those places, and um… you know, kind of anecdotally begin to understand what's working. that might support that, but if there's a… if there's an organizational barrier, just like silos, Silos between ag people and the web people, um… those are hard to… those are hard to break, um, especially in larger organizations.

Any other, um, questions or comments? Well, I mean, let's, uh, let's move on, and um… So, the, uh… Those of you that answered the poll, most of you are still working in chats. Um, I'm going to, uh… Provides you with… so the prompts that I've given you already in the slides are probably the easiest way for you to enable your chats. But for those of you that are interested in developing a custom… Um, GPT, a messaging agent like we're talking about, um, or, uh, Google Gym.

Um, these are really, really very effective ways to have a co-writer working for you. on the AI side of things. So, the process is to collect your internal documents. We do this for our clients.

Generated deep research report on the client's industry. Um, and I'll show you the way to get a good deep research report. Use that and all of those documents in a custom GPT or an AI assistant, And then you begin asking questions. So the things you want to collect are personas, obviously, brand guidelines, any industry reports that you have, Brand voice documents, writing samples are very helpful, and analytics data, if you can get access to analytics data.

Um, to help the, um… the language model understand where people are coming into the site, what's working for them, what products are most common. What calls to action are working the best. All of this will feed into that. The folks over at, um, Deep, uh, Trust… insights have the best prompt that I've seen for this.

So it's called the Casino Prompt. And, um, if you take this prompt, drop it into your chat, Uh, it will walk you through the process of defining who the contact… what the context is, what your audience is, the scope, the intent, the narrator, the outcome. And you're asking for, you know, in terms of scope, you're asking for things like, um, I want to understand what the, um, Everything about the industry, the business models, the finances, the competitors, Um, I want to know everything about this industry.

And the report you get is going to be, 20 to 40 pages. Um, it will be a very good read, um, unless you are already well-steeped in Um, especially for agency folks, if you want to understand your client's business, this report will be great. But it's mostly valuable. Um, for… training this model, so that the model understands the industry very specifically.

So you use this to generate a prompt. And then you use that prompt to generate the deep research report. And all of the language models have this deep research mode, You paste the prompt in there that was generated by the casino prompt. Um, and as I said, you'll get a long report, um, detailing all that, uh, all that your agent is going to need in order to, uh, to move forward.

So, in this case, I'm using Google's GEM. You could do this with a custom GPT, uh, Claude Project. Um, create a new gem, Um, enter this simple agent prompt, but I would, um, I would recommend using the prompt, um, that, uh, I provide you over on, uh… on the site. And let me drag that over here for you.

So, uh… Over here, this is… this is the skill for, um… Custom assistance, and the prompt you'll want to use is this one. So you just copy this, and you put it, and that will be the one that I think you'll have better luck with. This is, again, that QR code, that link to get to that page that I just showed you, this one. The one before that.

this page. There is instructions for, um, custom GPTs, and there's instructions for those of you that have agents, uh, skills, like Clonko work. Um, this is a skill file that you can use. So, when you're building your agent, you give it a name, Uh, you paste a name and description, you paste that system prompt that I just showed you, or the one from the previous slide, And then you dump in all of your documents, it now knows a whole lot about your industry and your business.

Um, here's some example, um… documents that I put into this particular one. The prompt I have, all you have to do is enter the URL of the website that you're working on, and it will begin an onboarding process, and walk you through what additional information that it needs. In order to, um… become a good, uh, partner in writing and rewriting. The skills have a number of capabilities.

Um, so you can ask it to evaluate copying. what, you know, what mode am I currently writing in? rewrite for the different modes. There's a dejargon capability, so if you say de-jargon this, it will, uh, uh, take… make sure that you don't have any jargon in your, uh, copy.

recommend page placement. So, um, if you're like, okay, I've got you know, this copy for these different modes, how do I… put this on the site. Should it go near the top? Should it go near the bottom, etc.

And, um, in this case, I have a much more in-depth capability for generating ICPs. Um, so I think, um, if you don't have personas and ICPs, um, or, they aren't angled towards messaging sufficiently enough for you. Sometimes they're a little too focused on demographics and… product desires and stuff like that. This will generate ICPs with these modes of research in mind.

So that detail is on this page for you to go visit. And I think, you know, with that, you should be able to, um… really build out one of these messaging agents. Next week, uh, here at the same time, I'm going to be talking about ad landing page alignment. And what we're doing with the language models to evaluate Um, the ads you have, and making sure that your landing pages Keep the promise, and… drag people through.

What we find is that There will be ads that generate a lot of clicks, but not a lot of conversions, And right now, what people do is they move on to a different ad. Our point is, wait a minute, if you could… Make some small changes to the landing experience, you might find that those clicks did have some qualified prospects. And customers in them. So that's what we'll be talking about next week.

Um, and are… if there are, uh… not any more questions, we're a little bit over our allotted time, but… Um, I appreciate you guys staying along with me. Any other questions or thoughts? Concerns? I'll put my email address in here again, so that as you're implementing these, skills and things, feel free to send me an email.

Um, this is, uh, these are new tools for us, and so I'm very interested in Um, getting feedback on how effective they are and how we can… make this, in general, our services better. I have a comment, um, can you take one? Yes, absolutely. Yeah, I just want to thank you for, um, talking about the CTA the way you did, because, um, I feel like… There's this thing that people think that It has to be high pressure, you should put it throughout all the copy, and I've never loved it, because I used to be in sales, and I was really good at it, and that was never my selling strategy.

And it just seems to be this thing, like, make it brighter, make it, like, and it's like, no, I mean, if you read the top part, like, it's kind of rude to ask them right now to do that. So that was really good. I actually never had any one point that out before and explain. that that's for the more spontaneous person, which is why I don't resonate with it.

And yet, you know, it's this understanding that this is what you do on a web page, you know? And it's like… Yeah. Um, yeah, I work with lawyers, and they used to have different lawyers writing things, and then I would edit it, and, you know, someone would come in with such hardcore a hardcore approach like that, and it's like, well, this doesn't fit the voice of the rest of… but I see where they're coming from, and then you start kind of losing confidence of, like, oh, well, maybe I should be doing this.

You know? But you just… thank you for giving, uh, like, a reason to think it through better and understand… like, I mean, imagine that fall map did that. piss everybody off, and no one would want to ever buy that, right? So, anyway, so thank you, I appreciate that.

There's a… there's a formula that Roy Williams has that really helps me, uh… kind of understand the different roles that our communications have, and so… Salience is that state in which Um, our brand or our message has, um, sunk in to the chemical memory of our visitors. So, we've kind of made it through all of the gatekeepers in their brain. And it is a function of frequency times relevance. So if we think in the terms of ads, frequency, we know that they need to see our brand a certain number of times, And past that, they've developed a, um… an automatic level of trust.

You know, regardless of… almost regardless of what message you're associated with that. So it's frequency times relevance. So the things that are more relevant don't require as much frequency. Um, but when you combine those two things, and every ad is gonna do two things, people are gonna see it and not respond to it.

But they will be seeing your message and your logo over and over again. When something is relevant, When you add it to the frequency, then those two things combine, and they're like, oh, this is somebody that I should explore. That's when you get the click. That's when you get the buy.

So, um… people are a little bit too focused on that relevancy part. Like, let's make it, like, get them to click, click them to click. Um, and it ends up making your brand feel like a salesman, right? Um, so, um… what you want to do is also take into account frequency.

So it might be softer language, but you are getting value over time. your, you know, your ads decay over time because people get tired of just being buy, buy, buy. When they're not getting any idea, especially those humanists, think of them. they're not getting an idea of what your values are.

So you need to… your ads need to play both roles to maximize them. Um, the question in chat was, who was that? Roy H. Williams?

Um… there's this really interesting business school here near Austin called the Wizard Academy that he built. He's an old ad man, you know, $25,000 a day admin, um, and he's made people a lot of fortunes. But he built this really interesting business, uh, business school, where they say, why don't we try to do on purpose in our communications? what writers and poets and artists do accidentally.

Because they do a good job of getting into our… Hearts, which is that salience that we're talking about. So I would check out wizardacademy.org if you're curious about that, and sign up for his Monday morning memo. It's usually enlightening and… And quirky.

Thanks very much, Heather, for making note of that, though. Anybody else? I will let you guys get on with your Thursday, and hopefully see you back here, um… Next week, to talk about ad landing page alignment. It's a powerful tool.

Bye, everybody.