Episode Transcript
[00:00:08] Speaker A: What's up, everybody? Welcome to the Amer Life Marketing Mentors Podcast. My name is David Belville.
[00:00:13] Speaker B: And I'm Joe Barker.
[00:00:14] Speaker A: And today we are talking all things prompting, why it matters, how agents and advisors can use it to their advantage, and how to set yourself apart from everybody else. So without further ado, Joe, let's jump right into it and set the stage for us. Why does prompting matter when we're having a larger conversation about AI?
[00:00:34] Speaker B: I think it just helps streamline. Actually, it does. It helps streamline that process of getting the result that you're looking for faster. So I think the, the sooner that you can learn how to prompt, the sooner you can understand those best practices, then you'll be that much happier with the results that you're getting out of AI. You know, I tell some people, it's like, go going into Starbucks, and if you're like, hey, I want a coffee. I mean, they're going to give you a coffee. But if you really wanted a very specialized drink with certain milk and certain sugars, I mean, you have to say that and you have to, you have to give the direction. So that way the result is that much better. So, I mean, really, all in all, that's why it's critical, and it just makes your life so much easier once you learn how to do it.
[00:01:12] Speaker A: Well, not to mention, too, if everybody's going into the coffee shop and ordering a coffee, everybody's going to end up with the same.
[00:01:19] Speaker B: A hundred percent. Yeah.
[00:01:20] Speaker A: I mean, I, I love that comparison because that's kind of where you can get into. Into trouble if you're not prompting with best practices. You start ending up with content, ideas and generated responses that everybody else is getting as well, including your competition. So today we are going to talk about how to prompt with precision that will give you those responses that you're looking for, and you can focus more time on other.
Excuse me, aspects of your business.
[00:01:52] Speaker B: Absolutely.
[00:01:53] Speaker A: And so there's a formula that we like to follow and ensuring that there are certain criteria in each prompt. Do you want to outline those for us?
[00:02:01] Speaker B: Yeah. So we know that there are certain things that we definitely want to hit, and we can actually go through them one by one, if you want, and then explain kind of why it's important. Sure. So I'll start with the first one. The first one is the role. So the role is critical every time that you're starting a prompt to tell it who you want it to act as. And there's a couple of reasons why that's super important. One it sets the tone of, of really who it is and what you're trying. It's almost like you're saying who the assistant is that you're working with. I want this to be an email or you're acting as an email marketing specialist, you're acting as a content specialist, you're acting as a social media ads expert. As soon as you set that role, it says like, this is who my assistant is, is the way I always look at this is who I'm working with right now. At the same time, from the AI's perspective, it brings in the knowledge of that person or of that role, which is critical to meet those best practices of what would a social media expert or an email marketing expert know in order to build, you know, those best practices into it? So the role is critical, and that is definitely the very first thing. Before you start any prompt, tell it exactly you are this, or I need you to act as this.
[00:03:05] Speaker A: Yeah, Every single prompt should, should start the exact same way if you're following that best practice.
So the second factor that you want to include in every prompt is giving it a goal. So we've established a role. Now we need a goal. So if you're acting as a social media strategist, be specific, because there are many platforms out.
And so if your goal is to create social media copy, then that's gonna be a pretty general response and you might not get what you're looking for. But if you're specific about your goal, you need copy for an Instagram caption. Now you're starting to get closer to the response that you're looking for. But when you layer in all the elements, you really start to get ahead. So we've established a role and you need a specific goal. Tell that role what to do. What's the third part that we need to.
[00:03:59] Speaker B: So another part that you have to remember is the format that you want it to give back to you. So you can tell it, I want it in a PDF format or I want this for a certain platform, or I want this in a PowerPoint or I want this in PNG or JPEG. You can set anything that you want to set for that for whatever it is that you're looking for. And it will. It is amazing how many different formats that it can provide it to you in. You can even tell it, I want to be able to copy and paste this directly out of here.
So it's really important that you tell it exactly how you would like to see it as it gives that result back to you. So, and then that way the format's already done and you're not having to, to rearrange anything or reset anything.
[00:04:37] Speaker A: Yeah, well said. So we have a role, a goal in a format. The next pillar that you should always include is identify a target audience a hundred percent. Who is this content going to be for? Who is this messaging for? Be specific and ensure that ties back to your business goals. If you have audience audience segmentation, ensure that you're identifying which audience that you're speaking to. But a goal without an audience, I mean, at that point you may as well just be shouting into the void. So whatever generated response that you're looking for, make sure that AI knows who it's generating it for. That becomes really important in our industries because how someone looking for Medicare benefits is going to communicate differently than maybe young families looking for life insurance. And so being specific about what target audience you're going after will definitely help your messaging and make your usage of AI much more efficient.
[00:05:36] Speaker B: Yep. And you said it right there of the next step is now that you know your target audience is that topic. So you said it. You know, with that you're going after a family and they want life insurance. Well, make sure that you're putting that, that exact topic that you're looking for or narrow it down as much as you can into the topic of what it is you're talking about. You know, if we were to come out and say, hey, I want content about marketing for agents and advisors, there's a pretty wide array of topics that you could come up with and things that it could create on marketing for agents and advisors. But when we come in and we say, hey, we want to talk about AI in marketing for agents and advisors and utilizing email, that is very specific, that is really drilling down and the result that we're going to get off of just saying a general topic is going to be incredible compared to getting who knows what out of all of those results.
[00:06:23] Speaker A: Yeah. And something else that you'll want to layer in as well is the tone in which you need a response. So, you know, if you're generating website copy, then you might want an authoritative tone. When people land on your website, they want to know that you know what you're talking about. They want to know that they can trust you and they want to look at you as someone that can help them make important decisions for themselves. But if you are, you know, generating an email, subject line or content for an email, you might want to speak to your audience differently. So the tone in which we're hoping to get a response is also an important factor of this as well.
[00:07:01] Speaker B: Yeah, there's been a lot of prompts where I've either said, hey, I want to be educational or I want to, I want to be authoritative or whatever, you know, that language is. And really it's about thinking about that platform that you're putting it on and how you're utilizing it and which will help set the tone. And, you know, that really goes over the basics. But the other thing that you can do, which I, I think, you know, we're probably going to jump into here next, is those settings within the platform.
So there's a lot of things that, that overall you're going to want to keep that same tone or same structure, the same way that you present things. And that's where in those platforms you can go in and set settings. That is kind of like a lens that goes over all prompting and goes over all conversations. So it gives the AI a lens of like, okay, no matter what they ask me or how they ask me, this is how. This is the. We're going to be sure to incorporate this. So it could be a certain style that you like or certain words you don't like used, or there's a lot of instructions that you can give it to really, where you don't have to keep repeating it every single prompt of don't do this and do this.
[00:08:04] Speaker A: So that's such a good point. So most LLMs, whatever platform you use, has the ability to go in and edit those settings that you're talking about.
Another benefit of doing that is that that's just one more way to distance yourself from the competition. So, you know, if you have 10 people in a room and nine of them didn't personalize, you know, their settings, that one person who did is more likely going to, you know, get unique responses in return and be able to leverage those to not blend in into the background. So I think that those, those settings are important. You can give its character, you can tell it your own values and interests. So you know that that starts to get a little inside baseball. What we typically recommend is just get comfortable with using the platform first, those basic best practices, and then the more comfortable you're you, you, you are with using it.
Start to layer in these, these advanced tactics.
[00:09:06] Speaker B: Yeah, I would say as one of those first advanced tactics that I highly, highly recommend doing with any prompting that you're doing is at the end of the prompt and at the end of the request that you've put in there, know that's just the beginning of the conversation. And, and so say that to the AI and say like, hey, ask me one question at a time. A clarifying question to where you feel more comfortable or more confident in the response that you give me. Um, there's going to be tons of questions that probably could come up that could change the way that they it gives you the result.
So telling it one at a time to ask you those questions, you're just giving it that much more data and that much more information to give you the best result. And, and by saying one at a time, you're not getting hit, which I have before of 15 questions. And then I'm frustrated because that looks like way too much effort and I'm not doing that.
And so it's just so much more digestible from our side to say, hey, ask me one question at a time. Any clarifying questions that you have before you can provide the best answer for me, it's quick, it's easy. Usually it even tells you like, hey, I only have one more left. And at least you know when that's going to be done. And honestly, you can even tell it if it asks you four questions like, hey, this is the last question because I need a result.
If you're timely and impatient like I am, sometimes that's just a thing you can do.
[00:10:19] Speaker A: That's such great advice. While we're on the topic of AI asking you questions, you know, for everything we're talking about when it comes to prompting, that's tied to a specific goal or outcome that you're looking for. But in general, a practice that you might want to employ is just asking AI, your favorite LLM to interview you.
[00:10:39] Speaker B: Yep, absolutely.
[00:10:40] Speaker A: Especially when it comes to helping it understand your business goals, how you, you know, got into the business, what motivates you, what your values are. So, you know, these prompting best practices are great when it's tied to a specific outcome, but just for general knowledge, asking AI to interview you with one question at a time will really further its knowledge. And what that will do is create memories. And memories are an important part of kind of the rise of this technology as well.
[00:11:12] Speaker B: Yeah, in the settings that, you know, going back to those overall settings, that's where you can turn on memories so to where it will remember the conversation so highly recommend you go turn those on. That's also where you can turn off the sharing of the data to where you're not sharing it with others. But yeah, that is where you know, you can build it to where it does get to know You. One of my favorite things when asking it to interview you is to let it know that you're gonna be responding via voice in your own tone and the way that you speak and ask it to memorize that or at least get familiar with the way you speak and then ask it to provide the answers back in your tone and similar to the way you speak so that it truly is learning about you and it just gets it that much more efficient and that much quicker. Instead of having to edit and redefine or re. Ask things, it gets you to that end solution that much quicker.
[00:12:00] Speaker A: Yeah, I am always talking with my little AI buddy to the point where I, I need to recognize that I might be around other people.
[00:12:10] Speaker B: Yeah.
[00:12:11] Speaker A: And so, but I do that because to your point, it's allowing it to analyze how you talk so that you know, everything else starts to sound like you. And you know, if you are going to write prompts, you can deploy, you can deploy that same strategy and ask it to analyze how you write and memorize how you write for maybe any written communications that you're asking your LLM to help you out with as well.
[00:12:40] Speaker B: Yeah. In the beginning, I think, you know, if you can picture your, that LLM as your, your personal assistant and that truly is what it is. So imagine if you had hired a personal assistant. They're going to be doing the same thing, checking your emails, checking things to make sure that you know, you'll be proofing it for you, writing things on your behalf. It's the exact same thing except for instead of an actual person being there, you have, you know, this partner in AI on your phone helping you with that. So the more you can get comfortable with it, the better. And then the more information and the more you, the more you can have those conversations. It just gets so much more efficient over time to where a lot of the times now I don't even have to do a ton of edits because it knows how I speak so well that when I read it I'm like, oh, yep, that sounds like me. That sounds like something. And if it, if it's wrong, I tell it when it's wrong and I'll say, yeah, that is not something I would ever say. Don't ever do that again.
And it listens.
[00:13:34] Speaker A: You got to put it in its place.
[00:13:35] Speaker B: You do, you do and I do. Sometimes I get a little mad at it and I'm like, stop doing that.
But yeah, you do.
[00:13:42] Speaker A: I think that it's important that we also talk about guardrails. Right. Because for all of the benefits that AI and LLMs and this advancement of technology has, there are risks involved as well, especially working in the health and wealth industries where trust and relationship management and compliance are all extremely, extremely important to the well being of our business.
So when we think about guardrails, you mentioned the settings that you can go in and change. So you can just solidify certain guardrails that you always want in place or you can do it in each individual conversation as well. But do you want to talk a little bit to that?
[00:14:24] Speaker B: Yeah. You know, one of the things that you can do and again it can never act as compliance or an auditor or it has limitations. It is simply a tool that you're utilizing. But you can go into the settings and you can tell it. I am a financial planner and this is who I help and I need you to, you know, ensure certain regulations and double check certain things for me. But one important thing to keep in mind and it's just very, very glaring what it happens is it's AI and all of the tools can hallucinate and it can give you bad results.
And so it is. And when it does it, it's very confident when it gives you those bad results. I mean it's very confident in the bad answer.
And so you really do have to, you know, kind of put on that, that. Well, is this true with almost everything that it gives you 100 and so never, even if you put in those guardrails of saying hey, check CMS or check these regulations, you have to know it is still a device and it is still not 100% error proof. So everything that you have and, and especially anything where you're giving stats or sources or anything, just double check it, fact check everything that's coming through it. No matter how good your settings are or how much information you put in there, it is still a tool and it is still learning and it's, it's not 100 yet. It's getting better and better day by day. There are certain tools that I think are outpacing others and how good they're getting, but it's still a lot of errors coming through there. There's some where I'm like that is just wrong. That is just wrong. And it will be like, hey, nice catch Joe. And I'm like, no, don't try to.
[00:15:57] Speaker A: Sneak one past you.
[00:15:58] Speaker B: You should have caught that, not me. But and it was something so simple. So from something very simple to very complex, it can make a lot of errors.
[00:16:05] Speaker A: That's, it's a healthy skepticism that all users should, should have and continue to have.
You know, I, I also want to talk about what, I also want to talk about how you can improve, going back to prompting how you can improve your prompts by giving it examples to follow. Oh, absolutely. So by uploading relevant, you know, links or inspiration that you might have documentation that you want it to draw from or successful posts or campaign materials that you've had in the past, you can upload that into all conversations so that it's not flying blind and it has, you know, something to reference that you yourself have approved and been pleased with in the past.
[00:16:49] Speaker B: Yeah, it's kind of like whenever we've been rebuilding a site from somebody and we'll say to them like, hey, give us links of other sites where you really like the way that you know that the site is laid out or you like that user experience or the visual appeal of it. Give us those examples. Same thing when you're working with AI. The more examples you can give it and the more data you can give it, the better it's going to be. And then it takes away that part of the guessing game. Do you want a modern, do you want traditional? Do you want text heavy? Do you want photo heavy? It really starts to feed into that to where you can show, even if you're scrolling and you see a post that you like a lot and it grabs your attention, save those, just screenshot it and then put it into AI later and say, hey, these are examples of very eye catching posts or captions that I really liked. Keep these for future use of styles that I like or that I believe are eye catching. And then that way you can use those as reference.
[00:17:38] Speaker A: That's such, such a great point.
So we've covered a lot of ground here. We've established what every prompt should have in terms of basic best practices. We've gone over how to input settings to set yourself apart. We've established guardrails, we've established the fact that you must fact check everything.
[00:17:59] Speaker B: Absolutely.
[00:17:59] Speaker A: What haven't we touched on yet, Jo?
[00:18:01] Speaker B: There's probably, I mean there's so much we haven't, we haven't touched on. But you know, there's, there's a few like tricks that you can do, you know, to kind of streamline it. One of the things that we always say is, you know, if there's, if the conversation is veering off.
[00:18:13] Speaker A: Yeah.
[00:18:14] Speaker B: Or if it's not quite headed in the direction that you want it to, don't just continue the conversation and try to redirect it. Go back up and edit one of your last prompts to try to correct whatever it is that took it in the wrong direction or clarify something to where it gets it back on track.
So that's a great tool, is going up and editing instead of just continuing a conversation, because it can get really frustrating once you're off track of trying to get it back or it gets confused. So keeping in mind those LLMs when they're researching and when they're trying to give you answers, it is literally taking each individual word you give it, and it is researching each of those words for the definition. And that's how it is giving you those answers back. I mean, it's getting better and better, but if you have a word in there, one word could throw off why it's giving you the response it's giving. So going back and just simply hitting edit and changing it will save you so much time and energy and frustration when that conversation takes a left turn and you're. And you're ready for it to come back. So I think that's one and then another one, which I know I've heard you mentioned before, is the organization.
[00:19:18] Speaker A: Yes.
You know, this can really change the game for you, because especially as a business owner, you know you're gonna have repeatable systems, and you know that you're gonna have conversations that you want to reference time and time again. So there are tools on each platform that you can use to organize each conversation. So if you know you're having an ongoing conversation about email marketing, then give that its own space and home that you can revisit time and time again. Same with social media. Same with outreach messaging. Same with competitor analysis. Whatever the topic might be, I very much encourage you to organize those conversations. It will keep you on track, but it will also be so valuable for you to go back and reference and going one step deeper into that, a tool that we've been using a lot. You can now share conversations with your coworkers, colleagues, or, you know, fellow collaborators with your little AI companion. I think that's great. That's so awesome.
[00:20:23] Speaker B: Yeah, absolutely. Instead of having to pull it out and move it into texting or whatever it is, we can just hit share, and then the entire prompt and the entire feed gets shared between us, which just streams like that, streamlines that entire process and gets everybody on one page and. And it's sharing prompting at the same time. The other thing that I'll point out, and one that I don't, you know, we've touched in other videos, but I Think it's important is when you're looking at the AI tools, every tool is a little bit different in its strengths.
So where you might use one tool for content, you might use one tool for images, you might use one tool for fact checking or that breaking news, those trending topics. Get to know the different tools. And like, for Copilot, we use that one all the time because it connects all of the Microsoft Suite ones. Each one kind of has a different.
A different unique advantage over others. Sure. And I think a lot of people think they're all the same, and they're really not. Some of them are really, really, really good at coding. Some of them are really good at images. And so the more you can get familiar with them and the way that you can use them, you'll begin to categorize and organize yourself just based on that too. Like, hey, this is the tool I use for everything trending, this is the tool I use for social media content. And you'll begin to get really organized, and those tools will get better and better because they, they do specialize in certain things.
[00:21:42] Speaker A: Yeah. As you were saying that I. I've kind of thought back to previous advancements in technology. When I think about the dawn of search and all of the different search engines that used to exist. Oh, yeah, Ask Jeeves. Yeah, you remember that.
So, you know, the market was really saturated with them, but over time, there became one clear winner in Google. The same with the smartphone revolution. How many different versions of smartphones were there? And now, although you do still have numerous devices to choose from, I think that the iPhone has really distanced itself as the front runner. And so we are in the infancy of AI and where we are today, in 20 years, it's almost impossible to imagine where we'll be. But bringing it back home here is that.
Experiment with all these tools, find which one that you're most comfortable with and just know that over time this is likely going to change. But if you're giving yourself an opportunity to learn each one as they continue to advance, you're also giving yourself an advantage.
[00:22:48] Speaker B: Yeah, absolutely. The front runner right now, if you were to look at them, is chatgpt. If by far, you can just go to the App Store and see how many downloads it has compared to Claude, Perplexity, Grok or any of the others, it absolutely is killing it. But if you look at advances and if you start to look at recommendations and who is really starting to pull ahead, there's other tools out there that are, are coming in hot. Yeah. And now you have Google, who's come full force into the game with it all by bringing in the AI mode and switching its interface to that conversational style searching.
So we're just at the beginning of this and I think there is going to be some major shifts. You're going to see Google try to hold on to things and it may work. And it's going to be interesting over the next couple of years to see which tools advance the most and which ones garner the most trust and accuracy.
[00:23:37] Speaker A: Yeah, there is going to be a lot that we need to stay on top of as 2026 continues. I'm sure that the conversation we're having today is going to be much different than the conversations we have December of this year.
So just stick with the Amerilife marketing mentors all year long. This is a conversation that we're going to continue have and we'll stay on top of. Any major updates that you need, head over to amerilifemarketingmentors.com you can subscribe to our newsletter. That's going to be a great resource for you to stay on top of our content.
But bringing it back to prompting and AI Joe, what's your final thought? What's the biggest takeaway from today?
[00:24:18] Speaker B: I would say if you've never done it before, get in there and start playing around and trying it and you'll see how the prompting really does help you advance those responses. If you've been doing it for a long time, there's a ton of, you know, extra things that you can do. We probably will have to do an entire episode on, on advanced prompting, which we'll, we, I'm sure we'll be doing here shortly. Um, so just keep learning, keep, keep playing with the programs and seeing the capabilities that it has. I still have moments where I'm like, why am I doing this? Why am I not using AI for this? And then it clicks like, oh wait, I can use it for this too. So I think just keeping an open mind and testing it and just remember, don't give it personal information, be smart about it and always double check the responses that it gives you.
[00:25:02] Speaker A: Yeah, it's kind of crazy how quickly it's become such an integral part of our day to day.
[00:25:06] Speaker B: Yes. Oh, absolutely. I love it.
[00:25:09] Speaker A: It's your new best friend.
[00:25:10] Speaker B: It is truly.
[00:25:12] Speaker A: Well, I think that's a good place to end it. Joe, thank you so much for everybody who gave their time today to learn about this. If you want to learn more about prompting and AI best practices again, follow along with us. All year long. Amerilifemarketingmentors.com on all of your favorite social media platforms. And for the Amerilife marketing mentors, that's Joe Barker. I'm David Belville, and we'll see you next time.
[00:25:36] Speaker B: Have a great one.
