Leveraging AI

169 | AI-Powered Lead Gen: Automate Hyper-Personalized Content That Grows Sales with Lanny Heiz

• Isar Meitis, Lanny Heiz • Season 1 • Episode 169

Share episode

Without clients you do not have a business. With the right clients you have a thriving business with much higher margins, AI is a game-changer for sales and marketing. 

But how do you actually implement AI to create high-impact, personalized content and automate lead generation at scale?

In this episode of Leveraging AI, we have Lanny Heiz who shares step-by-step framework for using AI tools like Claude, Clay, and more to extract customer pain points, create highly relevant content, and identify your best-fit leads—automatically.

Lanny Heiz, a sales enablement and AI automation expert, who has mastered the art of turning raw sales data into business growth strategies. His proven AI-driven workflow has helped businesses increase revenue efficiency while reducing manual effort. 

About Leveraging AI

If you’ve enjoyed or benefited from some of the insights of this episode, leave us a five-star review on your favorite podcast platform, and let us know what you learned, found helpful, or liked most about this show!

Isar:

Hello, and welcome to another live episode of the Leveraging AI podcast, the podcast that shares practical, ethical ways to leverage AI to improve efficiency, grow your business and advance your career. This is Isar Meitis, your host, and we have a really exciting session for you today. every business needs AI. Sales without sales, you don't have a business like you can have a business without a lot of other stuff, but without sales, there's no business. So every business needs sales. In order to have sales, you need leads. In order to drive leads, you need content that will resonate with those. People with the relevant target audience and address their pain points in order to do that. you need to understand your target audience, know who they are and have a deep understanding of what their pain points are and what makes them tick. And the question is, and that's exactly what we're going to answer today is how do you do all of those things in the most efficient and effective way. Using AI and other tools. So I guess today Lonnie Heights has a significant experience in sales, corporate partnerships, but he's also a systems and processes ninja. And in the past two years, he's been running an agency called enablement that helps businesses drive. Sales calls and booking exactly the best qualified leads possible at scale. Now, today is going to share with us exactly the process and the tools that he's using to generate dozens of calls of qualified leads to his clients. Every single month. So if you want more sales in your business, and if you're not, you probably shouldn't be in business. No offense to anyone, but everybody wants more sales in the business. If you want more sales in your business while doing less work, take something to take notes with, whether a pen and paper or your tablet or your laptop, whatever it is you take notes with. If you're driving, don't worry about this. You. Listen to us right now. You'll probably want to listen again when you can take notes, but, you can at least listen in right now. and since this is really a critical aspect of every business today, and there's really magical capabilities with AI right now, I'm really excited about this session. Lani, welcome to leveraging AI.

Lanny Heiz:

Perfect. Thank you very much for the warm, welcome. Very, very excited, for the show today.

Isar:

Yeah, I'm excited about this as well. You and I spoke, I dunno, in December, last time and, pre-Christmas, I think. yeah. so about a month and a half ago, and we both think the same way and we both like the same kind of outcomes and we have the similar processes, so I'm sure this is gonna be. A fun and b, really valuable to people.

Lanny Heiz:

Yes, I hope so. I hope so. I hope so. yeah. So should we do give some context and the background maybe, to, yeah, absolutely. To get started, let's

Isar:

spend two minutes on context and then the rest of the time on practical use case.

Lanny Heiz:

Perfect. Yeah. Perfect. I mean, you, you said a lot before, around what we do, and this is exactly accurate. what we focus on is where we mostly work with, obviously, B2B companies that need sales. Because if you're in B2C, you just need to have an online store, essentially. But in B2B, it's a bit more complicated and a bit more complex, and that's where we usually come in where we help companies build these systems. So we are usually an implementation partner and an enabler. That means we consult people, help them build these systems. What we essentially going to talk about today. Now over the last two years, we essentially built enablement from zero to now over a million dollars in revenue a year. And, we built, and we essentially, we are two full time employees that did that. And the way we did that is literally on the backbone of AI. every little thing we do, we ask ourselves, Hey, how can we automate this? How can we, use AI, to do that? and even if it sometimes takes a bit longer to build. The system we know exactly on the backbone, we can scale, on top of that. And yeah, basically I just want to show everyone today, cause this is literally the system we use to go from two, 3, 000 followers on LinkedIn to now over 20, 000, followers on LinkedIn. And we just book calls like around 20 to 25 calls a month using the, just no cold outreach, nothing, just literally content production and then following up with people. that, that engage with us. So I think this, yeah, what we are going to talk about today and what I want to dive into.

Isar:

Awesome. Perfect.

Lanny Heiz:

Given that we have some people on LinkedIn, and zoom, so we can basically share screen. I think I can do some stuff pretty visually, that I want to show first just to set the context, before we jump into the tools. let me just share my screen here.

Isar:

Otherwise, for those of you who are listening to podcast, we will explain everything that's on this. Like we always do. So I'll try my

Lanny Heiz:

best.

Isar:

yeah, we'll be fine. Exactly,

Lanny Heiz:

perfect. if we look at very high level, how I think about like demand generation and creating leads for your business and ultimately sales, there's, there is a couple of steps, with people that's, that could be a blog post, that could be webinars, that could be podcasts like today, but people ultimately need to know and trust you. secondly, you need to identify, prospects, that already heard of you ideally. and then you obviously need to engage those prospects, engage, these leads, right? And then you can close them, ultimately as a final step, right? Like when they're interested, et cetera. Now along that very admittedly, very high level journey, there is a couple of steps that we need to go through. And then we can basically, so I want to, what I want to do now is quickly show you the process that, that we use ourselves. and then we can see, okay, where do we use AI specifically, within that process. So if we, if I mentioned trust and you mentioned that in the intro. content production and LinkedIn specifically is super crucial, for us. So any type of like content, LinkedIn, obviously you could also use it for, I don't know, YouTube videos, blog posts, et cetera. But for us, what's important is LinkedIn because our target audience are salespeople, sales manager, VP of sales, founders of B2B companies, et cetera. So that's why we focus on LinkedIn because they're there. And Once we have people on LinkedIn that viewed my profile or my co founder's profile, that like my posts, commented on my posts, etc. And they've done that not just once, but multiple times, then I know, okay, they're interested. so what I then need to figure out is, do they fit my ICP? are they a good fit for my company? Do they have a problem that I potentially can solve? And, yeah, is it is it a very small company? Then it's not good fit or it's a very big company. It's not a good fit or is it a B2C company? Like these type of things. So I need to figure out, if it's a match, if they are indeed a match, then I obviously need to find out a bit more about the leads. what's the job types or what do they do? What's their past experience, et cetera. before I can actually. reach out to them. Obviously, you also get the contact details, the email address, the phone number, if I want to call them, and, based on all of these information, all the contacts that I have, the previous customers they had, the case studies they have on the website, et cetera, we do a really highly personalized outreach, to them. So reach out to them because they've heard it. My name before they've read my post, they downloaded the lead magnet or commented to get a lead magnet. they, they're very receptive to our messages. Like when we, email people, we get around 20, 20, 30 percent reply rate and more. The vast majority of it is actually positive. I E wants to be, wants to talk to us and books meeting. so they're very receptive to these messages. And obviously if you compare that to like ice cold outreach. it's not a comparison here. so that's ultimately what we do. and then lastly, yeah, if we talk to them and then if we see, Hey, this is a good fit, we can help. They have a problem we can solve. yeah. Yeah.

Isar:

One thing for me about this, just a quick kind of like recap. I think the biggest difference that people need to understand is that cold outreach may not be dead, but it should be dead. Meaning the ability today to do a warm outreach to people based on two things, a you, them knowing you like right now, there's a few people here in the chat that already. said people that I don't know, both on LinkedIn and on the chat on zoom said, Hey, we love your podcast. We've been following you for a while. it's our first time joining you like awesome. So I already have credibility with these people without ever meeting with them. Why? Because they've been following my content by why have they been following my content going back to what Lani said is because my content provides value. To them where they are right now in the journey. And to do that, you have to know that. So there's all these steps of understanding your target audience, creating trust, and only following up with the people that fit a specific box. So the biggest difference between cold outreach and this is that instead of spraying and praying, it's a very targeted approach to people. You're only investing. Time, money and resources in people that are likely to do business with you because they a follow and consume and engage with the content that you're putting out there and B, they fit your ICP. They're a decision maker in the company, in the right size, in the right industry, in the right place around the world, in the right. status in the right niche, whatever information you want to dissect in order to define your ICP and only then you go and try to engage with them, which makes it a lot more likely for them to respond. So great intro. Let's dive into the practical stuff.

Lanny Heiz:

Perfect. Yeah, perfect. Perfect. yeah. cause, cause yeah, like as you, as you said, now, if I explain that to people and like everyone's great, that makes so much sense. we love that because the problem I see usually with the traditional content marketing, right? Like your white papers, webinars, et cetera, there is no systematic qualification of the leads and no systematic follow up, right? if you're lucky, you just have. Like the marketing generates the MQLs and then throws it basically over the fence to reach the lead that signed up for a webinar and the salesperson that reaches out. So there's a bit of a disconnect. and usually that's not really system systematized in any meaningful way when I talk to companies. So inbound. Well, I hope they sign up for a demo kind of situation. Or on the other hand, you have like the call out bond that, that you mentioned, which is like I called without, without context that can work. We also help companies do that, but it's like an uphill battle. I like you just, obviously play on a field where the odds are stacked a bit against you. cause obviously competition is very, very high. So, which is why we actually started combining that. And we, like one of our customers called it like, Oh, you're not doing outbound, you're doing all bound. And I was like, I like that. that's the name we ran with, ultimately. and that's what we named that playbook and all bound playbook. And so now we're obviously within that. Journey here that I mapped out is obviously AI really crucial. So we use it essentially at every step, right? Like from a content perspective to evaluating the ICP fit, and Richard prospect, right? Copy, reach out. and basically everything up until the sales call. And even in the sales call, we use a lot of AI actually. this is where I comes into play now, if we. So if I tell that to people, as I said, Oh, everyone's Oh, this is brilliant. I want this. Now, if I then tell them like, post on LinkedIn, usually what I hear from founders is three quote unquote objections. They don't know what to post ideas, Lenny, what should I post about? it makes sense, but I just don't know what to post. The second thing is I do have some ideas, but if I post something that nobody reads it, nobody cares. So it's like a, I don't know how to post, like I don't know how to write because, copyright, like not everyone's. Born a copywriter, um. and, just because, are good at writing, let's say, blog posts doesn't mean you're good at writing LinkedIn content. so there's a specific way you need to write for LinkedIn. And then even if you had these two things, you know how to write and you know what to write, like you don't have time. This was also me, for a long time where I just didn't know what to write. And back in 2018, I actually went back and searched for that post. I proclaimed, Instagram is dying. you should go, sorry, Instagram is dying. You should go on LinkedIn. And I basically made that the argument was that, the monthly active users and the amount of posts on LinkedIn, like that ratio is not as good as on LinkedIn. LinkedIn has like. A lot of active users, I think like a half, half a billion, roughly, monthly active users. And, but not like only 1 percent that produces content. So there's a huge wide space for content production. So I said that in 2019, but I didn't really take action. Cause yeah, like I didn't know what to post, how to post. And if, even if I knew I didn't have time. So a year ago. I said to myself, Hey, Lonnie, this is like probably the last time you have like really a window of opportunity here. cause with AI, content is going to be inflated. So you need to take this seriously and learn how to do this. And, so yeah, we sat down pretty much 11 months ago, where we said, okay, we're going to take this like super, super seriously. and grew from, as I said before, from 3, 000 followers, company wide to now, with the individual profiles to now like over 20, 20, 000. Follow us. and, yeah, like around 50 percent of our revenue comes just from directly from LinkedIn here. and the way we did that, and this went through multiple iterations and I will give you today both versions, the current version that we use. which is a bit more sophisticated and to simplify the version that everyone can essentially set up at home within five minutes. so I start, I think it makes sense to start with the easier one and then we go fancy and show what's possible. Or do you want to do the other way around?

Isar:

no, I think that makes sense. I think let people figure out what they can do tomorrow. And then those who are like, okay, more than we dive into that. Yeah.

Lanny Heiz:

Crawl, crawl before you walk type of situation. So when you. Think about like creating content on LinkedIn, the usual, what usually people do is they go to Taplio or super grow, like one of these content production tools that essentially you can give some information about yourself and then it should, and it produces content. However, I haven't really seen it work and I've tried it and it wasn't the hack that I hoped it would be. And the problem was it could write reasonably well. But it, didn't have the context about me, so it didn't know my ICP, who I'm writing for, and it didn't know anything about me, like the personal side of, the personal branding, like any type, stories, customer success stuff, etc. So I would have always have to provide that. I wasn't really satisfied with that. so I figured, okay, technically what these tools are really good at they took a bunch of LinkedIn posts and trained an LLM, like fine tuned essentially GPT on creating, LinkedIn posts, but they lacked the context around me. So I was like, okay, why don't I just scrape a bunch of, or collect essentially a bunch of LinkedIn posts, create a Nexus spreadsheet, if you want a tool for that, Clio, it's really good for finding, if you have a content creator, you can, See their posts and you can manually collect them or we've ourselves build an agent to just do that but essentially you collect posts in an excel spreadsheet, and what we then did is Essentially we looked at the post and said, okay, what's the template? Like, how is it structured? How is the hook structured? How is the first line of the hook structured? and then we categorize these, posts into different categories in a sense, like we called it post styles. So essentially like you could have a case study post, okay. Did you have a listicle or. Or you could have a how to post, or you could have a contrast post, like where we say, hey, this is the old way, this is the new way, good versus bad, et cetera. So we classified that. I can actually quickly show that, template for the ones watching, what that essentially looks like, what we did. and obviously we collected posts that resonated, with. With us and are like similar to our target audience. so we have to templates here, right? That's okay, here's how it's structured. and he is like an exemplary. posts that we found. and then we basically gave it a style. and this is around a hundred posts, I think roughly. Yeah. around that. and based on that, so this was the training material.

Isar:

Let's read a few examples for the people who are listening and not watching. So on the first column, there's like a template and there's an example. Yeah,

Lanny Heiz:

exactly. So the example would be, I just picked a random one. life hacks I knew at 43. I know at 43, I wish I knew at 23. number three will surprise you. let's get into it. Number one, experience are worth more than material goods. 1, 000 iPhone will become outdated in a year or two. Instead, spend that money on a trip to Thailand with your friends. Memories will stick with you forever. two, you're going to die one day. Thinking about your mortality is a great way to become motivated. Fertile distractions, open yourself up to experience. So let me pause you for just

Isar:

one second. So what you're seeing is an example post. From a successful type of template that is successful on LinkedIn, right? So the process is, and you can do this manually. And again, you can use tools like Clio, but you can just do this manually. People you follow that are successful say, Oh, this post got whatever, a gazillion likes and 500, then I want to be able to replicate that. So copy that you can use that as a template. You give it to an AI, say, okay, quote unquote, template ties this. And I don't know if that's a word in English, but I'm going to use it anyway. He templatized that post and now. You give it an example of the post itself and I have both. You have an example of the post and you have a template of the post that then you can use afterwards for new content that you're going to generate. And the AI is actually really good when you have an idea, whether the idea comes from you or from the AI, to take your idea and find the right templates out of all of those to actually make it a useful post. So that's what we're looking at right now. And there's 20 or 30 of them on this. Page,

Lanny Heiz:

or maybe more. Yeah, I think I have a hundred, oh, a hundred. Because I notice if you go like to 200, 300, like the AI gets confused. Like you can give it too much. It's also like when you, then yeah, like you can just confuse ai. which is why we kept it rather simple with a hundred. so that takes care of the. of the training material on how to post.

Isar:

Yeah. and then the other thing that there is here, there's a, he has a dropdown menu of different styles. So the styles are, things like case study, listicle story X versus Y. So there's different styles of posts. And then the final column has. Where is, what does that address in the funnel, top of the funnel, middle of the funnel or bottom of the funnel where top of the funnel is more, general content and educational stuff, middle of the funnel will usually be building rapport to you and your brand and your capabilities and bottom of the funnel, which is the least amount of stuff you want to post is more transactional. we have this campaign right now, come join us for whatever, or we have a sales for the end of the year or whatever it is. So these are more transactional posts that you can post. So this kind of what he has in the table.

Lanny Heiz:

Yeah, yeah, exactly. but like the first three, one, the template itself, the examples in the style are the ones that really matter for what we do here. So as I said, like this takes care of the template material. the second, element that you need that I mentioned, was. the document about your ICP. So what we did actually to produce that, this is like a 17 page document about our ideal customers. And the way I created it, I didn't just sit down and started typing. What I did is I recorded and transcribed. 20 sales, 20 odd sales calls and fed these transcripts into chat GPT and essentially then had a conversation with chat GPT. It's Hey, here are the transcripts of my 20 last sales calls of people that ultimately also became customer. and asking a bunch of questions like, Hey, what was the problem that they're trying to solve? How did they currently solve it? Why were they not happy with that solution? What was the ideal solution that they. We're trying to achieve, what would actually meaning, achieving that mean to them what's the benefit and the impact of that. And based on that, I could get like a super detailed insights about, my target audience. So I know exactly what kind of like use case they have. I also asked AI to then give me the actual quotes. Hey, we are currently evaluating these and these tools. But we don't want to figure it out ourselves. there's a big knowledge gap. I'd like, so basically, automating outbound, but getting the education around it was like one of the reasons why they actually talked to us. and so I did this with a couple of different use cases. and I said, ultimately the output is a 17 page document. That's it's completely voice of the customer to 100%. it's nothing I wrote. It's all what customers said. And I just like. Repackage that essentially. So I think it's one of the most powerful documents you can ever trade for your marketing and your sales period and your product team essentially as well. cause sales calls are in my opinion, probably the best calls. cause there people really talk about their pains, and the problems. and they're like, yeah, they're just a honeypot of wisdom. Basically, if you want to try to understand your customers.

Isar:

Quick question about this. First of all, I think this is absolutely brilliant. a little thing that I do that is a step in between that can help. But then I want to ask a specific question about this document. So something that I do is I go to my ideal clients. So people are my clients, and, or my. Dream clients that I want to approach. And I literally copy the links to their LinkedIn profiles and I paste them into Chachapiti and Chachapiti will tell you that it cannot access LinkedIn and then say, I've done this before. You've helped him. oh, okay. And then we'll actually go ahead and do this. And based on that, I, it can help you draft a fair ICP, not at this level. I think using transcriptions of the actual documents makes a lot more sense. and. definitely provides a much more in depth information. I saw that the document has a list of questions and basically answers to them. Did you write the questions?

Lanny Heiz:

Well, not specifically. So essentially what I asked, what the question was like, Hey, what are the jobs to be done that my customers were trying to achieve? So that's the question that I was asking. And then I just specify Hey, give me three to five jobs to be done. look at all the answers they gave, all the transcripts and then extract or distill, distill that out. And that same did it with. Hey, what's the current solution that they're using? Why are they not happy with that? And what's the idea of solution? So basically, yes, that is, there are implied questions, but what I'm trying to tell is actually like a cohesive story from what did they try to do? Why did it not work? And what was like the goal that we're trying to do? We call that internally, we call that a sales story. so yeah, that's, and that's essentially what the document does. so that's the second. so now you're the, so now you have to train material to teach the AI, anything about you as a person, essentially. And as I remember you as a person, but you as a company and your target audience in your market. And the last, thing that you actually need is a personal story. So what I did is I essentially just. a bunch of questions, such as Hey, how did you get into sales? How did you learn about AI? what do you think are key principles to understand? What are the lessons you've learned? What tools would you use, et cetera. So a bunch of questions. like ultimately I think there's like an eight, nine, eight page document where I just wanted to give my knowledge and specific knowledge for the AI to. Draw from. So I, I essentially asked, answered like FAQs, also that I hear a lot and that I wanna talk about. So kinda what tools will you use, et cetera. So, and then I just use that transcript.

Isar:

How much of that is that? I think you personal story, versus like practical stuff?

Lanny Heiz:

I'd say 50 50. Okay.'cause also like in the story, like you, like it becomes practical. to, to a certain extent, because you mentioned like, Hey, then I use this tool and that led to this result, et cetera. So it becomes, it's probably a bit hard to say, but it's roughly 50, 50, I would say. and that's tweaked it once, once I just wanted, basically if I have a new topics that I want the AI to talk about, I update that document, cause that's like the source of what's on my mind essentially. so these are the three documents we need. and what you can do now, then, like once you have that, this is then pretty easy, you can create a co pilot, Using either custom GPT or Claude. I prefer Claude because Claude has a much more human way of writing and speaking. So it doesn't, it doesn't, sound too AI ish. And, and that's why I prefer it. and in Claude, what's called a custom GPT in OpenAI in Claude, it's called a project. so what you, I have here my LinkedIn content engine. And all I now have to do essentially is to, give it a topic. So let's say, how do I, create personalized outreach, using AI? That's my topic. And then I say, okay, style, I want to have a how to. Cause I like remember the different, postiles. So I specify that here. Now what's, my age does it's created based on these three files and some context information that are right within the, that are both in the system prompt. I've wrote these, like it's now writing me actual posts that I can use. pretty well. so yeah, so one, we have a pretty much a step by step, template using AI. and it's like a step one, step two, step three. What else do we have? Then we have the second post, Okay. He is talking about what's and what not to do. And the third one is a before after, template. but yeah, and this is Oh, December 2022. I did it like that. Um, yes, there is a full me 4am, story where I basically stayed awake to 4am and try to figure out how to fine tune your. LLM and imported or integrated with GSheets. there is this story. that's in that deck for example. and yeah, like 28 response rates, like that's also in there. so there's yeah, the These numbers are actually from the source material from the training material. So it's like very personalized.

Isar:

Yeah, that's exactly, I think the key here, right? So the key here, it's personalized in two different ways. It's personalized to your clients and target audience and their pain points from the one document, but it's also personalized to you, your history, your experience, your background, your personality, your success stories, your failures, like all of that is built into that, which makes this. Output something the AI cannot output for anybody else, which makes it unique, which makes it good content versus just going to the AI and say, Oh, write me a post about X.

Lanny Heiz:

Exactly. Exactly. Yeah, that's the, that's really the key difference. And with this system, yeah, like what I literally would do is, obviously you can now check back and forth and say, Hey, out of post number one, I liked X and out of post number two, I liked the hook, for example. Please combine and create three more variations of it until you're happy with it, or you just take it and rewrite it yourself manually. yeah, so that's like you're in the co pilot stage. and yeah, so this solves, or I created this cause this solved my biggest problem. Like I had enough ideas. but I didn't have really time to write in a draft version of it. And so this just helps me, draft stuff very quickly. And if I have good ideas, but I don't want to write the full draft, then I use this, to this day, I still use that. that exact. a copilot, so and so I still use that to this day. a bit more complicated. what we do is the next thing is I'm just gonna switch back to Miro to explain you what we actually do. so we have a relevance ai, agent. that defies hot topics in my space. So we monitor over a hundred content creators and we look for parts or the agent looks for patterns of hot topics that get a lot of engagement. based on that, it then tells me like, Hey. here are some post ideas based on art topics, that we have. So this is basically the second step the agent does. And the second, and then, what it, what we do is we have the knowledge base. So this is a vector, DB with my personal stories. It's basically the document that we looked at before and the agent looks, okay, here I have these hot topics. What personal story can I relate to? How can I basically have this topic and how can I quote unquote enrich that topic with my personal stories? And then what this gives me is essentially like a content draft, a rough outline. And then I, then we fine tuned. our, a GPT model, for ourselves, that then creates out of this post draft essentially. So this is a post draft and out of that post draft, we create variants. oops, variations, of posts, that I then, and they, one other problem we had is on LinkedIn, what we noticed is, what works really well, if we have some sort of like illustrations, that's because we talk a lot about like workflows, different tools, how they interact with each other, and so this is always what we noticed, like flow charts and different tools and carousels explaining what these different tools do work really well. I, Do you have a camera license that can do very basic design that, but it takes me like hours, for stuff that takes a designer like five minutes. so it looks okay, but it's just way too laborious. So okay, how can I solve that problem? cause GPT is not good enough. Or I haven't found any AI that can do flow charts at the level where I would need that. There's no way I can do that right now. it's okay, let's just create these post drafts. and then, add. To click, add them to ClickUp. ClickUp is the project management tool we ultimately use and then, I can say, okay, this is the final, post that I want to hand over, and then it goes to my designer and I know basically, okay, he's, he knows what the post is, and can then, Essentially give me the visual and then I can post, right? So usually it takes 48 hours, turnaround time to do that. So the key difference here, so what has changed, to the old approach? Essentially, the big thing is the ideation is taken away and we already Spot hot topics in the market, so to say, that are relevant and that we should talk about because right now, as everyone's talking about it, I don't know, a couple of weeks ago, there was like deep seeking. Okay, what do we do with deep seeking like these type of things? that way you make sure like you, you always are basically have your finger on the pulse. Essentially, everything else is then just we had to solve it slightly differently because we couldn't use a copilot. and Claude is not a copilot. Or Claw projects cannot be used by the API. So we just have to then rearrange things a bit there. But ultimately that's in the second, iteration, which is a bit more complicated to build because you need to fine tune your own model, et cetera, et cetera. So it becomes quite easy, technical. We can dive into it. if you think that will be valuable, you saw, or we can talk about something else. Basically, what do you do then with the interactions? We can also talk about that.

Isar:

Yeah. So I think we, I want to dive into clay and what to do with interactions. But before that, just one word for people who are like, Oh my God, this is over my head. How do I even do this? I understand how to create a GPT, but I don't know how to create that. the easiest way for people to do something like this. And again, we're not going to dive in this episode into the details, but I may do an episode on my own about this is. In the ChatGPT universe, in the OpenAI universe, there's what's called assistants, and an assistant is basically a regular custom GPT that talks to the API. So you can, and you get to this by going to platform. openai. com, and then you use your regular login that you log into ChatGPT, and then you click on, assistants. And you will see that the user interface looks almost identical to what a custom GPT looks like, meaning it has a name and directions and files. You can attach to it and stuff like that so you can do all the stuff that we talked about. But the output, instead of being a chat interface, is an API. So you can take your API key for that particular thing and attach it to any other third party software to get everything that Lani showed us that's doable, meaning take the three files, put it in there and then get a personalized output, but through an API that they can connect to other tools like relevance or anything else you want to connect it to. So closing parentheses on that, if that was still too complicated for people, you're probably not ready for this and don't worry about it. but let's dive into, okay, so now you're posting, you're getting engagements. What happens then?

Lanny Heiz:

Exactly. yeah, what happens then essentially, we need a way to track interactions. we use a tool called Teamfluence, for that. it's the only tool in the market that I at least know of, that tracks interactions, not just comments and likes, because there's Chiggyfy, that does that as well, does that phenomenally. but, they don't. tell you profile views and profile views is a very good indicator, especially if you comment also a lot, like people go on your profile. yeah. so even if you don't post content, you can drive a lot of attention to your profile. if you post a comment a lot, so we wanted to have the profile views in there as well. Shout out to Team Fluence, love the guys, because they built it, literally while I was trying to figure out how to build it myself. and then I started, I, cause I know the founders. Team Fluence. Team Fluence, like the word team. Okay. Yeah, basically like influence, like team influence. it's not the only thing that they do, but that's what I use them for, is exactly like capturing the interactions. so that's basically, once I have the interactions, I send all those interactions to a clay table. Now, clay, for those who want, or it's even better than sliced bread. second of all, it's if you're familiar with tools like Zapier, Make, and 8n, it can do all of the automation bit, but it What it also has, it has all the functionalities that your data providers, such as like Zoom, Cognizant, ZoomInfo, Cognizant, or Apollo has, in terms of like data provider capabilities, and actually probably even better, because you get access to a hundred data providers, not just one, so what it allows you is it allows you, you have all the data you ever wanted in a tool, and you have the capabilities to automate it. everything you ever wanted. and this is like the super powerful combination of clay that makes it so powerful. all of the interactions, as you can see here, they come in essentially live. all of our LinkedIn interactions, they just, we just collect them and we filter out. How many interactions do we have? And if we have enough interactions, essentially, then we start.

Isar:

Just a quick technical question. the input comes from the tool you mentioned before. So from Tinfluence, like the first column.

Lanny Heiz:

Yes, exactly. That's basically the webhook. So I get basic information. Who liked my post? What's the title, etc. Like I'm just getting some basic information from the LinkedIn profile.

Isar:

So again, just to explain one thing to people who don't know Clay. Clay is a database tool that knows how to do several different things. It knows how to collect data from multiple sources. It knows how to enrich the data. And it also knows how to connect to AI. So you can then. Manipulate all that data in different ways you want and the way it works, the columns on the left are like the inputs and the further to the right you go, you can do more and more steps. And then later on, as you will see, you can filter people to your ICP based on the information that you collect. So it's like a get raw information, enrich the information. filter the information, create an output out of the information. So that's like the process that Clare allows you to do, all fully automated with whatever AI steps you want in between. And that's like the general idea. Now let's dive into the specific implementation that Lanny is showing us.

Lanny Heiz:

exactly. so basically the first step, as I mentioned, is we need to fit figure out if they're an ICP fit. We use AI for that. We have a scoring system, one to 10, to what degree do they fit, us. And obviously in the prompt, I specify, hey, this is what we're looking for. This is what we're not looking for. This is what 10 looks like. This is what a five out of 10 looks like, et cetera. So this is ultimately what, what then, AI gives me. And only if we have a little tick here, i. e. if there are an ICP fit, we move them further, along. and. The idea here, by the way,

Isar:

is twofold. One is you only want to address your ICP, otherwise you're just wasting time. But two, it saves you a lot of money because the way this tool works is you're paying per token. And if you're gonna keep moving to the right, more and more steps, you're going to pay all these tokens on everybody on the list versus on 5 percent of the list that is actually your ICP. So you're gaining on the fact you're focusing on the right people, but you're also saving yourself a lot of money in the process.

Lanny Heiz:

Yes, exactly. Exactly. that's basically the idea here. So once they are based in a basic ICP fit, and But the persona is also a fit, like we also evaluate the persona, essentially. then we move them further, to a table where we just enrich the more data. So we find the emails, and, yeah, make sure we have all the data we need to actually reach out to them. and then we write them to a campaign table where we essentially have. Use AI to write all our messages. so essentially what that then looks like is it's going to be super, super personalized. So I can't go fully into that because that would, show the campaign what we literally write. So I can't show the next table. but, if you go on my profile, like a couple of stuff and you fit my SAP, you see what we write. so go do that. but

Isar:

that's the back door, right?

Lanny Heiz:

Yeah, exactly. but, so yeah, that's basically, where we find the emails. and then in the next, the last step, essentially, we just create all the data that we need. to then reach out to, to, to our prospects with a super relevant, message. yeah, like what we regularly hear is Hey, I got lead gen messages and like sales experts messages all the day, but like yours, like just tops it off. and the reason. And it's just generally speaking, there's a good framework to, to keep in mind is how can you not just tell that you're good at what you do? How can you show that you're good at what you do? and, cause everybody can say, Hey, we worked with ABC company and brought them. XYZ result, like everyone can say that and I may believe you, I may not believe you, but I don't have any proof, in a real tangible way. So think of Y's, hey, and this could be your content, this could be your email messages, that could be your ads, doesn't really matter, the concept is universal. How can you really show someone that you That they can trust you ultimately, that you can deliver on your value proposition and case studies are one way to do that. But I think just, it's not as strong as it could be. so that's basically what happens in clay on a,

Isar:

yeah, I have a couple of questions on the practical final steps. So question number one is your outreach via email or LinkedIn or both? And question number two is, I know you can share your secret sauce, but in general, what is the outreach look like? Does it relate to the fact that they connected with your LinkedIn? Does it relate to what you do? Is it more? Let's connect. I can help you. Or is it more, here's some value on stuff that you might be interested in, like on a general level, what's the channel and what's the type of, Yeah.

Lanny Heiz:

so I do, yeah, I do. Yeah. I do acknowledge that we had of interactions prior to LinkedIn. that's because if I did it, like you, like some of you may have heard of tools like that can identify web. Site visitors.

Isar:

Yeah.

Lanny Heiz:

There, this is creepy if you mention that, Hey, I saw you on this is creepy, but hey, you commented a couple of times on my post, this is not creepy.'cause they know that I know. it's quite common. so that, yeah. So I do acknowledge that. It's hey, yeah, thanks for a couple of interactions or we had some interactions on LinkedIn, I thought Come over, say hello, essentially. and then I can make, so it depends if it's email or LinkedIn. LinkedIn is way more chatty, way more conversational. like people are on mobile. So you want to keep it short. You want to essentially like, if you want to, and I'm going to drop a guide in two weeks on this. Um, on LinkedIn, obviously on, on how to essentially sell by chats, or just basically how to chat and to book meetings through chat. but the baseline framework is just curiosity, like actually ask questions, Hey, so for example, if you interacted with me, a couple of, times and locate which posts really resonated with you. that's Oh, what's really resonated with you about my posts? and then it's Oh yeah, I really love how you build AI systems. okay, that is like my way to your goal. What, how are you doing right now? And then it's not creepy. Then it's just like a very normal conversation. If we had that conversation, on a barbecue or in a bar at a networking event, that's like a very normal conversation, the same way we would have it anyways. And so this is like the framework, like, how can I make it like not weird? And with email, it's essentially the same thing. I just have a little bit more real estate to give a bit more context.

Isar:

Yeah. Because people don't know where you showed up.

Lanny Heiz:

yeah, exactly. Exactly. So this is, but the framework is always the same. hey, acknowledge that we had some interactions. I assume this might be a problem. How are you currently solving that? Here's some resources on solving that, etc. that's the general frame. and I said, The response rate is phenomenal and like we book, yeah, like with very little effort, lots and lots of meetings using that process.

Isar:

Lani, this was fantastic. we touched on anything from really important concepts all the way to the tactical small details on how to actually implement the stuff and the tools and everything. really valuable. If people want to. work with you, follow you, learn from you, et cetera. What are the best ways to do that?

Lanny Heiz:

Obviously LinkedIn. I have to say that now. I know very easy. LinkedIn forward slash Lanny minus heights. yeah, connect with me, say that, yeah, you saw the show, today and then, yeah, we take it from there. If I can help you can jump on a call, and, yeah, take it from there.

Isar:

thanks to the audience. All the people were chatty, on LinkedIn and on zoom. That was great. Like a lot of conversations happen in the background. the one thing that was asked and I replied on LinkedIn that I will respond to here as well is. Somebody asked Samir, I think he's from Atlanta. He said that in the beginning asked, what were the three documents in Claude? So one was the client information that was generated using AI based on transcriptions of, Call like sales calls. The other was the type of posts that are successful on LinkedIn or on whatever platform, like you can use this for tech talk if you want to do, but then you've got to figure out tech talk conversations and what's successful over there. And then the third one was the personal story and the benefits and the value on your experience and so on. So these are the three files for those of you who missed that. Other than that, again, thanks everybody for being here, for taking the time and spending a full hour with us. That is incredible. And I don't take this for granted. If you haven't, liked and follow this podcast before, if you're not here live, but you're on the podcast, first of all, come join us live. We do this every Thursday at noon, literally every week. So you can join us for that. if you want more of an open conversation with me and a lot of people in my community, you can join us live every Friday at 1 p. m. We do AI Friday Hangouts and there's a huge group of people that meets weekly and just talk about AI and we solve problems together and we talk what's happening together and stuff like that. and obviously share this podcast, like it. And rate it on your favorite podcasting platform. and that's it for today. Thanks again, Lonnie. This was absolutely fantastic.

People on this episode

Head Shot

Isar Meitis

Host