Episode Transcript
[00:00:00] Speaker A: On our last episode, you picked your AI platform. So now the real question is, what do you do with it? That's why today we're covering the use cases that earn agents their most time back. A prompting formula that works on every tool that we covered last week, and one rule about client information that you need before you type a single name into any of them.
[00:00:19] Speaker B: This is part two of our AI for Agents series. I'm Jo Barker.
[00:00:22] Speaker A: And I'm David Belleville.
[00:00:24] Speaker B: And this is Amerilife Marketing Mentors podcast.
[00:00:32] Speaker A: All right, Jo, when you look at an agent's actual week, the appointments, the paperwork, the follow ups, what work should I be taking off their plate?
[00:00:40] Speaker B: An agent needs to look at everything that you just mentioned that you're doing and see where. What are those tasks that really cause you a lot of friction and just cause you a lot of frustration and then ask yourself, is this something an AI tool could help me with? So I wouldn't talk necessarily about any of your client data, anything that involves any personal information, but when you're looking at the whole process of what you do in a day, where are those points where you're like, man, if I just had somebody to help me with this, it would make my life so much easier. That's usually where a AI tool can come in and help you with that. A lot of the times for salespeople, they're really good at selling, they're great at selling. Marketing is like a second thought for them because they have a job to do and their job isn't to be creating content all day long. Their job is to be talking to people. So that's where you see a lot of agents take advantage of the different AI tools for their content creation and coming up with ideas and, you know, breaking down and making it simpler for them to create a marketing strategy and deploy those marketing tactics.
[00:01:36] Speaker A: Yeah, there are many use cases and we are going to dive into many of them today. But one call out I want to make, and this is according to HubSpot State of Marketing report they do every year. Teams that about a third of teams that are using AI are reporting that it's saving them up to 20 hours a week, if not sometimes more. And so that's, you know, that's a couple of days shaved off that you can now dedicate to, you know, your clients or any other work that you have been putting off. And so let's dive into it today, Jo, and talk about what use case do you think is being underutilized the most by our audience?
[00:02:14] Speaker B: I don't know if you can say it's being underutilized. They just haven't discovered it yet.
And I think it's different for everybody. So there's some people out there where they rely a hundred percent on referrals and they're not comfortable digitally and they're not comfortable online, but AI can still help them with that referrals, and maybe even it's setting up emails or setting up communications with them. So I don't think there's a one size fits all for everybody, but I think it's looking at what makes you successful. What does that process look like? And then breaking it down and saying, okay, how can I utilize a tool that can help make this process simpler?
There's, I think there's a lot of things where if people knew the capabilities, they would absolutely take advantage of it. But I also feel like there's a flip side of that where there's so much capabilities it's almost overwhelming. And before you know it, you've got 25 conversations happening with AI with all these different projects that are out there. And now you've got to figure out it's almost like you spread yourself too thin with too many projects happening.
[00:03:10] Speaker A: That's a great point. I think that people who are unfamiliar with AI or want to get started but don't know where to start, they think that, well, everybody else is so much more advanced with it, or I've already missed the boat, or when I do start, I need to be having autonomous agents working on my behalf, replying to emails and updating my website. And it doesn't have to be like that. In fact, there are many, many experts with AI who likely aren't using all of its capabilities because it's so vast. So understand what Jo is saying is that just pick one or two tasks that are holding you up throughout your week and, and just start there. Because, Jo, we know from experience that the more you use these tools, the more use cases you actually do discover.
[00:03:54] Speaker B: Yeah. And I want to give an example of something that's outside of the LLM wheelhouse, which is what we talk a lot about is those LLMs and those chat platforms. But I want to take it a step outside of there, of other AI capabilities. So an example would be, is if you're an agent and you're utilizing say Excel or Google Sheets and you're manually updating lead information and you're manually doing work, those are those friction points to look at is what are you doing manually? That maybe there's an AI tool out there to help you to where you're not doing it manually anymore. An example I'll give is we were working with an agent and they were keeping track of all of their client or all of their lead or prospect information in an, in a spreadsheet. They were then taking that spreadsheet and feeding it into their email platform. Well, the fact is, is if you utilize AI, you can do an integration that is taking that sheet and doing it for you and you're not having to handle any of it. And as soon as a new lead comes into that she, it's automatically going into your email marketing platform. So those are the type of things to think about as you're looking through the process. What are you doing manually that maybe you don't have to do? And there's probably a tool out there
[00:04:59] Speaker A: that can help you with that 100%. One of the use cases that I have found most valuable, that I think our audience would resonate with, is deciphering between data and analytics for those that aren't numbers people, but they want to understand what their social media data is telling them, or they want to understand what their email data is telling them. And AI is a, is a great tool to help you understand and extract information that can really help you make pivots and strategic changes that can move the needle for your business.
Do you have a favorite use case for AI?
[00:05:31] Speaker B: Yeah, absolutely. So a great example of that is if you take Google Analytics, which is tracking the user behavior across your website. So you're not looking at individual user behavior, but you're looking at overall user behavior of what pages are they going to, what entry page did they come into? You can pull some of those reports. Now don't get me wrong, Google Analytics has insights built within the platform, but you can also pull those reports out and put them into an LLM that you use that connects other platforms and ask it to analyze that data, find traffic patterns, find bounce rates. If there's a page where you're seeing where they're going from the homepage, they're going to the about page and then they're bouncing back and going to this page. They're obviously looking for something, there's something missing. Utilize AI to help you see what it is that you're unable to find.
[00:06:13] Speaker A: You also often recommend that people use AI to audit their website. And that's a really cool use case that usually took a whole team of people to understand and do and accomplish, but now you can do it in literal minutes. And that too has been really Beneficial especially for the Amerilife marketing mentors, because we are a new team and we did just launch a website not that long ago. And so, you know, whether it be this podcast or our website or, or the social media content that we put out, all of that is, you know, the data telling us what we should be talking about. And so, you know, these can be elements that you incorporate into your business
[00:06:54] Speaker B: as well when you talk about the website. A habit that we used to always see is when people create their websites, they say what they want to like they, what they want people to hear. And so we see that a lot. Where you have agents who are immediately on their homepage is, is all about like, I, I, I, I can do this, I can do that. And realizing when AI gets in there and you ask it to do a scan of your website, it's going to point that out to you and it's going to say, let's talk more about your client, less about, you know, what it is that you're trying to accomplish. Let's talk about how you're helping your client. And then you just change those wordings around and the sentiment changes to where you know. And in some cases it's passive language where you need to be active. And so there's those type of things that can help you with. And then as you mentioned, you can also use it behind the scenes to help you do a SEO scan, understand what's happening with your website, see where you're slow. It's going to tell you all those things. And you don't have to use a bunch of different tools to do it. It's one prompt.
[00:07:47] Speaker A: Yeah. And we're going to always be transparent with you guys. And there is a lot of ground to cover when we're talking about all of these different channels. And so even those of us that are in it every single day, AI is just such a helpful tool to help us keep track of everything that we do have going on. And then it's just a timescale saver as well.
However, if you're not using AI the right way, it could actually have the opposite effect. You might have to start over again. Or maybe you're spending spinning your wheels trying to generate an answer that you're pleased with. So let's get into prompting best practices because this is where it all begins.
There is a formula that we want to ensure that our audience knows. Do you want to start with the first one?
[00:08:30] Speaker B: It's giving it a role. So make sure that you tell the platform who it is. That you need them to act as if you're doing email, say email marketing. If it's broad and it's marketing, say marketing strateg.
If it's a research assistant, say that. Say it's a research assistant.
[00:08:42] Speaker A: All right, the next thing we do is we need to give that role. So for the sake of this example, we'll say act as an insurance agent.
Now, they need a goal, so we need to tell that insurance agent what to do.
[00:08:54] Speaker B: Yep, absolutely. So we need to give it that goal and then we need to tell it the audience that you're speaking to. So you could be an insurance agent, but who is that audience that you're going after? Are you selling short term? Are you selling, you know, Medicare? There's a big difference between who your potential audience could be. So number three on that with the role. And the goal is going to be the audience.
[00:09:11] Speaker A: And where does your audience live? That's the next part is you need to tell what this, what this content, what this message, where it is going. So you've got the role, you've got the goal, you've got the audience, and now you need a platform. Where is this message going? Is it an email? Is it a social media caption? Are you writing website copy? Those are the things that we're talking about when we say that you need to have a format or platform. And the last one, Jo, is give it a tone.
[00:09:38] Speaker B: Yes. That is, I think, where a lot of people miss, and I think that's also where a lot of people forget to edit. And so a lot of the times that's where you see a lot of people where they're like, it all sounds the same now because they're just taking ChatGPT and they're forgetting to give it that type of tone.
And so that's where you can come into play. Because you don't always want to sound the same. Sometimes you want to be educational, sometimes you might want to put some humor in there. There's different types of, you know, just tones you want to bring into the conversation or you want to it to be perceived as. And I think when you forget about that and you just copy and paste what it's saying, then that's where you end up with a lot of the posts that just all feel the same.
[00:10:12] Speaker A: Another great way that you help others understand what we're talking about is a GPS analogy. Right. You. You wouldn't just tell, you know, your Uber driver to take you to a mall. Yep. So this is what we're talking about. If you could, can you touch on that analogy?
[00:10:29] Speaker B: I mean, imagine if you, you know, if you're saying, hey, just take me to a mall. Well, there's a lot of questions that come along with that. The problem is, is that the LLMs aren set up to come right back with a question. The LLMs are set up to give you an answer. It wants to please you. So when you put it into a gps, it's going to pull up all these different malls and it could be from everywhere. And you already know where you want to go. Same thing when you're doing a prompt. You already know where you want to go. Give it to it, provide it to it, say it. So instead of saying, hey, I want to go to a mall, it can be, hey, I want to go to a mall that's located in Clearwater and I want to make sure that it has a store that sells these type of shoes. Also, I'm going to be there around these times. And I also want to make sure there's no tolls on my way and I want to make sure I'm not getting any, you know, spending any additional money. There's so much that you can expand upon and the more efficient you are at giving it the information upfront, the better the response is going to be.
[00:11:19] Speaker A: And we're now comparing that. And to bring it back to AI terms, to someone just simply asking, write me email.
You know, write me an email versus somebody that's saying write. As an email marketing strategist, your goal is to write a purp personalized thank you email to this list. I mean, you walk through that formula and all of a sudden the generated outputs are vastly different.
[00:11:43] Speaker B: Oh, absolutely.
[00:11:44] Speaker A: And what's important to note here is that a majority of folks out there might just be taking that first option. So by understanding these best practices and applying them, you're more than likely starting to distance yourself from your competition.
[00:11:57] Speaker B: And keep that in mind too, as you're building out content and how our consumers are starting to do it. You know, it may not just be, hey, find a Medicare agent near me. The average keywords on Google is 5 keywords, 5 to 7 keyword. The average search on an LLM is 30 to 40 keywords or 30 to 40 words. And so instead of find a Medicare agent near me, they're saying, hey, find a Medicare agent near me. I'm specifically looking in the Clearwater area who takes appointments Monday through Friday between these times. I also want to make sure that they specialize in new to new to Medicare. Also check their Reviews for me make sure that they have over 4.5 stars or over 4.8 stars. There's so much you can say and the more precise you are, the more precise. And it's amazing how detailed of an ANSW you can get and it can match you better to what it is that you're looking for.
[00:12:43] Speaker A: And the more you use these platforms, the better they are at understanding you. And you'll. You. It sounds silly, but you will develop a relationship and this AI does genuinely get to know you.
[00:12:53] Speaker B: Oh, absolutely.
[00:12:54] Speaker A: So if you're using it for work cases, then you're over time you're going to already have that head start to where it might just take that single prompt and you're getting those answers you're looking for a lot quicker. But. But one thing to note on that point is that you should never just take that first response because we have a tool at our disposal now that's conversational. It's not just that 5 keyword search on Google. You really want to press it and continue that conversation. So you are getting the best possible outcome.
[00:13:22] Speaker B: Yep. You want to keep having that conversation and then ask it to ask you questions. I mean, there's things that you're not going to think about and then you can say, hey, ask me one question at a time until we get to a point where we feel like we're, you know, you're comp. More confident in providing me a solid answer. You can easily say that. And I say it all the time because if you don't say give me one question at a time, it's going to give you 15 and you're going to get frustrated real quick. So it's an easy way to just say, hey, ask me one question at a time until you feel like we have enough information.
If you do that and it comes back and you're like, oh man, it missed that. Go back up and edit it and say, be sure to also ask me this.
[00:13:54] Speaker A: Yeah.
[00:13:55] Speaker B: And then that way we'll still ask you all those questions and then it will touch upon that other topic that you think it missed.
[00:14:00] Speaker A: Another great tactic that you can use is have AI interview you.
[00:14:05] Speaker B: Yes.
[00:14:05] Speaker A: So I know that this is something that we both do. You want to train your AI, you want it to understand your business. Right. What your goals are, what your mission is, what you sell, what your messaging is. And it can often feel overwhelming to know where to start if you're just kind of brain dumping all of that into AI. So instead prompts it to do an in depth interview. With you and maybe ask it to cap the questions. Otherwise it might just ask indefinitely. But ask it to do a 10 question in depth interview about topics that you think are relevant to your business. And that's just one little trick that you can use to also train your AI.
All right, Jo, as we wind down this conversation, we want to make sure that our audience understands the guardrails that should be in place. And one thing that's major, you mentioned this earlier, is ensuring that no personal identifiable information ever touches one of these AI platforms, right?
[00:14:57] Speaker B: Yep. No personal information, don't utilize it for legal or compliance review.
It's never going to be perfect and you're never going to be able to go back and be like, oh well, AI told me I could do that. It's never going to hold up. So just make sure you still follow those compliance things. Make sure that you're still reading and double checking and putting your own authenticity to anything content wise that it's creating for you. And just remember it is still a machine that's working with you. So it's never going to be perfect. And that's where you really want to, you know, keep working with it, but keep it within those guidelines and don't step outside of those.
[00:15:29] Speaker A: Are there any other mistakes that our audience should avoid similar to not relying on AI for compliance? Are there any other pitfalls?
[00:15:39] Speaker B: Yeah, I'm sure there's probably a lot. If you look at the lawsuits that have happened out there, one of them is the data. I mean, some of the, some of the data that it gives back to you, it hallucinates. And so just taking it for fact is, is risky, especially when a lot of people are feeding a ton of data into these platforms and they're just taking it for its word because it's so complex that it's accurate. It, it may not be. And in my opinion, even with Claude and I feed it a ton of data, I see errors a lot. In my opinion, it's not the best at computing data, it's great at coding not the best at data, and it hallucinates more than I would definitely like it to. So those are the things, and I think we said it earlier, that fact checking is going to be critical because the first time you put something out there that's inaccurate, one, you're putting your name on it. So now your credibility is going down. And then if it doesn't sound like you, you just lost trust.
And so that's, it's really hard to gain that trust, to earn that trust. And now you got to keep it. And you can diminish that relationship rather quickly by trusting AI too much.
[00:16:40] Speaker A: One thing to keep in mind when you're using AI is that what you tell it not to do is often as important as what you are telling it to do. So set up these guardrails with what we're talking about and tell AI to avoid these things. The last thing I'll say on the topic is there's also a lot of words, sentence structures and phrases that are automatic AI tells. Now these platforms have been around now for months and with ChatGPT's case, over a year.
There's a lot of AI generated content out there and it's becoming easier than ever to tell what it is. And that's another quick way to lose trust.
[00:17:17] Speaker B: I can tell now when it is a clogged made website, I can tell because the template, it almost always looks the same. So as soon as I see a website that's newer, if it has a certain layout, I can, I can tell immediately. Club did it because there's so many of them that look alike now. Same thing on LinkedIn. There is a certain phrase that it always does like where it's like, well, it's not this, it's this. Yes, it is non stop and everybody's posting it and I can't even entertain reading it because I'm like, oh my gosh, stop doing that. So we're well past the M Dash stage and stuff like that. We're, we're now realizing those phrases are non stop and so you almost can't trust it as much at all anymore.
And it is kind of sad to see how many people still don't realize, wow, we know that's not you. The other thing that surprised me is when people comment on LinkedIn and it's a telltale sign that they used AI to comment. And I'm like, you can't even come up with a comment by yourself. You have to use AI to give your own. Like now we're taking it too far in my opinion.
[00:18:16] Speaker A: No, that's a great point because we're in a trust building business and when you're on platforms like LinkedIn, your audience wants to know that you know what you're talking about. So if you're relying on AI, anybody can do that. So why should I go to you if that's what you are doing? I think that's a great final point. And just so you know, next week we are getting into how to develop content with AI and do it the right way. So it sounds like you and distances yourself from the competition. But in the meantime, for the Amerilife Marketing mentors, I'm David Belville.
[00:18:48] Speaker B: And I'm Jo Barker.
[00:18:48] Speaker A: And we'll see you next time.
