331 | How to Find Your Highest-ROI AI Automations With Corey Ganim

0:00Isar Meitis:Hello and welcome to 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 everyone more or less on the planet right now wants to automate their business with AI. The biggest problem is that most people do not know where to start. And it is a really big problem because AI will do more or less anything. So you got to figure out, okay, what do I want to do first? How do I want to do this? What is the right steps to do it? And in today's episode, we're going to teach you exactly that, but we're also going to teach you the following steps, how to then actually automate the things and what is the process.

0:40Isar Meitis:Our guest today is Corey Ganim, and he is the founder of Return My Time, which is a company that helps other businesses understand well, as the name suggests, where they're spending too much time and they don't necessarily have to. and his sole focus is to help businesses, you, understand where you can save more time. Now, his process is broken into three different steps, audit, optimize, and automate. And as you can imagine, I'm very excited about all three steps. And also, as you can imagine, you can use AI in all these steps. Now, the cool thing is you don't actually have to use AI to automate everything.

1:18Isar Meitis:One of the things the process will teach you, and you'll learn exactly how in just a few minutes, is you may not need an AI solution. There might be an existing out-of-the-box solution that you can just use, and AI will help you find the right solution for you based on the audit that you're doing in the very first step. So as I mentioned, I work with a lot of organizations, and a lot of people just don't know what to automate or how to automate it, and having a framework that can walk you through all these different steps is very important. It is, in my eyes, more important than figuring out whether you want to use ChatriPity or Claude or Grok or whichever other tool you want to use, because the tools are all really good right now and having a good process to understand what you need to fix in your business, what is the right approach to do that is way more important than figuring out which tool is going to do it for you.

2:08Isar Meitis:They all will do a pretty good job in doing that. You can pick whichever one you want. And because of that, I'm very excited about today's episode. And so I want to welcome Corey to the show. Corey, welcome to Leveraging AI.

2:22Thank you so much for having me. Yeah, I'm excited to dive in today. I think there's quite a bit that we're going to cover when it comes to exactly what you mentioned, right? The framework around AOA, audit, optimize, automate. I think that's going to really simplify things for a lot of people when they think about using AI in their business and using it to take some of that repetitive work off their plate. So I'm excited to dive in.

2:42Isar Meitis:Yeah, listen, I think we're going from a stage in which people did not know what AI is. The next stage was like, oh, I'm starting to understand what it is. And I can start to understand that there's better prompting and less than prompting. And then the people go to the next steps like, okay, forget about the prompting. What is it that I'm trying to fix in my business? and I think it goes back to the very basics of just being a good leader and a good manager but being able to use AI to help you in that makes it very attractive. Exactly and to your point, right? Like half the battle is understanding the problem to be solved.

3:20I think a lot of people, whether it's an organization or whether it's a solopreneur, they think about AI, they kind of get a feel for what it can actually do and they just start throwing AI at all their problems. They just start trying to automate this thing and that thing. And they don't really spend the time to do step one in our framework, which like we mentioned earlier is audit. And really understand, you know, by doing a proper audit, what is it that we actually should be focusing our attention on? What is it that AI can actually be helping us with? And what are the things that AI shouldn't touch?

3:50I think that's another mistake people make is they try to use AI in parts of their business where AI has no business being it. Like knowing when not to use AI is arguably as important as knowing when to use it, in my opinion. So I'm sure we'll touch on all those points, but yeah, I think there's a lot of different directions we can go. That's for sure.

4:07Isar Meitis:Yeah, I love that. So really, let's get started. So if somebody is like, okay, I know I want to use AI. I don't know where in my business is the best place. And I think there are a very wide range of answers to that. So let's talk about the audit step. Let's understand your approach and let's understand how that could be applicable in different scenarios. Yeah, definitely. So when we talk about thinking about where AI can fit into our existing business, really at the end of the day, to take a step back, every business owner, they don't care about AI, right? AI is just the tool of the week or really the tool of probably the decade that's going to help us achieve one of three outcomes.

4:45Because at the end of the day, business owners only care about three outcomes. They either want to make more money, they want to save time, or they want to increase the quality of their product or service, period. And AI just so happens to be the best vehicle in 2026 and beyond to achieve one or all three of those outcomes. So really the idea becomes, what are the things that we're not doing in our business that we could be doing to, like I said, make more money, save time, or increase the quality of our product or service? Now, when it comes to the audit, so again, the audit is something that we sell as a service, and it's something that you could run on your own business.

5:20You don't need to be selling AI audits to make use of this framework. The idea being, you know, the simplest way to kind of explain the audit at its core is you're just kind of getting a brain dump or you're getting all of the existing bottlenecks that you have in your business down on paper, really on digital paper, and then letting an AI tool, whether it's Claude or whether it's ChatGPT, analyze that brain dump, analyze those pain points, analyze those bottlenecks, and then go out on the internet and find you some simple off-the-shelf tools that already exist. that can solve those problems. And that's what our, we call it the AI tools assessment.

5:58It's a very basic form of the audit is designed to do. It's like, let's just find some tools that already exist that can plug some of the holes that we have in our business that before AI, we didn't have the ability to go and find those tools because there's tens of thousands of AI and software tools out there. But now something like a Clodder, a ChatGPT can do that research and kind of prescribe the perfect tool stack for us and for our business. Does that make sense?

6:21Isar Meitis:This episode is brought to you by the Multi-Agent Orchestration Course. It's a course that I have been teaching personally since April of this year, and every single one of the cohorts that we have opened have sold out. And the people who have been through these courses made incredible transformations for themselves, switching jobs, starting their own companies, and so on, and definitely for their businesses by being able to build really complex processes by following our methodology and our infrastructure settings that we teach during the course. The course is four sessions of two to two and a half hours each over four consecutive weeks.

6:55Isar Meitis:And through it, you can go from knowing a little bit about how to use AI and using it in ways that most people are using it today to learning how to build really complex, sophisticated systems that can automate more or less everything in your business. The current cohort that is open starts on November 2nd, and it's going to go on November 2nd, November 8th, November 16th, and November 23rd. And we are almost sold out for that cohort as well. It is very likely that we will not open another cohort in December, meaning the next one after that is going to be in January. So if you still want to learn how to build really incredible automations and learn everything that I've learned by doing this every single day for the last few years, don't miss this opportunity to come and join us on November 2nd.

7:34Isar Meitis:There's a link in the show notes. And if you're going to use the promo code LeveragingAI100, you're going to get$100 off of the price that is on the website. Like I said, there's a few seats open, but they are selling out fast. So if you're interested in doing this still this year, go and click on the link right now on your phones and complete the registration and the payment so you can guarantee your seat. I'm looking forward to seeing you there. And now back to the episode. Yeah, I love it on several different levels. One, there used to be the old saying in the, you know, any projects that anybody run ever is like faster, better, cheaper, pick two.

8:09Isar Meitis:Right. I can do this faster and better, but it's going to cost you more money. Or I can make it cheaper, but it's not going to be as good and so on. And I think AI is the first time we can actually break that vicious cycle where you can actually do faster, better and cheaper all at the same time. So this is connecting to your previous point. As far as finding not AI tools, I'm 100 % with you. Like everybody wants to build these really cool AI automations. What people fail to remember is that somebody needs to A, build them. Okay, let's say you invested the time to do this. Somebody needs to maintain them, meaning these things break.

8:38Isar Meitis:They change over time. Your needs change. the underlying operating system. Microsoft comes up with a new thing and it's all going to stop working. ChatGPT or Claude comes out with a new version. It's all going to stop. So all these things requires maintenance, which means it's not, oh, I'm going to invest the time right now and then I'm done forever. It's never done. You have to keep on working on it. And so having an off-the-shelf tool that is going to cost you 20 bucks a month is way more valuable than developing something yourself that you will have to work on from now till forever because it has parents and you have somebody you can call when the thing stops working or when it has issues or when, again, the operating system changes and stuff like that.

9:18Isar Meitis:So when people today, like you're saying, you know, when you're a hammer, everything looks like a nail. Everybody's like, oh, let's solve everything with AI. No, there are awesome solutions that do not involve AI. And just like you said, there are things that AI is actually less good at than some existing tools and more expensive and less predictable. And so there are definitely cases where other tools are more relevant. let's let's dive in and really look at what this audit actually looks like like what kind of questions do i need to ask myself as a business owner or what is the process to really understand where do i need to start yeah absolutely and so i've got i know um i know most of the folks are probably listening on the podcast here listening to audio so for the people listening to this i'm happy to drop these i'm sure you'll add it maybe it's like a link in the description but we've got a google doc here of some prompts that i put together that you can you can feed these to clod or feed these to ChatGPT, and it'll basically perform the audit on your business.

10:12So the first prompt that we have here is like a self-interview, right? So this is where the AI is gonna play the consultant and actually turn the tables and interview you. And that is, and I know you use this capability because you're deep in the weeds, but this is something that a lot of people don't realize is that AI tools, particularly Claude, I've noticed is a lot better at this than ChatGPT, but they can interview you and ask you questions to get the additional context needed to do a really good job at whatever you asked it to do. And now I've actually made it a habit. This is kind of a side note.

10:44I made it a habit that whenever I ask AI to do something for me, I always end my prompt by saying, feel free to ask me any additional questions needed so that you can get the context that you need to do a really good job at this task. And sometimes Claude will ask me, it's like, well, hey, you know, did you want this as a Google Doc? Or did you want this as an Excel? Or, you know, hey, you said this, but did you mean this? Right. And, and so it's kind of my way of being able to fill in the gap so that Claude can give me a much better output. So when we do this self-interview prompt, that's basically what it's doing.

11:15It's just going to be asking us questions the whole time. So the, the prompt that I've got here, it's pretty robust. So I'm not going to read the whole thing verbatim, but the gist of it is, so you're going to act as an, and again, this is us talking to the AI, act as an AI consultant, running a discovery call on my business. My goal is to find the bottlenecks costing me money, time, or client experience, and where AI can fix them. About me, and again, that's where you're just going to brain dump your business type, what you sell, your revenue, your team, et cetera. And then you're going to say, ask me one question at a time and wait for my answer.

11:46And I want you to cover the one result that I'd want most in 90 days, again, which maps back to our three outcomes, more money, hours back for better experience, how leads come in, what happens in the first 48 hours, where deals stall or slip, a walkthrough of my typical day, the tasks that I dread, what only I can do and what breaks when I'm away, what clients complain about or ask repeatedly what tools I pay for, but underuse, right? So there's like another couple of paragraphs in regards to that prompt. And again, you guys can grab it at the link in the description in a copy paste format, but that is how you can get AI to self-interview you.

12:20And better yet, if somebody runs that, that self audit on themselves one time, it's going to take roughly 20 minutes to run through. They're going to get more value out of that one prompt than they've probably gotten out of their whole AI journey thus far because of the fact that the answers are going to be personalized and tailored to your exact business. I think that's another reason people struggle. The reason that AI struggle to get adoption among the mainstream is that a lot of the advice is generic, right? This is what helps the AI tailor the advice exactly to you and your business. And better yet, if you schedule this, schedule this prompt to run once a quarter so that you're basically re-auditing yourself once a quarter, every quarter you're going to get those tailored insights that quite honestly, you would have to pay a consultant a few grand, you know, to do pre-AI, at least if not more.

13:10Isar Meitis:I love it. One more thing related to, well, two more things related to the whole self-interview thing. One, if you are running a department, you can make small customizations to this, to each person in your department. If you're running the whole company, you can do the same thing for the different tiers of management and then the people underneath those and send every person a slightly customized version of this. And then you can take the answers and aggregate them into whatever aggregation you want. So you can have a company-wide priorities and not just individual priorities. so that's one thing that I want to say the other thing is going back to what Corey said is I have several different skills that are built to interview me more or less on everything and the core is exactly what Corey said the biggest trick is to tell it what's the goal like what is it trying to understand what is it trying to find and then encourage it to ask as many questions and contextualize between one and the other so it's not allowed to ask more than one question because you wanted to take the input that you gave it to understand what would be a better next question.

14:12Isar Meitis:And I don't limit it. A lot of people saying, oh, don't ask you more than three questions. Like, no, I want you to grill me. Like I want you to go for as long as you want. If it's this kind of thing, right? It's not always, but if it's a, I'm trying strategically to understand where I want to move my business, what are the key pain points and so on, go for as long as you want. Again, it costs me 20 bucks a month to do this. Or if you're like me, $200 a month instead of paying a consultant$700 an hour. So let it run, ask you as many questions as it needs. You're going to get crazy amount of value just from that.

14:45Isar Meitis:And again, if you do this as an aggregated thing, so let multiple people run through this in your organization, give all the answers back to one AI, you will learn even more because as leaders in an organization, we have a lot of blind spots to what's actually happening in the weeds in many different aspects. And this will allow you to float things that you're like, oh my God, I did not even know that it's such a big problem at this and that department. So these are my two cents. Now let's continue. So you go through the process. You get Claude to interview you. What happens next? Well, and one more tip that I'll give too, just before we move on to the next piece, what you could do is you could give this prompt to ChatGPT and actually talk to ChatGPT in voice mode.

15:26Now, Claude has a voice mode as well. So you could do the same thing in Claude. I just find ChatGPT's voice mode to be a lot more intuitive and quite frankly, just a lot easier to talk to. Again, you could do the same thing in Claude and it's probably easier and a better user experience to literally be able to talk to the AI and have a conversation about your business. You're probably going to get better results that way than you would if you're having to type out. Because naturally, if you're talking, you're just going to give more detail than you would if you're typing. And then you can take your voice conversation and you have a transcript of that inside of Claude or inside ChatGPT.

15:57and then you could take that transcript and then have it do one final analysis pass. So that's just like another tip, kind of another way to approach it.

16:04Isar Meitis:I agreed 100%. I don't type anything anywhere anymore. Like I either use the built-in voice or I use the built-in Mac transcript features. I find that it's good enough, but there's many third-party tools, whether Whisperflow or OpenWhisper and all of those that you can just run and it will transcribe everything. So including when you're writing emails and so on. I'll say one more thing related to what you said, because I think it's very important. one of the things you can do with your team is run this as a group exercise. So you can record a call, whether it's on Teams or Zoom or Google, Beat or wherever, wherever you want to record the call on.

16:39Isar Meitis:And you can just have a conversation with the team and say, okay, what are you struggling with? And so on. Kind of like follow the questions in the transcript that the AI is going to ask you, but ask it a few people, get their feedback and get their inputs, and just give the transcript to AI. and you will do an incredible job in distilling the golden nuggets from the conversation and putting it in a structure. Exactly. You're spot on. And so another, we've got kind of like a prompt B. So this is an option for people that if they want to skip the interview and they just want to kind of one shot it in a single prompt, this is a prompt that they can use to, again, skip the interview and basically just brain dump their week and brain dump some of the stuff that's already on their mind and get a lot of the same, just like bottleneck data.

17:21So this is where, again, this prompts a lot shorter. So I'll say the whole thing. So this is the prompt. You're going to say, here's a brain dump of how my business runs and what frustrates me. And then, you know, you're going to use something like a whisper flow or something like a voice transcription tool to just brain dump everything on your mind, right? Here's what I'm struggling with. Here's my bottlenecks. Here's where I spend my time. Here's what I hate doing. The more detail, the better. And you're going to include your business type, your revenue range, and your team size. And then you're going to end it by saying, find every bottleneck hiding in this, including the ones I didn't call out directly, build a table with bottleneck, estimated hours per week, outcome, right, revenue, time, or client experience, and what it's costing me, then list the three questions whose answers would most change this picture.

18:04And I love that ending. So that's where we're saying, we're basically asking the AI to continue to challenge us to say, you know, all right, at the end, ask me three questions that I can honestly answer that are going to push me to make a change and actually move the needle on some of the stuff that's currently holding me back. yeah very cool i i agree i think there is there are cases where you don't have to go through the

18:27Isar Meitis:full thing because you already know what's your immediate pain point like if you just like oh my god if i could spend instead of three hours a day is on emails i could spend an hour on emails i will do it right now because it's and you know you already know that's your biggest problem or if i could handle whatever so in those situations it definitely works very well to just go straight to the point, but still having structure, like you said, helps. Right, exactly. So then we've got two more prompts here to kind of round out our full self audit here. So the prompt C, which again, we just went through, basically, you could choose A or B, the two that we just gave you, whether you want to do the full structured interview, or whether you just want to do a brain dump.

19:04And then prompt C is this is when we get the AI to both score and prioritize our bottlenecks on a matrix of what we call effort versus impact. So that's where the AI is going to look at all the things that we said. It's going to identify the biggest pain points. Then it's going to find the ones that are both relatively easy to fix, but also provide a high impact. So that's what we call like high impact, low effort. We would call that a quick win. Moves the needle, but it's not a huge pain to fix. So the prompt is score each bottleneck above. Impact, you know, that's where we kind of, and I'm not going to run through, but like that's where we give it the criteria for, you know, a nine to 10 impact means critical to revenue or it takes me five plus hours per week right and so it kind of lays out that full criteria which i won't go into but then we say sort them into quadrants so a quick win is anything with an impact score of seven or above and an effort score of one to six major projects is anything with an impact score of seven or above but an effort score of seven or above a fill-in is anything with an impact score of six or below when an effort score of one to six, and then a skip, meaning it's not worth our time because it's high effort, low impact, is something that has an impact of six or below and an effort of seven plus, right?

20:20And again, the numbers aren't as relevant. The whole idea is with this prompt, we're just getting it to give us the lowest hanging fruit opportunities. That's pretty much what it comes down to. And then to round out the prompt, we're telling it for each quick win, recommend one fix, use a dedicated off the shelf tool first, an automation, meaning like a Claude Skill, second, and a general chatbot only if nothing else better exists. Search for current pricing, include setup time, and a conservative estimate of hours saved. Finish with ROI, meaning hours back per month, times my hourly rate, and you'll give it your hourly rate, minus the tool cost, and then the one fix that I should do first and why.

21:00So that is where you get it to score your bottlenecks and then prioritize exactly where you should focus your time. Does that make sense?

21:08Isar Meitis:Yeah, absolutely awesome. I want to recap a few things because there was a lot in there. The first thing is you need some kind of a scoring matrix. And that's basically what Corey is saying. I came from the software world. In the software world, we always look at value to the clients. So whatever that does to whoever the client is. And then the second parameter was effort, which is exactly what Corey is saying. And we measured effort in t-shirt sizes. That's very common in the software space. so extra small small medium large extra large extra extra large and you can give numbers to each and every one of those once so you kind of know what that is and then the ai knows what it is right so extra small is anything you can do in less than an hour small is anything i can do in a few hours and so on and so forth and you know extra extra large is okay we'll take three weeks or more to develop this so it's a major project sometimes it's worth it if it's going to completely transform your business but then you can at least have the right uh mix of immediate wins that you can do very quickly versus large projects that run in the background that are going to take longer and will be more transformative.

22:08Isar Meitis:So this is one aspect. You got to quantify the qualifiable. So you're going from unstructured data, which is you just brain dumping and answering questions to a structured data where, okay, now I can decide priorities and put value to different things. So this is one aspect. And the second aspect is really the recommendations. and the recommendation that Corey is suggesting is third-party existing tools, then build your own automations, then just use free chat as the levels. You can build your own priorities. Whatever works for you is perfectly fine. To me, it makes perfect sense, especially once you start doing the last component, which I want to talk about, is scoring this based on actual ROI.

22:51Isar Meitis:So how do you calculate your ROI? Well, Corey just told you how he does his, but you have other things. Like what are the things that are going to cost you money? what are the things that are going to make you money or time and then multiplied by what's your time is worth and and i always tell people that usually your time is worth a lot more than you think meaning even if you're running on quote-unquote billable hours and saying oh i'm billing people for 200 hours or a thousand dollars or it doesn't really matter your time usually is worth more than that and the reason i'm saying that is because this takes into account just the time you're working on that particular client.

23:26Isar Meitis:It doesn't take all the overhead you have to do to make the company work or your job be fulfilled. And so to do that, you probably need to take into account more things. So it's an exercise on its own to figure out what's your hour is really worth or what it should be worth when you're pricing it this way to know, oh, this is going to save me an hour a week. Oh, that doesn't sound like a lot. Well, depends on what you do with that hour that you get back. Right. Well, I think that's a really good point to make too, especially for, yeah, for the people that do like client hourly work, like you said, I think the default is to think like, oh, my billable rate is$250 an hour.