[00:00:45] Zach Hammer: Welcome back to the Real Estate Growth Hacker Show. On [00:01:00] today’s episode, we’re going to be talking about why you likely don’t need another prompt, the most perfect prompt, this massive collection of a thousand prompts that no one wants or asked for. You likely don’t need that. What you actually need is something else.

[00:01:13] Zach Hammer: And that’s what we’re going to talk about today. With me today, my co host, the co hostess with the mostess. I got Charlie Madison back in the house wearing back in style, the Hawaiian shirt looking very tropical today, Charlie, welcome back.

[00:01:27] Charlie Madison: Finally, it’s back.

[00:01:29] Zach Hammer: You were not feeling quite yourself for a while. So, no Hawaiian shirts.

[00:01:33] Charlie Madison: Winter is over.

[00:01:34] Zach Hammer: Fair enough, good call. So yeah, In this world of AI, to give some context, like the name of the game, the thing that people talk about a lot is prompting and prompt engineering and writing the right prompt and doing that almost to the point where I think people have started to go down this path of expecting the prompt to do a lot of heavy lifting that really you shouldn’t be expecting it to do.

[00:01:59] Zach Hammer: So how does [00:02:00] this show up? This shows up by thinking, I need to change this word or phrase to be the perfect word or phrase in order to get the end result. I learned about this cool prompting technique, so I need to do that in order to get the right end result. Really what I found is like that stuff does matter to a degree, right? Like it does influence things. It does have an impact, right? There are good strategies around this. We’ve covered power ChatGPT principles. We’ve covered my mega prompting framework. It’s not like I think that prompting doesn’t matter. But in this world of like real estate and leveraging AI to get consistent results, the thing that people are most often missing is if it’s not giving you back something consistent, you probably aren’t giving it clarity on exactly what you’re looking for and exactly what it needs to look like. And how you solve that is by giving it templates and examples, right? So that’s typically how you will most readily solve that. So when I say that, what do I mean? Well, for me, templating it’s about you looking at what is the perfect example of something when it’s done [00:03:00] right,

[00:03:00] Zach Hammer: what does that look like? You take that and you start extracting out the key ideas. If you’re templating an email, what’s going on in the subject line? What’s going on in the first line of the email? What happens after the first line, right? So you build it out more like it’s a building block of concepts rather than just expecting it to know what you mean when you say write the perfect email, right?

[00:03:20] Charlie Madison: Right.

[00:03:21] Zach Hammer: What is that? There’s a million ways that you could write an email that would be good and ideal. There’s lots of different thinking and reasonings behind it. It doesn’t know what the perfect email it is.

[00:03:32] Zach Hammer: When you ask it to write the perfect email, it’s going to give you back some amalgamation of everything that it’s ever heard when somebody says perfect email. It’s most likely going to be hot garbage that it gives you back that maybe has some hints of intelligence, right? But when instead you start saying, here’s an email style that has worked well for me or is proven or is tested. I want to take this thing that I’ve already tested and proven, and I want to reuse it over [00:04:00] and over again.

[00:04:00] Zach Hammer: And I want it to follow that. That’s where you build templates. Does that concept make sense?

[00:04:06] Charlie Madison: Yeah, I like that idea because I feel like a lot of times when I’m using ChatGPT, it won’t duplicate itself and it feels like my results are more like a shotgun. Like it’s not even that it won’t duplicate, but there’s a wide array of results and what I hear is, if I’ve got this template, these borders, it keeps it within the range of what I’m looking for.

[00:04:29] Zach Hammer: Exactly. So, like most often if you’re asking for something and it gives you back what you want, sometimes it gives you back something that’s off. We talk about this often, is essentially a mirror for your own clarity. So if it’s not consistently giving you back what you want, you aren’t being clear.

[00:04:47] Zach Hammer: And the way that you increase that clarity is you give more specifics, right? Even a professional builder, an architect, and say, build me the perfect house.

[00:04:56] Zach Hammer: You’re an amazing architect. You know what the perfect house is. There isn’t a [00:05:00] perfect house.

[00:05:00] Zach Hammer: There’s the house that’s perfect for you. How are you going to use it? How do you need it? What are houses that you like? What are the other houses that you’ve seen that you liked? What do you like about them? What don’t you like about them? Those sorts of ideas. Even when you’re dealing with other humans. Build that clarity, you have to provide clear examples.

[00:05:15] Zach Hammer: Now, if you approach that same builder and you say, hey, here are some examples of what I like. Here’s what I like about them. I believe my ideal house would have these elements. They’re a lot more likely to be able to come back to you creating something that’s really close to your vision that you lay out, and essentially AI works the same way.

[00:05:36] Zach Hammer: What I find most often, this is the disconnect that people seem to miss that is really powerful if you can build it as a skill set. And it’s the ability to abstract the concepts behind what makes something successful. So here’s what I mean by that. When you look at a social media post YouTube video whatever that you like, instead of [00:06:00] watching it, reading it, or reviewing it from the vein of just consuming.

[00:06:04] Zach Hammer: Describe, what are they doing here? What are the words that you would use to describe, like, What is that opening sentence? What is the next thing, right? If somebody goes through and says, they open with a hook that’s captivating and uses a question. Then, they introduce themselves. Then, they introduce the key concepts of the show. Then, they go down point by point through the outline, following these steps, these ideas, and conveying this sort of thing, right? Like, when you go through and do that, and you actually write that out and you build that out as a template, for one. That’s one of your best ways to actually understand why something’s working or why something’s not, and potentially learning more about what makes a difference or what doesn’t. Which helps you to be able to get more better results of the future when it comes back to you with something that’s off, you know, why, where, and how.

[00:06:48] Zach Hammer: Right? But the other reason is it’s what it needs in order to give you back those ideas too. Does that make sense? Like I feel like this is one of those concepts that it makes sense in my mind, but I don’t know [00:07:00] if it connects with other people in terms of like, how you would actually do that.

[00:07:03] Zach Hammer: Do you feel like you would know how to go out and do that process? Or does that feel like that would be something that, that would still be hard?

[00:07:10] Charlie Madison: Yeah I am not good at breaking things down into their individual parts by myself. I would have to try to use ChatGPT or someone else to break it down for me. Once I get so many parts, I lose track of it.

[00:07:25] Zach Hammer: That makes sense. And I found the same thing. And the good news is you can use AI for this. I’m in the process of developing some GPTs and some prompts that are designed to help do this as well. This is both unfortunately and fortunately. The good news of the bad news, right?

[00:07:41] Zach Hammer: So the good news is you can have AI help you with this. The bad news is it won’t be able to do all of the things for you if you don’t understand the concepts, right? If you don’t have the words to describe it. The good news of that is that means that you still have a lot of value with [00:08:00] your humanity and your expertise and your unique point of view.

[00:08:03] Zach Hammer: The bad news is, that means that AI won’t be able to do it for you and make that thing completely easy for you, right? So you still have to build up your expertise and understanding of the underpinnings, the underworkings of what’s making this thing work or not, and try to put words to it.

[00:08:19] Zach Hammer: Here’s part of the permission that I want to give you and anybody listening. It’s not going to be perfect when you start, right? You’re going to try and you’re going to try to describe something and you’re probably not going to have exactly the right words, you’re going to try and build a template around it, and it’s going to work maybe closer than without the template, but it’s not going to be perfect. But over time of trying that process and seeing how it plays out, seeing where it goes wrong. You’re going to start to build that internal intuitive understanding of what you’re looking for, what you’re not, and how you could get better at clarifying those concepts. So, what I found so far is that AI is pretty good at giving you a rough template if you know how to [00:09:00] build out a prompt that’s designed to get it to do that abstract thinking in the right way.

[00:09:04] Charlie Madison: Right.

[00:09:05] Zach Hammer: But you’re still gonna have to, Go ahead.

[00:09:07] Charlie Madison: Isn’t that what we do? I’m thinking of kids learning to walk. We try something, it breaks, we try to improve on it and we just can’t keep refining. That’s how I learn anything.

[00:09:19] Zach Hammer: I’m never completely sure if this might be one of those things that I have as just a super power. That I don’t recognize it as a super power because it feels normal to me. But I’ve learned that’s just necessary.

[00:09:30] Zach Hammer: You will learn more by looking at somebody who’s doing something successfully and breaking down what they’re doing, then you typically will by having them teach you what they’re doing. What’s even better is both, right? Where it’s like, you look at it, you try and break it down, try and understand it, and then you get their perspective and that helps you clarify it.

[00:09:50] Zach Hammer: That same process tends to work well with AI as well, where you can have AI try and break it down for you but you want to be able to look at the template and say, where is it wrong? [00:10:00] Why is it wrong? And assume that it is wrong on some things. And try and come up with those reasons, because that will give you the better understanding.

[00:10:06] Zach Hammer: So, the end result though of your template is essentially, there’s a couple of key things that I like to do. So, I like to have specific areas in my templates that are designed to be replaced conceptually. What I do is I tend to use curly brackets to describe those.

[00:10:22] Zach Hammer: So it’s like I use a curly bracket and then within the text, I describe what is it going to replace. Okay? So what is the concept? An intriguing hook that’s typically starting with a question. I’ll have that in curly brackets, that knows that’s what it’s replacing.

[00:10:35] Zach Hammer: The other thing that I’ll have in a template is plain text that is not designed to be replaced. It’s designed to be the same every time. I’ll use that for things like if I’m building a job description.

[00:10:45] Charlie Madison: Yeah.

[00:10:46] Zach Hammer: Then it might be like the company name always stays the same, job description, job title. Then the actual job title is a thing that’s replaced, right? But the job title indicator would not [00:11:00] be. So it’s plain text and I include that in the template and I indicate this stays the same. One of the other things that I often do is that I’ll include comments in the template.

[00:11:08] Zach Hammer: Sometimes where I’m giving further context to like a bigger section where here’s the goal of this section. Here’s what we’re looking to do. Part of what matters is the formatting of how you do that. So how do you build out those templates? And so for me, I actually, in the templates that I’m building, I include a legend that explains, here’s the replacement text.

[00:11:26] Zach Hammer: When you see plain text, it stays the same. Don’t change it. Then if you see these symbols, that means that it’s a comment. Don’t include it in the end result, but use it as guidance for how you build the end result. Does that make sense?

[00:11:37] Charlie Madison: Yeah.

[00:11:38] Zach Hammer: One of the other key things at this point. I essentially consider it part of a template, even though it’s technically not, is the examples, right?

[00:11:46] Zach Hammer: So it’s really useful to be able to give it both the template, but then an example of how that template would look when it’s filled out. And again, this goes back to the same concept that we talked about last week. I don’t remember when we talked about this. But that [00:12:00] typically if you treat AI almost like how humans learn.

[00:12:03] Zach Hammer: Then you’ll get better results. So in the same way, if you just give somebody a framework for how to write an email or how to write a blog post or that sort of thing, that will help and it’ll be better than nothing. But if you really want to be clear, you give them an example or two of somebody filling out the framework so that they could see it in application a few ways around.

[00:12:22] Zach Hammer: Does that make sense for how I think through building out templates and where and why and all of that?

[00:12:26] Charlie Madison: Yeah. I really like the idea of viewing them as guardrails so that I get some kind of consistent output. Because otherwise, it’s got its temperature setting. The reason it gives different responses every time is it’s literally built to be creative, to be different.

[00:12:43] Zach Hammer: Exactly. That’s the key. It’s both its power and sometimes its detriment if that’s not what you’re looking for. So, templates are one of the easiest ways to put those guardrails on it consistently, get it to still have the right amount of creativity but fall in line where it needs to follow your vision.

[00:12:58] Zach Hammer: With that in mind, [00:13:00] where are templates useful and where are they not useful? They are useful when you have clarity on what you want it to do consistently. So if you say, yes, this is what I wanted to do, I wanted to do it consistently. Chances are you want templates, you want something that says, here’s how to do it.

[00:13:17] Zach Hammer: Here’s how to do it right. Sometimes you need multiple templates. So when it comes to a social media post. You probably don’t have one template. That’s your right answer. Likely multiple ways that you’re going to structure text in order to convey something. You might have a different template for announcing an upcoming open house, versus how you want to talk about a client story.

[00:13:36] Zach Hammer: So it’s worthwhile to have different templates for different things and to convey different ideas or to strategically convey different concepts. That’s probably going to be the case for blog posts, for emails, for social media things like SOPs. There might be different types of things where you need multiple templates for kind of a bigger bucket of things.

[00:13:53] Zach Hammer: And that could be really useful to have those. So when are templates good? When are they bad? Sometimes you don’t want it to be [00:14:00] consistent. Sometimes you want to leverage the power of AI’s creativity to give you different ideas, different ways of saying something, different ways of conveying it and, so if that’s the case, if you’re essentially using AI as a brainstorming partner, where part of what you’re looking for is you’re looking for variety, you’re looking for different ways of doing it, where you’re saying, I want you to try writing this as a bulleted list.

[00:14:23] Zach Hammer: Now I want you to write it as if it’s just being said with emojis. Now I want you to write it as if it was a paragraph. And you’re just trying to creatively figure out, what is the best way for me to convey this idea? What is the best way to describe this? Then you don’t need templates for that.

[00:14:36] Zach Hammer: You don’t want templates for that. You bring templates when you know what you want, and you want to achieve consistency. Does that make sense?

[00:14:43] Charlie Madison: Yep.

[00:14:43] Zach Hammer: The other aspect that comes into play, is having the rest of your things in your prompt dialed it, right? That’s where the mega prompting framework comes into play.

[00:14:50] Zach Hammer: So at this point, what I typically think is if I’m building out a mega prompt, what I’m essentially saying is that I want to achieve something consistently, if it’s an end [00:15:00] asset that I know what I want it to look like, then I’m likely bringing templates and examples. to the table as well.

[00:15:05] Zach Hammer: So I’ve got my mega prompt that includes all the context that we’ve talked about with the mega prompt framework. I’ve got my mega prompt plus templates and examples. Then from that, I get consistent end results. So that’s why to me, most of the time when people are really like fiddling with their prompt, They probably need to be putting more time into building a template to build that clarity.

[00:15:25] Zach Hammer: Because it’s a lot easier to convey to AI and say, I want you to write an email in this way. It’s a lot easier to be really clear by showing it the example of what you mean, rather than trying to find the exact right word that means that to it. The phrase a picture is worth a thousand words. There’s some rhyming there to a template is worth a thousand words as well. When you could show, this is the picture of what I want you to do, it understands that much clearer, much better than you trying to find the descriptive words that a template would do a better job, [00:16:00] enacting.

[00:16:00] Zach Hammer: Does that make sense?

[00:16:01] Charlie Madison: That makes a lot of sense. Yeah, I almost view it too like the template means, here’s the things that I want. It’s almost like a mold. Here’s the things I want the same or I even had this image of putting up walls so that water flows in the way I want. Here’s what I want the same, but I want you to be creative.

[00:16:20] Charlie Madison: Take all the mold, but be creative in this way.

[00:16:24] Zach Hammer: Exactly, and that’s the power, right? So it’s templates in a way that we haven’t had previously. So it used to be that templates were almost like, all you could do is fill in the blank with an exact word that’s maybe.

[00:16:36] Charlie Madison: It was like a mad lib.

[00:16:37] Zach Hammer: Yeah exactly. Now it’s able to replace your templated concepts with creativity, with being able to say, okay, so we’re writing a social post about this thing and we want to start with this idea.

[00:16:48] Zach Hammer: I could take the context of what we’re talking about, take your template and rewrite it with those things in mind and be able to get a good consistent result while still having that right level of creativity to fill in the gaps, to [00:17:00] reword things, to make sure it’s clear, make sure it flows, make sure conceptually all the ideas are there.

[00:17:04] Zach Hammer: But we have forgotten the power of templates because of the power of the creativity and really, if you’re not getting the results that you’re looking for, it’s probably because the right answer is in between, right? It’s that blend of giving it the guardrails to know how to get as close to success for you as possible while still having the freedom to be able to be adaptable and leverage your context to rewrite things for that purpose effectively.

[00:17:30] Zach Hammer: Does that make sense?

[00:17:32] Charlie Madison: That makes a lot of sense.

[00:17:33] Zach Hammer: Awesome. There you go. That’s what I wanted to cover today. So hopefully that is helpful for folks. If you like the idea of being able to know how to build templates, get help building templates, be able to do that in a way that’s relevant and directly applied to the real estate business.

[00:17:50] Zach Hammer: This is part of the work that we do in the AI Mastermind for Real Estate teams. Where we’re looking at concepts and figuring out what are the templates that make sense? How can we adapt templates to this [00:18:00] purpose that we get consistent results for writing SOPs for we haven’t done this yet, but I imagine it’d be useful for filling out contracts, adapting contracts, making offers, thinking about how to structure offers. All of these kinds of things, all these sorts of ideas maybe I know for instance, you have a specific way, a specific framework for how you think through writing like a listing description, right? So being able to take that concept, templatize it so that when you get a new listing, you can make that process of writing that listing description flow through what you’ve proven to work, right?

[00:18:30] Zach Hammer: Maybe somebody else has a slightly different way of doing it. Honestly, when it comes to something somewhat creative like a listing description, I don’t believe that there’s an exact right or a wrong way. I feel like there’s the way that makes sense to you that’s likely built on powerful principles.

[00:18:43] Zach Hammer: But there are going to be multiple ways to still get to a good end result, and it could be useful to be able to get ChatGPT to do that result that you like consistently, right? So if you like that and you want help doing those sorts of things, that’s what we cover in the AI Mastermind for Real Estate teams.

[00:18:59] Zach Hammer: If you [00:19:00] want to learn more about that, get involved and see if we’re a good fit for you and see If you are a good fit for us. Then reach out to us at RealEstateGrowthHackers.com slash contact. We’ll get you more information about that mastermind. But otherwise, hopefully this was helpful for folks on learning how to get better results out of AI and where templates might make sense and why a prompt might not be the right answer.

[00:19:19] Charlie Madison: And I love it.

[00:19:20] Zach Hammer: Awesome. Until next time, folks, we’ll catch you on the next one.