In this episode of Scaling Smarter with AI: The Art of Scaling Business with AI, host and guest David Hirschfeld, founder of Tekyz, dive deep into how AI can be strategically leveraged to overcome common scaling challenges. They discuss using AI for idea validation, niche identification, lead generation, workflow automation, and contract management, sharing real-world examples and practical techniques. David emphasizes the power of AI in creating playbooks, automating bottlenecks, and guiding early business decisions, while highlighting the need for adaptability in the face of rapid technological change. This conversation provides a roadmap for startups and scale-ups to scale smarter, faster, and more efficiently by embracing AI’s transformative potential.
Scaling Smarter with AI: The Art of Scaling Business with AI
This discussion centers around the challenges of scaling a business and how artificial intelligence (AI) can be leveraged to overcome these hurdles. The Host, along with guest David Hirschfeld, founder of Tekyz, explore various AI applications for startups and scale-ups, from idea validation and lead generation to workflow automation and contract management. The conversation also touches upon the transformative impact of AI on business practices and the importance of adapting to this rapidly evolving technological landscape.
Leveraging AI for Startup Success
David emphasizes the importance of thoroughly vetting a business idea before product development. He suggests using AI for research to determine the viability of the idea, identify potential niches, and pinpoint the root-level problem statements within each niche. He highlights the need to focus on a specific niche with the highest perceived impact and cost associated with the problem being addressed. David’s company is developing an AI product to assist founders in this process, helping them identify early adopters and validate product-market fit before significant investment.
AI for Scaling Existing Businesses
David explains that scaling requires adapting strategies as a business grows. He advocates for creating “playbook strategies” to capture tribal knowledge and facilitate training and onboarding. AI can be used to organize and query these playbooks, enabling efficient knowledge transfer and reducing reliance on key personnel. He also suggests using AI to generate questions about scaling, providing a framework for developing a comprehensive plan. David stresses the importance of detailed prompts and specifying the perspective (e.g., CEO, marketing analyst) for more tailored results.
AI for Automation and Workflow Optimization
David highlights the significance of automation in scaling, particularly for addressing bottlenecks in various business processes. He recommends identifying manual processes and prioritizing automation based on the biggest bottlenecks. He cites Michael Dell’s approach of focusing on and resolving bottlenecks to create a smoother workflow. David also discusses the use of AI in specific areas like accounting, where it can automate data collection and free up time for strategic planning.
Practical Applications and Tools
The Host shares his experiences using AI for contract management, creating custom GPTs for workbooks, and generating core values for businesses. He also mentions using Notebook LM to create audio versions of newsletters. David discusses using AI for website analysis and content creation. Mimi, the editor-in-chief of Startup Club, raises concerns about the varying responses from different AI platforms, particularly in legal matters. David advises triangulating responses from multiple AI tools, especially for critical decisions.
Conclusion
The discussion underscores the transformative potential of AI in scaling businesses. By leveraging AI for idea validation, workflow automation, and strategic planning, startups and scale-ups can overcome common challenges and achieve sustainable growth. The conversation also highlights the importance of using AI strategically, understanding its limitations, and always verifying information before making critical decisions. The speakers encourage listeners to explore the various AI tools available and adapt them to their specific business needs.
Transcript
[0:00–0:13] Host: Scaling smarter with AI, the art of scaling business with AI.
[0:13–0:29] Host: You’re listening to Start Scale, Exit Repeat, Serial Entrepreneur Secrets Revealed. And you probably don’t know this, but this is actually a live show. We do this on Clubhouse on a club called Startup Club with almost 1 million members.
[0:29–0:39] Host: If you’re not already a member, definitely join. If you haven’t already signed up to our newsletter, definitely sign up to that newsletter because that newsletter is just a wealth of information.
[0:40–0:50] Host: We come out with one every month, and it is incredible. We got everything from the Time to Sell Index to leading articles by myself and now by others for Startup Club.
[0:50–1:02] Host: But today’s show, we’re going to be really getting into scaling and scaling smarter with AI. And we have a guest author coming on.
[1:02–1:20] Host: Our guest is David Hirschfeld, the founder of Tekyz. Tekyz, yeah, that’s right. And he’s really quite good at understanding the whole scaling and AI.
[1:20–1:33] Host: When I did go look at his website and tried to understand it, he really has a good understanding of that. I don’t know where he is right now. I know it’s a little bit of a difficult platform to maneuver sometimes, but maybe you can ping him into the room.
[1:34–1:47] Host: Before we do that, again, I just want to really emphasize that you can sign up to Startup Club for free. We have a lot of different things happening at Startup Club. We’ve got the weekly show we do here on Clubhouse.
[1:47–1:58] Host: We do this every Friday, 2 o’clock Eastern. You can listen to this in podcasts or on YouTube or come in person and listen live. A blog, which is getting a lot of attention by Google.
[1:59–2:12] Host: I’m amazed at how well it’s doing on Google. And the fact that when an article is posted on that particular blog, it gets to the front page of Google. Almost like up there with some of the articles I’ve done for Forbes when I write an article.
[2:13–2:24] Host: The score is pretty high on Google. I’ve written a number of articles for Startup Club. and it’s hit the top page of Google as well. I think one of them was the top books for entrepreneurs in 2025.
[2:25–2:38] Host: And if you type that in Google, you’ll see Startup Club, the article that was written on Startup Club. And then another one was top movies to watch for entrepreneurs in 2025. And I think we actually got the number one link for that.
[2:39–2:50] Host: And I’m working on another one right now, top streaming shows, but that one’s actually going to be published on Forbes. It’ll be out in the next week or two, hopefully. I’m mentioning this because you may have some expertise.
[2:50–3:04] Host: You may have something that you want to share with the startup community. And not only can you come on stage and ask questions, share your advice, your expertise, but you’re also welcome to become a guest author.
[3:05–3:18] Host: And I believe now we have a link on the website, Mimi, is that correct, where people can go and under About Us and become a guest author or guest speaker. but we are definitely looking for more contributors for Startup Club.
[3:19–3:31] Host: And if you’re one of those people out there who has an expertise and would like to share that with a large community, then we would be very happy to have you on Startup Club and doing some writing for Startup Club.
[3:32–3:44] Host: Start, Scale, Exit, Repeat. It is the largest section of the book. It is the most complicated component of Start, Scale, Exit, Repeat because it’s not easy.
[3:44–4:06] Host: 99% of companies in America fail to scale. And there are many reasons why they do that. I point out the number one reason is the entrepreneur themselves. And now we do have David Hirschfield, a 35-year software developer veteran, and he’s helped a lot of startups, a lot of technical innovation, a lot of business growth.
[4:06–4:19] Host: And he started TechYZ in 2007, And he’s also emerged as a strategic advisor specializing in AI-driven workflow transformation for scale-ups.
[4:21–4:31] Host: David, welcome to the show. We’re a bit of a different style of podcast. We like to have a live audience, live show, and we do this every Friday, 2 o’clock Eastern.
[4:31–4:43] Host: You can come on stage and join us. If you’re in the audience right now, you want to ask a question of David, feel free to come up on stage and ask a question. David, welcome to Clubhouse and Startup Club.
[4:43–4:54] Host: We’re excited about the topic today around how you scale your business or how you scale your startup using AI. So right on the bottom right, start, scale, exit, repeat.
[4:54–5:06] Host: There are about 30 chapters in the scale section. And I think the one that interests me the most or the one that I think is most fascinating is on the X factor.
[5:06–5:18] Host: and how startups can launch a business, but if they can find something unique and different, go into the white space. That’s what Joe Foster, who founded Reebok, calls it.
[5:19–5:33] Host: Go into the white space. And if you can do that, then you can position yourself to win in a marketplace. It happened in a couple of my companies. When we figured out our X factor, the dominoes fell and we completely dominated that particular industry.
[5:33–5:48] Host: Think dominoes pizza, 30 minutes are free. national car rental just show up at the lot and go. There’s a lot of examples out there of companies who, when they figure that out, that key component, that key strategy, they can just take off.
[5:49–5:58] Host: Mimi, maybe you can back call him. I’m just going to add David as a friend. I can see he has zero friends. Oh, we lost him. We lost you, David.
[5:59–6:00] Mimi: I’ll invite him back right now.
[6:00–6:11] Host: Invite him back in. So let me tell you a story about Joe Foster. Joe Foster, who founded Reebok. First 19 years business life, he got to $9 million in sales.
[6:12–6:24] Host: He spent almost 10 years trying to get a distribution deal in the United States. And so 1979, he applies for three five-star awards. And he actually wins them.
[6:25–6:37] Host: He won them on Runner’s World magazine. He won these three five-star awards. And he caught the attention of Paul Fireman, who was a distributor in the United States, and who agreed to take on the Reebok product line and distribute that.
[6:38–6:57] Host: One of the employees in the United States saw their wife doing something called aerobics. They were using ballet shoes. So he actually recommended to the management team that they do something that no other shoe company has ever done before, which is get into design aerobics shoes for women.
[6:58–7:09] Host: And there’s a lot of pushback because running was everything. He just won the awards for Runner’s World, not for aerobics. But they went out on a gamble, and they designed shoes for aerobics.
[7:09–7:25] Host: And somebody on television wore those shoes. It was Jane Fonda. And when she wore those shoes, the company went from $9 million in 1979 to $900 million in 1984.
[7:25–7:41] Host: Became the largest shoe company in the world in 85, 86. and after that, obviously, Nike eclipsed them. It was just an incredible story of how when you find your X factor, when you go into the white space, as Joe Foster calls it, you can accomplish a lot.
[7:42–7:54] Host: We’d like to share your expertise with those here today and also those on our podcast. You can also post a question in the channel and we’ll try to follow the channel, but I highly recommend that you just jump on stage and share your experiences.
[7:54–8:10] Host: I know I’ve used AI a lot to help scale the business. I’ve used it for so many. One of them I’ve used it for is contract management. So right now, every contract I get, I create a custom GPT and upload that contract into the custom GPT.
[8:10–8:23] Host: And then it becomes basically an amateur lawyer. Okay, so I’m not saying you’re going to replace your lawyer 100% right now. But I think you can do about 90% of what your lawyer does you can do with contracts.
[8:23–8:33] Host: It’s just a lot easier than cutting and pasting. And if you have a big question about a contract, you just jump over to the custom GPT and you ask that. Colin? David, welcome to. Finally.
[8:34–8:37] David Hirschfeld: Sorry. I was in the wrong place. And now I’m in the right place.
[8:38–8:49] Host: Awesome. Very good. By the way, I could talk forever about scaling and AI. Because those are the two parts of my world right now that I’m living in. I’m surrounded by it. I use AI probably six hours a day.
[8:49–9:00] Host: It’s no longer Google on my Chrome. It’s perplexity. It’s completely changing the way startups and founders are launching their businesses or running their companies.
[9:00–9:11] Host: But, David, thanks again for having you on here. And we gave you a little bit of an intro earlier on. I know you’re an expert on scaling, and you’ve launched a company called Tekyz.
[9:12–9:19] Host: We’ll talk about your company a little bit later, but let’s jump right into AI and how you see AI helping startups scale.
[9:20–9:30] David Hirschfeld: Okay. I’m also curious, you said you’re using it like 12 hours a day. How are you using it? Okay, six hours a day.
[9:30–9:40] Host: My wife, she’s a little too irritated about that. But, no, it is a default. So I use it for, I mentioned earlier before you were on, I use it for contract management. I create custom GPTs.
[9:41–9:52] Host: I’ve also created for the book, Start, Scale, Exit, Repeat. We did a workbook, Mimi and I and the team here, and we created custom GPTs throughout the workbook. Scale section.
[9:52–10:02] Host: And there’s everything from I was just doing a podcast with JLD the other day on Entrepreneurs on Fire. And one of the custom GPTs we did was on companyvaluesai.com.
[10:02–10:19] Host: So we uploaded the content from the book into the custom GPT, set some parameters, and now you can go to companyvaluesai.com, and it will literally give you four to six rhythmic value statements from people we interviewed in the book and in the book.
[10:19–10:30] Host: So it’s that kind of stuff that we’re doing. I could go on for 30, 40 minutes of the types of little hacks that we use it for. But I think I’ve been really impressed with custom GPTs.
[10:30–10:43] Host: I really like Notebook LM. On every newsletter now, we do a Notebook LM. And that’s the — it’s like a — then we turn that into a podcast. So we put it on YouTube. If people don’t want to read the newsletter, they can listen to the notebook.
[10:43–10:49] Host: I find it funny because we do a lot of work on this newsletter. But when you listen to the notebook, it gives you a bit of a different perspective.
[10:49–11:00] David Hirschfeld: I’m curious from you. When Notebookit first came out, like maybe in the first two or three days, I pointed it at my website and had it create the podcast discussion about my website.
[11:00–11:17] David Hirschfeld: I was blown away by it because, oh, magical, what AI can do. But what I found was listening to these two people talk about my website, it created some messaging for me that I didn’t realize we had on the website that really didn’t speak to what we do.
[11:17–11:31] David Hirschfeld: and it brought that into context in a way that was meaningful. So I realized I had to make some changes to the website. I would not have gotten that if I didn’t listen to people debate the pros and cons of our business based on what they found on the website.
[11:31–11:42] David Hirschfeld: Things like that are huge. And all they did was point it at the website and listen to it talk to itself about the website. So that doesn’t really help people scale necessarily.
[11:42–11:59] Host: You know why I like it, though, is we do it with all the articles and newsletters now, and I think it enhances the newsletter. I think it adds a whole new dimension to the newsletter. I can be driving to work now, click on the Startup Club newsletter on YouTube, and get my monthly newsletter through that versus — now, I know it won’t cover everything.
[11:59–12:15] Host: It’s amazing. We have a 50 growth hacks and we had one on that It didn go through all 50 growth hacks It picked out two or three or four of them and talked about those and emphasized those Then you get a sense of what in the newsletter You say oh I got to go check out this newsletter I think it has a dimension I didn’t know the piece of it.
[12:15–12:16] Host: It’s just one of many of these pieces.
[12:17–12:40] David Hirschfeld: What we typically, I’ll take these interviews and also on my podcast when I interview somebody, we take the transcript, we generate a medium article, put the transcript in there, put a summary in there, put a bulleted index of topics that are talked about in there, and we post a podcast episode or YouTube video in there as well.
[12:40–12:51] David Hirschfeld: What do you think about also having Notebook LM generate a review of the discussion and putting that in there as well? Yeah. It basically does an analysis of the thing, right?
[12:52–12:52] David Hirschfeld: Yeah.
[12:52–12:55] Host: It’s not going to give you a 4.5 out of 5 stars.
[12:55–13:02] David Hirschfeld: No, not that, but maybe have a podcast about the podcast. Let the two people talk about what you guys talk about and what they think of it.
[13:03–13:14] Host: Yeah, I think we’ve got to use it carefully because it can get boring really fast. Like, when I go back right now, not as smooth as they are, though, if you ever listen to them, they’re really trying to emulate the population when they speak and whatnot.
[13:14–13:25] Host: I think we’re more authentic, and I think people will pick up on that. So I think there’s a place for it. But all human beings talking the way we’re talking about particular things and trying to help startups figure out how to scale their businesses.
[13:26–13:27] Host: I just don’t think you can replace that.
[13:27–13:37] David Hirschfeld: But some people either don’t have the time to listen to the whole thing today, but maybe there’s a three-minute summary discussion about it so that they can get the highlights in a verbal way. It’s just an idea.
[13:38–13:44] David Hirschfeld: I wanted to speak it out loud because then I’ll remember it, and I’ll hear if it makes any sense to me while I’m saying it out loud.
[13:44–13:54] Host: David, somebody gets an idea for a business. Yeah. Okay, my idea for this business is we’re going to take recycled bottles and we’re going to make baby carriages or something like that.
[13:54–14:02] Host: How do I use AI to get that idea to action? How do I use AI to help me get that idea out of the gate?
[14:03–14:15] David Hirschfeld: Okay, so the biggest thing with startups, the reason people are entrepreneurs is because they go, that doesn’t make sense or there needs to be this because of some pain that it eliminates in people’s lives, right?
[14:15–14:29] David Hirschfeld: That’s where entrepreneurs come from. And then entrepreneurs are the ones that gripe about something, think there’s a better way of doing something, and then decide they want to act on it. The very first thing you need to do is what’s the genesis of this idea?
[14:29–14:39] David Hirschfeld: Is there a real reason that needs to be done other than it just seems like a good idea? And then you start with the research, like what is the viability of this?
[14:40–14:58] David Hirschfeld: Are people already doing this? Can AI quantify for me how much of a pain point this is? and then say, okay, it seems like it is a pain point. There is definitely an industry out there that needs this based on social signals and research databases and things like that.
[14:58–15:12] David Hirschfeld: And say, okay, fine. So now I need to go a little deeper, right? I need to figure out who are all the niches I might be able to market this to. And this really needs to be every entrepreneur’s number one job when they think of a concept that they want to pursue.
[15:12–15:44] David Hirschfeld: what are all the root level problem statements and you figure out a root level problem statement when you’re in the shoes of the stakeholder the buyer in that niche in one of those niches and you ask why does this problem statement matter and you usually have to ask that question two or three times or four times before you get to the root level problem statement of why it matters and it has to start out with this problem it makes me miserable or i hate this problem or i feel threatened by this problem because if I don’t fix it, this will happen.
[15:45–15:57] David Hirschfeld: And that’s when you know you’re at a root-level problem. Anything less severe than that isn’t a root-level problem statement, and so you don’t really understand the triggers that you need to understand before you’re going to be able to reach people to buy a product.
[15:58–16:19] David Hirschfeld: So once you understand all the root-level problem statements that you’re solving across all the various niches, then you have to find one niche out of that list of niches whose problem statements, when you score all their various problem statements, has the highest perceived impact and independently the highest cost because of the problem that they’re struggling with that you’re trying to solve.
[16:20–16:32] David Hirschfeld: That’s the number one. AI is really good at helping with this. In fact, we’re building an AI product. We’re in the early stages of building an AI product to do this for founders because it’s such a tedious task to do this.
[16:33–16:36] Host: You’re designing a product that’s going to help them vet their idea, basically?
[16:37–16:47] David Hirschfeld: It’s going to figure out who’s the early adopter. Yes, exactly. Hopefully they’ve done some betting even before they get to this point, but maybe not. If it’s easy enough to do with the tool, then maybe that is the first step.
[16:47–17:07] David Hirschfeld: So it’s going to identify who’s the early adopter, which is the one or two problem statements that have the biggest impact to them that you’re addressing. People think you can sell, unless you’re a mass market product and you’re on Kickstarter and doing a prelaunch sale to an ass market, You can really only focus on one niche at a time.
[17:07–17:22] David Hirschfeld: And the reason you need to figure that out is because the tighter you can figure out who that early adopter niche is, the tighter that niche is, the more you can charge and the higher your closing ratio. If you’re broad, then you can’t charge as much, and your closing ratio is going to be a lot lower.
[17:22–17:24] : Your marketing costs are going to be much higher.
[17:24–17:38] David Hirschfeld: You’re telling me there’s a lot of water in the desert. Exactly. Somebody who has a dire need for that water but has no money, and that’s not a good target. You may have somebody who has a lot of money but doesn’t feel the need for it, and you can’t get their attention when you’re marketing.
[17:38–17:50] David Hirschfeld: So that’s why you need both. Costing them everything, that person may not feel the need to do anything about that problem right now, and that’s not your early adopter. Even though you can say, but look how much this is costing you.
[17:50–18:04] David Hirschfeld: They say, yeah, I know, but it costs everybody in my industry that much, and my company is just not going to get behind me on this because they have other things they’re trying to accomplish right now, or it’s too many things that we have to change internally, bring that on, even though it would really benefit me.
[18:04–18:24] David Hirschfeld: It’s not going to benefit the company as a whole. So that’s why you have to measure both those things independently. I was just having a conversation today with a founder who’s trying to scale and running into this problem on a product where he’s won awards about how this problem will literally increase the value of a company, and he’s having trouble because he doesn’t nail the early adopter down properly.
[18:25–18:35] David Hirschfeld: He’s going after people that want the product, but they’re not willing to push past the corporate inertia to then bring the product in and get it going because it’s going to be a big change.
[18:35–18:49] David Hirschfeld: So he’s going after, and he doesn’t have the CEO’s buy-in for implementing this product because of the way he’s been marketing and targeting and messaging. He’s been going after the CFO or the director of marketing or people like that.
[18:49–19:00] David Hirschfeld: And they want it, but they don’t want it badly enough that they’re willing to fall on a sword to try to bring it in because the CEO hasn’t gotten behind it. So we had a whole conversation about the same exact thing this morning.
[19:00–19:10] David Hirschfeld: You have to figure this. And you don’t want to wait until you’ve put several million dollars into your product, like in the case of this guy, to try to figure this out. Figure it out before you even start building the product.
[19:12–19:17] David Hirschfeld: And you can test it and validate it solidly before you even start to build your product.
[19:18–19:34] Host: I’ve used AI to try to vet ideas. I’ve used AI to help me pick a domain name. I’ve used AI to help write business plans. I’m just finishing a private placement memorandum right now. It’s been $20,000 15 years ago to a law firm do a private placement memorandum.
[19:34–19:49] Host: This thing’s amazing. It’s done 40 risk factors. Now, I did a lot of the writing myself. I had it cleaned up, so it just makes my writing faster, more efficient. And by the way, if you’re writing an article for Forbes or for Startup Club, you can’t use AI at all unless you’re using it for research.
[19:49–20:03] Host: It’s got to be original material. But if you’re writing a business plan, you’re writing an investor plan, the more that you copy or use other people’s words in their documents, the more likelihood that you’re going to get that funding.
[20:04–20:15] Host: So there’s a totally different concept than in school. Plagiarism in PPMs and S1s, I’ve done an S1 too. S1’s extreme. Like S1, they won’t even look at language. It’s not even — the SEC look at language.
[20:16–20:24] Host: It’s not even been already approved before. They literally have to write language in there that actually is approved, and that’s where AI can really come in very nicely.
[20:25–20:28] David Hirschfeld: You still want to get it reviewed by an attorney to make sure that –
[20:28–20:32] Host: I’m still going to get an attorney reviewing it, but I’ve done like 95% of the work here.
[20:32–20:42] David Hirschfeld: Exactly. Documents. Exactly, and it’s not going to cost you that much now. It’s going to cost you a small fraction of what you would have paid to have the attorney create something for you. Yeah.
[20:42–20:51] Host: AI is amazing. It’s a reg D, so we’ll file with the SEC, and we’ll do the reg D for sophisticated investors. But let’s talk about that is your area. That’s your expertise in your company.
[20:51–20:52] David Hirschfeld: Startups and scale.
[20:52–20:58] Host: How can you use AI to help scale a business? Because it really is a tough thing to do. Most companies in America don’t scale. That’s reality.
[20:59–21:14] David Hirschfeld: The very first thing you do, because what you do to get your first customer is not the same thing that you necessarily do to get your next 10. And then once you’ve got 10, now that same playbook isn’t going to get you from 10 to 100 or from 100 to 1,000.
[21:14–21:27] David Hirschfeld: And every time you hit the next log limit, you have to step back and say, okay, now what are we going to do to get from 100 to 1,000? Because it’s not going to be the same sales, marketing, infrastructure, support.
[21:27–21:51] David Hirschfeld: Something is going to change for us to be able to achieve that. And what is that? The very first thing in the scaling process is your ability to create infrastructure in your organization so that you can bring people in to accelerate growth, whether it’s support, whether it’s marketing, whether it’s sales, whether it’s operations, it’s logistics, how do you bring new people in?
[21:51–22:06] David Hirschfeld: And what is your ability to bring those people in and train them without sucking the life out of the people that have all the tribal knowledge in your organization about that job? So the very first thing I say is create a playbook strategy.
[22:06–22:19] David Hirschfeld: And a playbook means that you’ve got somebody who’s got their tribal knowledge about how to do a job really well and all the nuances, other than just the basic steps, but all the nuances about that job.
[22:19–22:38] David Hirschfeld: And now you have to use AI to capture all that information and organize that playbook. So the idea is that once you’ve got that playbook document and then use AI to, like a notebook LM, about that playbook so that you can bring somebody else in, now they can ask questions.
[22:38–23:01] David Hirschfeld: They have an easy way now to query if they have questions about their job or a particular decision point comes in and it’s nuanced. But that should be documented in the playbook because the person with all the tribal knowledge downloaded it into a document of some kind that’s in Notebook LM or they recorded it and that got ingested in Notebook LM or whatever, or Zoom video and that got ingested in Notebook LM.
[23:01–23:14] David Hirschfeld: So there’s a lot of different steps here that AI can come into play. Okay. You tell AI, this is my business. Here’s my website. Here’s documentation that documents my business.
[23:14–23:25] David Hirschfeld: And then you say, knowing what my business is, and this is the revenue I’m currently at and so many employees I have, I want to grow to three times to five times this in the next two years or whatever the goal is.
[23:26–23:39] David Hirschfeld: What questions should I be asking about my company before I create a plan to scale my company up? And it will give you the questions to ask. So that’s where I usually start is ask the questions you should be asking.
[23:39–23:50] David Hirschfeld: Ask it for the questions that should be asked. And then have it answer those questions for you. Say, great, I like those questions seem relevant to me or they’re not, and explain why, and it will make changes to it.
[23:50–24:00] David Hirschfeld: Now answer those questions based on what you know about my industry and competitive pressures and things like that. Make sure those prompts that you write are very detailed. They provide context.
[24:01–24:27] David Hirschfeld: Tell it what perspective it should be considering this from. from a CEO perspective, from operations, from a marketing perspective, as you’re a marketing analyst. Act like a marketing analyst doing this plan Act like the CEO doing this plan And you get different results if you do it that way and they be much more specific And then as it’s answering those things for you, there may be something in there that you like but you needed to expand.
[24:27–24:39] David Hirschfeld: You say expand on point number two or expand on point number three. And pretty soon you’ve got a whole scaling plan. Then you can start asking questions about, okay, now I need to execute this. How do I start to implement some kind of execution plan for this?
[24:39–24:47] David Hirschfeld: So these are things that this is like every single day doing this kind of stuff with AI. You said you spend hours every day on AI?
[24:48–25:04] Host: Exactly. Now, one of the things that my co-host who’s not here today, she’s on a plane, but Michelle Van Tilborg, she does a pretty unique technique where she’ll say to the AI, okay, I want you to be my business coach, and I want you to ask me questions.
[25:04–25:16] Host: That’s great. What I can do to scale the company. And I like the way it does that. It’s now prompting, actually emulating a coach, prompting you questions, asking you questions as if it was a real coach.
[25:17–25:19] Host: That’s a pretty neat technique she’s been using and adopting.
[25:20–25:28] David Hirschfeld: I love that. And I’ve not heard that one before, but I’m always asking it to ask them questions in different contexts. That’s a new one. It’s now in my playbook.
[25:28–25:42] Host: You could ask, hey, be a lawyer, be a doctor. Oh, my gosh, I got my medical results back for my breathing tests and all this stuff. I got this arthritis and I’ve had it for 20, 30 years, and they’re all trying to figure it out, and I can’t understand these results.
[25:43–25:47] Host: Right. So I logged on, and I just said, be my doctor, and I hear my results. What does it tell me?
[25:48–26:00] David Hirschfeld: You’re not going to just do whatever it says, right? You’re going to use that to help inform you, but at least that’s what you should use it for. AI should always be used to inform you, not to decide for you.
[26:00–26:11] David Hirschfeld: Maybe it informs you on something so obvious you feel like that’s a perfectly good decision, but you should always, it’s not always right. It does still hallucinate. It makes things up.
[26:11–26:28] David Hirschfeld: And one of the ways you can see this is ask it to research you. Use the deep research tool, for example, on one of these tools, and ask it to research who you are and give it enough information so that it’s for sure going to find you because you’re an expert on you, right?
[26:28–26:41] David Hirschfeld: The one thing that they are a complete expert on is themselves. So ask it to do a full research and background and describe everything about who you are, what you’ve done, your history, everything.
[26:41–26:53] David Hirschfeld: And you’ll find that it will make some things up. It will make assumptions about who you are based on what it finds. So you still have to be careful. It’s so much smarter than it used to be. You still have to take it all with a grain of salt.
[26:53–27:05] David Hirschfeld: But what it can do, it’s just like magic. I used it the other day. I was in Lowe’s. I had a garage project that I was working on. My wife called me. She says, I need fertilizer for my hibiscus. Can you get me a bag of it?
[27:05–27:20] David Hirschfeld: So I looked. The word hibiscus was not on any of the 100 bags of plant food and fertilizer that I found. So I stepped about 30 feet back from all of the shelves that had all these bags of fertilizer, and I opened up ChatGPT, and I tapped.
[27:20–27:32] David Hirschfeld: I went to the voice portion of it and then tapped on the camera, and I pointed it at all these shelves. and I said I’m looking for plant food that will feed hibiscus and then
[27:32–27:36] : it said yes I see plant whatever the brand was molytone
[27:36–27:47] David Hirschfeld: or something like that and this is exactly what you would want to feed hibiscus they said it’s red and white and so I went walked up to a red and white bag which was not the right one and so I walked up to it I said is this what you’re talking about?
[27:47–27:58] David Hirschfeld: She said no that’s not it it’s farther to your left she couldn’t see it when I was walking up she said it’s to the left on the second shelf from the top. So I just turned my camera and pointed in that direction.
[27:59–28:04] David Hirschfeld: So as I see it, it’s about three bags from the end on the second shelf. And I walked over there. I love it.
[28:04–28:16] Host: I love it. That’s magic. Every aspect of life, you can use it. Here’s one that I did the other day on a podcast. It was a podcast with JLD. And it was the Ultramers on Fire podcast.
[28:17–28:31] Host: And we developed for the scale workbook, which has not released yet, custom GPT for core values. The idea is to create four to six core pillars of rhythmic statements based on questions it asks you.
[28:31–28:46] Host: We uploaded the information in the book, and I mentioned this earlier in the show. But people can actually go to it at companyvaluesai.com and create these core values. And core values are one of those things that you really need to be able to scale your business.
[28:47–28:59] Host: And AI can play a role in many of them. Let me give you another example. We have one called personalityprofileai.com, which goes to another custom GPT, which figures out your disk profile, your Myers-Briggs profile, to identify these positions.
[29:00–29:11] Host: And with these positions, you actually say, okay, this would be the ideal profile, personality profile of an individual that I want to hire. I want a sales leader. Okay, I want someone who’s got high dominance and high influence.
[29:11–29:23] Host: Great. And so what I can do with this program is you can have people — everything’s free. This is going to cause disinflation. I’m not going to go for deflation right now, but I think productivity boom is going to occur.
[29:24–29:33] Host: I think so, too. A lot of these programs in the past, they used to charge money for these programs. Now you can get them for free. What are some other components of scaling?
[29:34–29:45] David Hirschfeld: Automation. You have all these workflows in your business, right? And the bigger you get, the more workflows you have. This is why you want to document playbooks so that you can bring other people in and not suck the life out of your key people.
[29:45–29:58] David Hirschfeld: And also, because the more playbooks you have, the more intellectual property value your business has. Because if somebody acquires you, they’re also acquiring all this well-documented knowledge about how to operate and how to run and how to sell and how to grow.
[29:58–30:09] David Hirschfeld: But you always run into bottlenecks when you’re growing because there’s people-dependent processes that are repetitive. They may be in logistics. They may be in shipping. They may be in ordering.
[30:10–30:21] David Hirschfeld: They may be in invoice to order reconciliation because you have people. So you have all these things. They may be in marketing and outreach, but most of these things can be automated.
[30:22–30:32] David Hirschfeld: But you don’t want to automate them all at once. So you’ll list down everything that is a manual process and focus on the ones where you know that there’s bottlenecks in your business.
[30:33–30:39] David Hirschfeld: If you were able to eliminate any restriction in this area, your business would automatically grow,
[30:40–30:43] : except it would find the next bottleneck upstream or downstream.
[30:44–30:54] Host: And that’s what happens. It’s hard of scaling that it’s always bottlenecks. There’s always a bottleneck. If a bottleneck is not enough sales, why do we have enough sales? We don’t have enough salespeople. We don’t have the right salespeople.
[30:55–31:06] Host: There’s always a bottleneck. And every time you clear up that bottleneck, it points to another bottleneck. You can’t sell that bottleneck. And that’s one of the ways you can help scale your company just by limiting those bottlenecks.
[31:06–31:16] David Hirschfeld: I don’t have enough leads. So you come up with a program that’s generating more leads than you can actually handle. now I’m not qualifying my leads by taking us too much time.
[31:17–31:29] David Hirschfeld: So come up with some kind of workflow to qualify those leads, and now you’ve got a qualified prospect. I’m not closing one. This is what you do with your sales pipeline, right? It’s like that in operations, in finance, in recruiting.
[31:30–31:53] David Hirschfeld: Maybe you’ve got a recruiting bottleneck because you’re not getting enough profiles in or because you’re not finding enough of the right candidates. So there’s all kinds of automation you can do there. Maybe you’re an accounting firm and you want to automatically pull in information from all the different financial places where finances for a particular business are captured because you need all that to be able to do their taxes.
[31:53–32:10] David Hirschfeld: If you automate all that, you’ve got 80% of the manual effort that goes into running an accounting firm addressed. And you can focus much more on financial planning and strategic planning for the business as opposed to spending all your time doing the bookkeeping and collecting everything for taxes.
[32:10–32:24] David Hirschfeld: So you list all the areas where you think you might have bottlenecks, and then you prioritize them to which one’s the biggest bottleneck today, and that’s where you focus your effort. Michael Dell of Dell Computers, when he was asked how he grew his business so big, this was exactly what he said.
[32:25–32:45] David Hirschfeld: He said he just focused his energy on wherever the bottleneck was, and then he fixed that so that became a pipeline that things could flow through freely, which, of course, instantly identified the next bottleneck because then it would flow faster through there except to hit some other point upstream or downstream.
[32:45–33:05] Mimi: Okay, sure. Yes. Thank you very much for giving me time to speak. I do run a small business. There’s a service-based industry. You were talking quite a lot about using AI for generating business plan, values, automation, so on and so forth.
[33:05–33:17] Mimi: One question is that I have been having a go at several different sources of AI, from Gemini to ChatGPT to other ones.
[33:17–33:37] Mimi: And what I noticed that, actually, quite recently, I used both Gemini and ChatGPT for a legal matter. and I noticed that the answers they provide, especially when you say that, act like my solicitor and advise me on this.
[33:38–33:49] Mimi: Read this letter and tell me if you were the lawyer acting for the other side, why would you write such a letter or how you would react to this letter.
[33:49–34:22] Mimi: Oh, that’s great. I love it. And what I was noticing that they’re very different. and at that point I should have noticed that it’s like a for example with the same test results blood results or scan results you can go to two separate doctors and they can give you different opinions this is how I noticed that how different this AI I don’t know how because I guess they all go and scan the whole worldwide web and this is how they get the information but especially I found
[34:22–34:38] Mimi: Gemini. Gemini is more of the conservative one and the chat GPT is more of the aggressive one. And so now going back to the original question, how can you use AI for helping to scale your business?
[34:39–34:50] Mimi: Do you have any specific recommendations? Maybe, I don’t know, this one for business plan, that one for fundraising, that one to increase leads, so on and so forth.
[34:50–35:00] Mimi: Or do you think that it’s better to stick to one of them and the more questions you ask, the more training it gets, and the better results you’re going to get out of the answers it provides?
[35:01–35:02] Mimi: So that’s my question.
[35:02–35:15] David Hirschfeld: I have a question for you first. When you asked them to review the letter, what context did you give them about the company that sent you the letter and about your company?
[35:15–35:26] David Hirschfeld: Did they know where the company was located? Does it understand what the company does? Did they understand your relationship with the company, your business relationship, things like that?
[35:26–35:26] David Hirschfeld: Absolutely.
[35:26–35:37] Mimi: Actually, the company that I had this legal matter with, it was a massive news. I’m not going to name names here because the legal matter can’t really mention anything.
[35:37–35:51] Mimi: But it was like a national news about this company. and as soon as I was typing in something about this company, all of a sudden both of the platforms, they’re recognizing the issues of this company.
[35:52–35:58] Mimi: Then I would say that, okay, what is my issue? Why I’m into this dispute with them? So they had the background.
[35:58–35:59] David Hirschfeld: Okay, okay, that’s good.
[35:59–36:05] Mimi: Coming from World Wide Web, the background of the company for them would be exactly the same.
[36:05–36:15] David Hirschfeld: And the same with your company and what you do. So that was the first thing. And so that’s good. And your relationship with the company, did you specify that in the prompt?
[36:15–36:35] Mimi: Yes, yes. Actually I was going to quite a great extent to explain things upload previous documents Oh good For CatGPT you can upload documents in PDF in Word But Gemini has got, doesn’t really have that sort of multifunction.
[36:35–36:47] Mimi: And sometimes I had to cut and fix things into Gemini because it wasn’t accepting certain type of format of documents we uploaded on. But more or less, I was feeding them with the same information.
[36:47–37:01] David Hirschfeld: My suggestion is try two or three, especially when it’s something like this. You get feedback from two or three different Gemini, Claude, Perplexity, because it has a little bit different take on how it uses the AIs.
[37:01–37:14] David Hirschfeld: And then you triangulate between them. None of them are an absolute source for anything, right? They’re giving you sort of a predictive answer based on what makes the most sense, based on all the things it’s learned in the past.
[37:14–37:26] David Hirschfeld: And all these models are slightly different in how they do it. So I always triangulate when it’s really important like this. And when there’s potential liability or risk associated with how you interpret the answer, then I try to get it from several of them.
[37:26–37:34] David Hirschfeld: And I’m assuming you’re using that to inform you so that when you have a conversation with your solicitor, you can have an informed conversation, right?
[37:34–37:44] Mimi: I know you guys, you sound as if you’re based either in USA or Canada, but I’m based in the UK. And, of course, the legal system is very similar, but it’s still different.
[37:45–37:58] Mimi: What I have to say is that if I want to just rely, especially chat GPT, because it’s quite aggressive in terms of how it advises you, it would have been gone barely up, my choice.
[37:58–38:06] Mimi: It’s good to get some information out of it, but there was no way it could have replaced the solicitor. No way.
[38:06–38:27] David Hirschfeld: No, and people shouldn’t think that it will, but it can replace a lot of what they charge you for. like what Colin was saying, for preparing the document before you go to the solicitor and then have them review it and make whatever changes they need to make so that they’re willing to stand behind the legal agreement, that can reduce your costs a lot.
[38:27–38:30] David Hirschfeld: And in that respect, I think that it’s fantastic.
[38:30–38:46] Mimi: There were certain things that they were trying to put into the document, and my solicitor completely crossed them out, said that’s no good. because it didn’t have enough information about the cases that could have backed up my side of it.
[38:46–38:48] David Hirschfeld: The deep research tool to prepare them?
[38:49–38:52] Mimi: Deep research tool isn’t on ChatGPT?
[38:52–39:07] David Hirschfeld: Yeah, now I think it’s on all of them, but it came out on ChatGPT first just four weeks ago or five weeks ago. So if you go to ChatGPT and let’s say you pick 03 mini high, you should see a button in the search box.
[39:07–39:40] David Hirschfeld: There’s a search tool, right? You can click that so that it’s looking on the web while it’s doing its work, but there’s also one called deep research. And if you do that, it will, whatever prompt you give it, it will always come back and ask three or four or five questions to clarify a few things, and then after you answer those, it will go and it will research 10, 15, 20, 30 different resources and take about 10 minutes to do this and then produce an answer from aggregating all the
[39:40–40:00] David Hirschfeld: information that’s available today specifically about that topic. And it’s quite good. It really impresses me. I’ve used it for many different things. When it’s something complex and I want to make sure I’ve got the best research possible, it still will make mistakes. I’m not saying it doesn’t, but it’s much better than just asking a question, having it spit out the answer right away.
[40:00–40:13] Mimi: I have to say when it comes to writing letters, Gemini is the tenest tips of any of them. It’s so good. The way you write, sometimes it blows my mind.
[40:13–40:29] Mimi: It’s just fantastic. It runs better than it says it. Oh, yeah. But you have to tell it how to write things. But in terms of doing the research, I didn’t find Gemini to be as research savvy as GBT.
[40:29–40:43] Mimi: I haven’t tried the other ones. I just wanted to ask you a question about if you want to use AI for lead generation and capital raising for a business, which ones would you recommend better?
[40:44–40:46] Mimi: Or they are the same to you? I don’t know.
[40:46–41:01] David Hirschfeld: Actually, right now, there are some tools that are on the market for basically LinkedIn, AI applied to LinkedIn outreach, that sort of thing. But the problem with this is that it is changing so fast.
[41:01–41:12] David Hirschfeld: So one thing those tools don’t do a good job with is to orchestrate an entire lead campaign and automate it for you. They can give you a direction on how to do it. They can do a couple of the steps for you.
[41:13–41:24] David Hirschfeld: But then you need other tools that do the workflow automation for you. Maybe it’s Zapier if you’re Simple Tools or Make is a little more sophisticated or M8N if you’re much more technical.
[41:24–41:36] David Hirschfeld: and then you have to build your own workflows because the tools, they don’t do all of that. So when you’re saying lead generation, which part of that process are you asking about?
[41:36–41:47] David Hirschfeld: Because I think they’re all probably better than what we do without using any of them. I lean on ChatGPT most of the time just because it’s consumable.
[41:47–42:05] David Hirschfeld: I’ve got an app. I love the deep research tool. They seem to stay ahead of everybody else in general sense. If I’m doing something that’s more sophisticated in AI, I really like the Gemini pipeline because they have all the infrastructure and deployment tools available that are built into their stack.
[42:05–42:29] David Hirschfeld: for building our own custom AI models where we have our own machine learning engine added to, and we use that to the large language model, whether it’s ChatGPT or Claude or Gemini, is going to then use our database that we built around a specific vertical specialization, like the Launch First that I was talking about earlier doing that niche analysis stuff.
[42:29–42:55] David Hirschfeld: ChatGPT and Claude both now have this context protocol system. It’s called, like, it’s MCP is the acronym, and it allows you to build custom software that’s really intelligent that will, when it knows when it needs to check other sources, and you can provide it what sources are that it needs to check so that it’s a custom application.
[42:56–43:12] David Hirschfeld: So for lead generation, I would look for the tools that are implementing AI into their lead generation process because they’re going to be very specific and basically build a campaign per person.
[43:12–43:23] David Hirschfeld: Find the person that might be your ideal customer profile and then build a drip campaign written specifically to speak to them. And that’s the holy grail for lead generation.
[43:23–43:25] David Hirschfeld: and there’s tools now coming out on the market that do this.
[43:26–43:48] Mimi: So to make sure I understood this correctly, you said that you can just literally, for example, I can go to Gemini and just put down the company details, put down my details, put down the avatar, favorite avatar of the customer that we have, and I’ll ask to generate the campaign to increase leads by 20%, something like that.
[43:48–44:04] David Hirschfeld: It’ll build a drip campaign for you. All the tools will do that. I don’t know which one does that the best because I really have not tested the tools and compared. And the reason I haven’t done that is because I know there’s all these tools coming out, and now there’s tools available on the market that do this for you, and that’s what they’re doing.
[44:05–44:08] David Hirschfeld: They’re basically automating this whole workflow for you.
[44:09–44:11] Mimi: Amazing. Thank you so much. Cheers.
[44:11–44:12] David Hirschfeld: Cheers. Bye.
[44:13–44:14] : Colin?
[44:15–44:29] Host: There we go. I like to say there’s an AI for that. There’s an AI for that. There’s an app for that. There’s an AI for that. And AI, what happened November 23, it was a real shock in the world.
[44:29–44:46] Host: I call it a paradigm shift. We’re seeing a lot of shifts that continually occur, and you can actually ride those waves and those shifts. How does somebody — so can you just talk a little bit about Tekyz and what you do and how you can help people scale their businesses with your company and how they can reach you?
[44:47–44:47] : Sure.
[44:47–44:59] David Hirschfeld: And before I even do that, I want to respond to what you just said about the shift. So the big shift is that technology always expected us to learn how it works, and now technology is learning how we work.
[45:00–45:13] David Hirschfeld: And so the relationship we have to technology and user interfaces and applications, that relationship between us and the application is changing dramatically. And that change is accelerating really fast.
[45:13–45:28] David Hirschfeld: So our expectations are growing really fast in terms of what we should expect from technology because of AI, which is what you were saying. I just did a conversation on the idea that AI is basically making software human first now.
[45:29–45:38] David Hirschfeld: We are a custom development shop. So we do workflow automation. We work with a lot of startups and a lot of scale-ups to build their applications.
[45:38–45:50] Host: The company we have really doesn’t work with startups. We did products for sandals and banks. Are you offshore as well? Yeah, we are. It’s hard to sometimes get these companies to help startups because of the failure rate or, look, I just need a half a startup.
[45:51–45:53] Host: Is that something your company does as well? Oh, yeah.
[45:53–46:08] David Hirschfeld: That’s why I came out with Launch First, which is a methodology specifically for helping startups find a path to revenue and prove product market fit before we start building the product. Because I’ve worked with so many that have failed that shouldn’t have.
[46:08–46:27] David Hirschfeld: And they always fail for the same reason. They wait way too long to validate product market fit. And the only way you validate product market fit is by generating revenue from sales and doing it in enough numbers with a high enough closing rate where your cost of acquisition is one-third or less of the lifetime value of that customer.
[46:28–46:38] David Hirschfeld: And when you have that, you’ve got product market fit, you know, you have a viable business, then build the thing. And then people go, what? How do you sell your product before you have it? That’s what Launch First is all about.
[46:38–46:48] Host: Tip of the iceberg of your knowledge. We haven’t really uncovered all those gems that you can provide us. And you never know how the show can work. And we appreciate you, Margin, for coming on stage.
[46:48–47:00] Host: I really enjoyed having you on stage. If you’re listening to some podcasts, you can come to Clubhouse. Join us on Startup Club and listen to this live. Join us on stage. Share your expertise. David, this happened about two weeks ago.
[47:00–47:32] Host: Mimi is the editor-in-chief here at Startup Club. and every time we do an article it hits the top spot on google or the front page of google so if you have an article that you want to publish be great if you could be a contributor you’re not allowed to use ai except for research but great if you’d be a contributor and actually do an article about some of your expertise as well i’d love to club and you’ll get really good seo juice out of that i will absolutely do that one of the gentlemen who works on our team he does
[47:32–48:05] Host: the newsletter every month and i’m like this freaking newsletter is so good you’ve got to show your formula to the world and he wrote an article seven kick ass or seven ways your newsletter can kick ass i just googled how to create a great newsletter for startups came up number one on google that was just published a week ago anyone who’s in the audience or you’re listening to some podcast or on youtube you’re welcome to become a startup club contributor and there’s ways you can contribute. One is articles. The second is come on and do a show like David did. You can actually
[48:05–48:21] Host: even do your own shows, open up and become a leader on Clubhouse. I feel like there’s a strong community here on Clubhouse, on Startup Club, and we continue to produce good content. We have great community members, and the community comes together and makes such a difference here.
[48:21–48:30] Host: Thank you very much, David. You’ve been listening to Start, Scale, Exit, Repeat, Serial Entourage Secrets Revealed. We’ll see you all next Friday, 2 o’clock Eastern. Thank you.
David Hirschfeld founded Tekyz, a company dedicated to transforming business software development. With over 30 years of experience, his journey began with a physics degree from UCLA and a successful sales career at Computer Associates. After launching and selling his first software company in 2000, David found his passion for empowering entrepreneurs.
He developed the Launch 1st™ methodology, which focuses on generating revenue before coding begins. This helps startups gain traction while minimizing risks. With a commitment to innovation and collaboration, David leads Tekyz in providing AI-powered development and SaaS solutions, making a meaningful impact in the tech world.
Tekyz is set to launch two new AI applications: one for automating the Launch 1st Methodology Niche Analysis and Estimiz, an AI-based project estimation tool. Outside of work, David enjoys golfing and woodworking.
You can learn more about David Hirschfeld and Tekyz by following his LinkedIn profile — David Hirschfeld’s LinkedIn Profile.
For more information about Tekyz’s services and how they can help you harness the power of AI in healthcare, visit tekyz.com or contact the founder directly at [email protected].
AI-Driven Growth for Startups was originally published in Tekyz Blog on Medium, where people are continuing the conversation by highlighting and responding to this story.