Bản chép lời đầy đủ của Stop Prompt Engineering. Let AI Interview You Instead. | Craig Mooney

uh wanting to understand safety and how to protect yourself around safety. 00:02:25.840 --> 00:02:28.240 Strongly recommend joining his coalition. 00:02:28.479 --> 00:02:30.319 Secondly, Dr. 00:02:30.479 --> 00:02:47.199 Jeffrey McGee, the first guest on the Scale From Here podcast, he is um his episode was October 6th, and he is putting together a two-day performance-driven sales selling boot camp on December 2nd and 3rd in Houston, here in Houston. 00:02:47.439 --> 00:02:56.080 So if you're in the Houston area and you want to individually go or take your team to learn how to sell, go check out his boot camp. 00:02:56.159 --> 00:02:59.599 We'll put this link in the in the uh show notes as well. 00:02:59.840 --> 00:03:13.199 But it's salesboot camp.jeffrey, J E F R E Y, McGee, M-A-G-E-E dot com backslash, and you can sign up for his boot camp there. 00:03:13.280 --> 00:03:18.479 I mean, this guy's acclaimed, he's been doing these for a long time, and he does a great job. 00:03:19.120 --> 00:03:26.800 Uh, two shows ago, we had Greg Herrod on, and uh we were talking about um strategic communications. 00:03:27.199 --> 00:03:30.080 You're not a control freak, you just never learn how to delegate. 00:03:30.240 --> 00:03:30.719 Great show. 00:03:30.800 --> 00:03:34.719 We're getting great feedback on that show, so definitely go check out Greg's show. 00:03:34.879 --> 00:03:42.639 And on October 27th, we had Leslie Vinett's, and we again we're talking about AI made outbound easy. 00:03:42.800 --> 00:03:44.960 That's exactly why yours isn't working. 00:03:45.039 --> 00:03:47.360 So, this is going to be a great follow-up show. 00:03:47.599 --> 00:03:54.159 She was talking a lot about using AI to make generic messaging and how it's just not working and how noisy it is out there. 00:03:54.319 --> 00:04:07.840 So, if you're looking for uh outbound messaging and marketing messaging using AI and you want to learn the do's and don'ts, definitely go back and check out our October 27th show with Leslie Venetz. 00:04:08.319 --> 00:04:12.479 And then continuing on today, uh, we have Craig Mooney on the show today. 00:04:12.560 --> 00:04:26.800 And Craig is the founder of Auto Build Solutions, where uh they're helping component manufacturers and LBM suppliers escape reaction mode with automated connected systems that improve visibility and reduce manual work. 00:04:27.040 --> 00:04:38.000 With 10 years in building materials and trust sales, he uses the race framework to modernize workflows, implement AI, and reduce reliance on legacy software. 00:04:38.160 --> 00:04:46.480 Craig also led uh BCM 2026 session on building AI agents for the component manufacturing project lifestyle. 00:04:46.639 --> 00:04:47.680 Whew, mouthful. 00:04:47.759 --> 00:04:50.079 All right, let's bring Craig on to the show. 00:04:51.360 --> 00:04:52.399 How are you doing, man? 00:04:52.720 --> 00:04:53.360 Good, good. 00:04:53.439 --> 00:04:55.920 Thanks for the warm intro and the yeah. 00:04:56.079 --> 00:04:58.160 Sorry I made it long-winded for you. 00:04:58.399 --> 00:04:59.120 No, you're good. 00:04:59.279 --> 00:05:02.959 Hey, whenever I love when people they say, Can you send me a bio? 00:05:03.040 --> 00:05:05.680 I'm like, oh yeah, you want a paragraph of me bragging about myself? 00:05:05.759 --> 00:05:06.079 No problem. 00:05:06.319 --> 00:05:06.720 Right, right. 00:05:06.800 --> 00:05:10.399 Let me just have AI whip out a uh four-paragraph intro for you. 00:05:11.199 --> 00:05:12.480 Yeah, four paragraphs. 00:05:12.560 --> 00:05:13.600 I'd still be reading it. 00:05:13.759 --> 00:05:15.920 As you can tell, it's not my strong suit. 00:05:16.319 --> 00:05:19.439 So where you are sitting, man, where are you again? 00:05:19.759 --> 00:05:21.279 I'm in uh Oregon. 00:05:21.600 --> 00:05:22.240 Oregon. 00:05:22.800 --> 00:05:24.079 Oregon, not Oregon C. 00:05:24.399 --> 00:05:28.000 Yeah, some some call it Oregon and others are Oregon. 00:05:28.720 --> 00:05:29.360 Oregon. 00:05:29.600 --> 00:05:33.759 Yeah, O-R-Y-G-U-N is how uh how we pronounce it. 00:05:34.079 --> 00:05:36.639 So are you an Oreg Oregander? 00:05:37.519 --> 00:05:38.959 Uh Oregonian. 00:05:39.279 --> 00:05:40.319 Oregonian? 00:05:40.560 --> 00:05:42.800 Yeah, make it even make it even longer. 00:05:43.839 --> 00:05:44.959 That's amazing. 00:05:45.199 --> 00:05:48.959 Well, we're you know, AI is a thing, obviously. 00:05:49.120 --> 00:05:50.000 Everyone's into it. 00:05:50.079 --> 00:05:58.720 And when we first had a chance to connect, I was really liked the way you broke it down and your race methodology. 00:05:59.439 --> 00:06:01.759 Uh, we teased it a bit in the intro. 00:06:02.000 --> 00:06:12.319 Give us a high-level overview of of what the race methodology is when you're implementing, and this this refers to implementing AI into the sales process, correct? 00:06:12.639 --> 00:06:13.199 Correct, yeah. 00:06:13.279 --> 00:06:16.079 So I call it a self-driving sales system. 00:06:16.639 --> 00:06:32.639 And uh really what it is is uh the race implementation plan or found uh found uh uh framework is really review your work, review your workflows, architect what it would look like automated, and then confidently launch. 00:06:32.800 --> 00:06:39.040 And uh so that's the RAC, and then E is enhance or um evolve. 00:06:39.120 --> 00:06:42.800 So AI is constantly evolving, so that's where the acronym comes from. 00:06:43.120 --> 00:06:55.680 But the first thing when you're reviewing is see where you're at in the AI journey, and that's where we have the five levels of your AI journey, and we're gonna stick with the vehicle metaphor. 00:06:55.839 --> 00:07:09.519 So um, and that's where we have the uh the wagon, the the Model T, your daily driver, um, your uh race car or your sports car, excuse me, and then your self-driving race car. 00:07:09.600 --> 00:07:12.800 And so we can jump into the each stage of the journey. 00:07:13.040 --> 00:07:13.360 Yeah.

00:07:13.600 --> 00:07:17.360 So the the first one being the uh the wagon. 00:07:17.519 --> 00:07:25.360 I'm from Oregon, so really like the Oregon Trail, you know, that that that wagon image of just that long trek and manual labor. 00:07:25.439 --> 00:07:27.519 That's that's where you're not using AI. 00:07:27.600 --> 00:07:31.279 You're in a race, you're you're racing against race cars and you're in a wagon. 00:07:31.360 --> 00:07:32.240 That's just not gonna work. 00:07:32.319 --> 00:07:34.720 It's not sustainable in today's day. 00:07:35.120 --> 00:07:44.959 So in the in the the what's the uh curve of the that you have the early adapters and all that you would be you would be on the laggard side of things. 00:07:45.120 --> 00:07:50.160 You're you know, in the world of cell phones, you're still using a rotary phone with the cord and the wall. 00:07:50.399 --> 00:07:51.680 Yeah, yeah, maybe a fax machine. 00:07:51.920 --> 00:07:53.839 Anyone even knows what the hell that is anymore. 00:07:54.079 --> 00:07:54.399 Right. 00:07:54.480 --> 00:07:54.800 Yeah. 00:07:55.040 --> 00:07:55.680 No, yeah. 00:07:55.759 --> 00:07:57.439 I it went right over my head. 00:07:57.680 --> 00:07:58.000 No. 00:07:59.759 --> 00:08:02.319 But I mean, there's still people using fax machines. 00:08:02.399 --> 00:08:16.240 I guess some you know, some instances it makes sense, but uh, I mean, we're we're kind of at that at that stage now that we were 27 years ago or whatever, uh, when email started becoming more popular, right? 00:08:16.399 --> 00:08:19.600 Either you either use email or you get stuck in the fax era. 00:08:19.759 --> 00:08:24.000 So that's kind of where we're at, and that's where I see the wagon stage. 00:08:24.079 --> 00:08:28.160 You're you're resisting the email and sticking with fax. 00:08:28.399 --> 00:08:28.720 Right. 00:08:28.879 --> 00:08:34.480 Um so then the next stage is really you're um testing it out. 00:08:34.559 --> 00:08:37.200 You're you're searching with AI, and that's where a lot of people are. 00:08:37.279 --> 00:08:39.039 It's it's a search engine for them. 00:08:39.200 --> 00:08:42.559 And heck, I use it for uh my fantasy football team. 00:08:42.720 --> 00:08:50.559 Didn't really do well, but you know, a lot of people might use it for that, or barbecue recipes, or you know, what's wrong with me questions, whatever it is, right? 00:08:50.720 --> 00:08:53.039 Yeah, and uh WebMD. 00:08:53.360 --> 00:08:54.399 Yep, exactly. 00:08:54.559 --> 00:08:59.600 Um, I think I searched like one of the most common things people search, and it is it's like WebMD stuff, which makes sense. 00:08:59.679 --> 00:09:01.759 It's got a plethora of knowledge. 00:09:02.000 --> 00:09:06.320 Um, but you're still not using it to its fullest capability. 00:09:06.480 --> 00:09:09.919 And in that bell curve you were talking about, that's where we hit the daily driver. 00:09:10.000 --> 00:09:14.799 That's like your commuter uh car, the Toyota, the one that's like you know, trusty. 00:09:14.960 --> 00:09:18.080 You're not you're comfortable in it, but it's not gonna win the race. 00:09:18.240 --> 00:09:18.559 Right. 00:09:18.720 --> 00:09:18.960 Yeah. 00:09:19.200 --> 00:09:35.279 And and that's connecting with your email, your spreadsheets, and uh, you know, using the MCP connectors available to you to help write emails and you know, use that as an AI assistant to to summarize your emails, that kind of stuff, PDFs. 00:09:35.440 --> 00:09:35.759 Okay. 00:09:36.000 --> 00:09:46.639 Um and so that is 72% of uh all businesses are using AI, and I that's where the majority of them land right there. 00:09:47.120 --> 00:10:00.720 So taking it to the next level, we're starting to see people actually build with AI, building artifacts, tools that their teams can use, estimating tools, um, even you know, small CRM type of solutions. 00:10:01.039 --> 00:10:04.559 And so that's that's where you're in a race car situation. 00:10:04.720 --> 00:10:10.960 You're building something that's actually usable, might be a little too much power for your technical capabilities, right? 00:10:11.200 --> 00:10:14.320 Um, but you can really go fast with it and scale. 00:10:14.399 --> 00:10:19.840 But there's there's obviously some security risks, just like with race cars or a sports car. 00:10:20.000 --> 00:10:24.559 You know, it's like if you don't know how to drive it, you can you can take that thing the wrong direction. 00:10:24.879 --> 00:10:37.600 And uh there's also the challenge if you have um IP right and you're feeding into a non, what is it, gated or or guardrail? 00:10:38.639 --> 00:10:45.679 Yeah, an AI that has online access, it becomes common domain or whatever that is out there. 00:10:45.840 --> 00:10:49.840 And if people can access it, not that they can access it, but now it's out there, right? 00:10:49.919 --> 00:10:56.159 And it's using that to reference and and create other documents that people are calling upon. 00:10:56.320 --> 00:11:00.240 So you get into a little tricky uh some tricky things there. 00:11:00.320 --> 00:11:02.000 I think there was a legal case way back. 00:11:02.159 --> 00:11:15.840 A guy got um AI to proofread as legal documents, like the lawyers had sent them over an agreement or whatever, and it was confidential, but he fed it into AI and they were able to go back and retrieve it, and they're it's not confidential anymore. 00:11:15.919 --> 00:11:16.720 It's right. 00:11:17.759 --> 00:11:18.320 Exactly. 00:11:18.639 --> 00:11:28.000 And so what you actually that's a great point because what you're seeing now is is these businesses either don't have an AI handbook or playbook for their company. 00:11:28.080 --> 00:11:43.440 So people are using their personal accounts with company data, with company logins, and you know, you're making those connections, and now like your your business, you don't even realize that you're giving away all this data and logins to an AI model. 00:11:43.679 -->

00:11:49.519 Uh you don't know if the person, the your employee is um giving it full access or what that setting is. 00:11:49.679 --> 00:11:58.879 So it's really important for that first step for these businesses to really um harness that into uh controlled environment using team plans, business plans, and whatnot. 00:11:58.960 --> 00:12:00.639 Because I mean it's happening everywhere. 00:12:00.720 --> 00:12:11.360 Uh so it's uh like you said, put the guardrails up, put it in a controlled environment, and put your AI handbook together sooner than later. 00:12:11.679 --> 00:12:12.159 Right. 00:12:12.480 --> 00:12:12.799 Okay. 00:12:13.440 --> 00:12:15.440 And then so it was a race car. 00:12:15.679 --> 00:12:17.200 So yeah, it's a sports car. 00:12:17.279 --> 00:12:18.240 Um sports car. 00:12:18.480 --> 00:12:22.960 I guess yeah, it's similar, but you know, most people have more access to sports cars than race cars. 00:12:23.200 --> 00:12:31.679 And so the the last one is a self-driving race car, and that's where it is like uh you have AI embedded in your systems. 00:12:31.759 --> 00:12:34.720 It it understands your workflows, it understands your habits. 00:12:34.879 --> 00:12:39.039 You have these uh, you know, you've heard skills um written up. 00:12:39.200 --> 00:12:46.000 So it gives you those guardrails and it also gives it the direction to be able to work while you're not working. 00:12:46.159 --> 00:12:51.200 Um that's that's where give me an example of what that might be, if you can, Craig. 00:12:51.519 --> 00:13:03.679 So that would be like, let's say we have automations that connect with your inbox, you're looking for a purchase order or uh an I uh uh uh ITB, an intent to bid or bid request. 00:13:03.840 --> 00:13:10.639 Something that that is gonna trigger an agent to to push to an agent and say, all right, now you deal with it from here. 00:13:10.879 --> 00:13:18.720 And so what you what you're what you need to do is train your agents, your AI system to know what to do next. 00:13:18.960 --> 00:13:36.159 And for instance, you let's say you get a uh that bid request, it knows to find the contact in your CRM, the company it's associated with, the buying habits of that company, uh maybe maybe rank this bid request and say, hey, this is a high priority item, let's get it in the estimating schedule quicker. 00:13:36.399 --> 00:13:49.279 Um, but if maybe the the agent sees that it's a low um uh maybe a customer that's not ranked so high, they don't purchase frequently from you or whatever, it might not take it as a priority. 00:13:49.440 --> 00:13:56.399 So what it's doing is making these decisions, putting in your CRM, and then scheduling and monitoring it from there. 00:13:56.559 --> 00:14:10.480 So you can really have it go from bid intake all the way to order and delivery, managing the uh the schedules, coordinating between different schedules, and making sure it communicates with your customer for you. 00:14:10.879 --> 00:14:13.919 So this almost sounds like inbound sales to me. 00:14:14.080 --> 00:14:15.200 Am I right on that? 00:14:15.600 --> 00:14:18.639 It can be inbound or outbound as well. 00:14:18.960 --> 00:14:38.240 You can use it for outbound as well, in terms of you know, if you have a CRM and you have a specific cadence um of reaching out to people or uh you know specific type of uh target that you're going after, your ICP, it can weigh all of these uh these these trends and and data. 00:14:38.399 --> 00:14:42.240 As long as you have data, it can it can help you with anything, essentially. 00:14:42.799 --> 00:14:44.480 I like the inbound idea. 00:14:44.639 --> 00:15:20.720 If someone sends an email saying, hey, we'd love you to put together a bid for this project we're working on, and it can pull that out of your inbox, analyze that client, look at what they bought before, probably if they can get into your system and say, okay, the last three times we quoted these guys at a 12% uh gross margin, throw together a quote, and then it lands in you or your director of sales inbox saying, This quote was requested today at noon. 00:15:20.879 --> 00:15:22.559 Here we are at 2 p.m. 00:15:23.039 --> 00:15:30.639 Here's the quote, review it, make sure we got everything right, let me know what changes you need, and then we can send it on. 00:15:31.200 --> 00:15:32.399 Yep, exactly. 00:15:32.720 --> 00:15:37.840 And there's there are certain settings like you're saying, with uh let me know what you need or whatever. 00:15:38.080 --> 00:15:49.039 You can have your agents never make decisions for you or ask for permission before it makes a decision, or you can just have it set to you know make the decision throughout the whole process or whatnot. 00:15:49.200 --> 00:15:57.120 Um, so in terms of inbound and outbound, I start with inbound because it's such an administrative role typically. 00:15:57.600 --> 00:16:01.039 It feels a little safer because it's more transactional. 00:16:01.120 --> 00:16:03.279 You're there's not as many people involved. 00:16:03.600 --> 00:16:08.399 Yeah, I'm very much a proponent of keeping your outbound personal. 00:16:08.639 --> 00:16:14.159 I am, you know, it's it's getting it's it's a little easy to identify AI messages. 00:16:14.320 --> 00:16:19.360 Um, I think for prospecting, gathering the the people that you want to reach out to. 00:16:19.519 --> 00:16:35.039 If you give your your AI assistant or agent a uh you know an ICP profile, um your target market, specific things that you need, and have it get that list together, find the email addresses and and maybe craft the message for you. 00:16:35.200 --> 00:16:46.080 But I still think the I think your personality and in your relationship is gonna win more than ever now with with AI involved because people are wanting that connection. 00:16:46.639 --> 00:16:53.440 Outbound sales, I feel like there's so much freaking noise out there. 00:16:53.600 --> 00:16:58.240 Yeah, it's gonna go restaurants are

about to get really busy for lunch, man. 00:16:58.320 --> 00:17:02.720 Like people are just just let's just meet. 00:17:02.879 --> 00:17:14.000 I I don't need you to send me the six email sequence with all the catchy hooks at the end and um all the psychological warfare that goes into those things. 00:17:14.079 --> 00:17:24.799 Like, let's just sit down and it it's weird how all this automation's gonna might push take us from being, you know, during COVID, we never even meet anymore. 00:17:25.039 --> 00:17:25.440 Right. 00:17:25.680 --> 00:17:28.799 And now I think it's gonna push us into more face-to-face meetings. 00:17:28.880 --> 00:17:33.920 I think people are getting um, you know, team meeting, video meeting fatigue as well. 00:17:34.240 --> 00:17:35.200 Yeah, exactly. 00:17:35.279 --> 00:17:45.279 I think there's there's gonna be more relationship building and and having that patience because I think AI messages are trying to close the deal from the very beginning. 00:17:45.359 --> 00:17:52.799 Like, even the you've seen it, those staged um, you know, six messages drip that they're that they're doing these campaigns. 00:17:52.880 --> 00:17:54.480 It's like, stop trying to close me. 00:17:54.559 --> 00:17:56.240 Like, let's develop a relationship. 00:17:56.400 --> 00:18:05.119 Like, send me something cool about my company that nobody else would know, or um, you know, a video of me introducing myself. 00:18:05.279 --> 00:18:17.359 That's what that's what I'm seeing uh and hearing right now is like LinkedIn DMs video are actually being opened more than you know than text because you're having that. 00:18:17.519 --> 00:18:20.559 Oh, this isn't an AI bot, this is a real person talking to me. 00:18:20.880 --> 00:18:21.519 Interesting. 00:18:21.759 --> 00:18:39.359 Yeah, but there so that's a tricky one because I think Videard has a a you can clone yourself and it can send video responses to emails that you don't it creates it and sends it. 00:18:39.599 --> 00:18:40.720 That's crazy, yeah. 00:18:40.960 --> 00:18:43.359 It's not, but you can tell it right now. 00:18:43.440 --> 00:18:47.200 You can still tell AI videos are just not quite there. 00:18:47.519 --> 00:18:49.839 Give it a year though, and yeah, we'll see. 00:18:50.480 --> 00:18:51.440 Crazy, right? 00:18:51.599 --> 00:18:51.920 Yeah. 00:18:52.160 --> 00:18:57.759 Well, let's let's pull this into to the uh inbound side of things. 00:18:58.000 --> 00:19:06.799 So if we wanted to let's talk about the sports car and the autumn, the self-driving race car. 00:19:07.839 --> 00:19:19.599 In inbound, what are some of the, you know, I think you said it was 71 or 79 percent of business owners are using it, and you think it was in that kind of um, actually, I thought you thought it was you said it was daily driver, right? 00:19:19.680 --> 00:19:25.119 That's where more people are sitting, kind of that using it as search, using it for a little research. 00:19:25.200 --> 00:19:26.319 Hey, can you find

me this? 00:19:26.480 --> 00:19:27.759 Can you find me that? 00:19:28.720 --> 00:19:33.920 As it pertains to inbound in your in your kind of experience, what's that next progression? 00:19:34.079 --> 00:19:41.119 Give, give you what are some like tangible things I could go, oh, I could do that with my sales process. 00:19:41.200 --> 00:19:42.960 What could I do? 00:19:43.359 --> 00:19:52.799 What I would do is start talking with uh, let's say you use Claude or Chat GPT and start saying, here's how my process is, and interview me. 00:19:52.960 --> 00:19:56.720 This is what I think a lot of people miss is like they try to prompt engineer things. 00:19:56.799 --> 00:20:02.160 And and you know, we were kind of taught for the last year to write the best clearest prompt. 00:20:02.640 --> 00:20:09.759 But now I see it as more like interview me in how I approach my business, how I approach this specific use case. 00:20:09.920 --> 00:20:18.880 And then Claude or ChatGPT will give you a list of uh questions that you go through and learn from that, and then continue to ask questions. 00:20:18.960 --> 00:20:24.240 So you can say, hey, I'm trying to automate my uh bid intake uh process. 00:20:24.480 --> 00:20:30.000 Here's what I sell, here's how my business operates, ask me questions so we can make this happen. 00:20:30.240 --> 00:20:34.880 And so it will, I mean, it's it's pretty easy to do, and just just answering those questions. 00:20:34.960 --> 00:20:45.279 So then now you have a database, a knowledge base for your AI agents or assistants to actually make decisions for you or help with these decisions. 00:20:45.440 --> 00:20:50.559 So what you really need is to harness what they're calling is a harness of this data. 00:20:50.720 --> 00:20:55.359 Um, and again, that a lot of that has to do with what's in your head and how things work. 00:20:55.599 --> 00:21:04.720 Put it down in paper, meaning make sure your AI has the data of your decision making uh uh processes and uh it'll take it from there. 00:21:04.880 --> 00:21:13.279 That's that's I think the first step is make sure that your workflow is clearly identified with your AI, your LLM of choice. 00:21:14.000 --> 00:21:22.960 Could you tie if if it were if it's allowed your your software or whatever type of software you're using? 00:21:23.200 --> 00:21:28.079 I know I'm my AI is tied into HubSpot, for instance, right? 00:21:28.480 --> 00:21:38.400 Could I tie it into my ERP or whatever it is and say, hey, go in and find my sales flow, my inbound sales flow. 00:21:38.640 --> 00:21:41.440 And it's this is what it's labeled or called. 00:21:41.599 --> 00:21:43.200 Analyze that for me. 00:21:43.359 --> 00:21:46.000 You know, pull it out, get all the information you want. 00:21:46.079 --> 00:21:52.559 And could you could you start there, or is that more in the advanced side of things now where we're connecting softwares? 00:21:52.880 --> 00:21:54.319 No, I think that's right there. 00:21:54.480 --> 00:21:59.440 That's right at going from uh daily driver to the sports car. 00:22:00.079 --> 00:22:00.880 Now you're making those. 00:22:01.200 --> 00:22:02.319 So I'm not very advanced. 00:22:02.480 --> 00:22:04.160 I'm just like, yeah. 00:22:04.480 --> 00:22:16.640 No, I mean, again, most of uh most everyone is at the just starting to connect with your spreadsheets, that kind of stuff, and say, hey, you know, maybe build out this spreadsheet for me or tell me what's going on in the spreadsheet. 00:22:16.799 --> 00:22:28.480 So if you're connecting with HubSpot, your CRM, and you're connecting with an ERP, the next step is like, all right, now build me a dashboard so I can analyze XYZ. 00:22:28.559 --> 00:22:31.200 Here are the KPIs I want, and building that dashboard. 00:22:31.279 --> 00:22:43.039 And so you actually have what Claude calls an artifact that you can that you can uh refer to regularly instead of relying on say HubSpot to build you your reports. 00:22:43.839 --> 00:22:44.559 Very cool. 00:22:44.799 --> 00:22:46.240 So we're gonna go in. 00:22:46.400 --> 00:22:59.839 Maybe the first step is we're gonna connect our AI with our software, our ERP or CSR, whatever we want to call it, and get it to do an to get it to analyze. 00:23:01.599 --> 00:23:06.720 Before so here's AI taught me this. 00:23:07.359 --> 00:23:13.359 Um when I started put you know re rebranding the podcast and things. 00:23:13.839 --> 00:23:26.079 I had done so much research and asked it so many questions and and brainstorm so much with it, it actually said to me, Would you like me to put together a brand document for you? 00:23:26.559 --> 00:23:26.799 Yeah. 00:23:27.039 --> 00:23:28.319 And I was like, what is that? 00:23:28.480 --> 00:23:30.960 And it's like, we'll just capture all of this information. 00:23:31.119 --> 00:23:36.480 And it gave me, we decided what fonts we're gonna use, the color scheme, the everything. 00:23:36.559 --> 00:23:40.000 It ended up being like a 10 or a 12 page document, right? 00:23:40.160 --> 00:23:47.119 Yeah, and I'm just thinking for a business owner, it and it'll scan the like we I didn't have the website up yet. 00:23:47.200 --> 00:23:52.960 We're we're building the website, but it'll it'll say, give me your URL and I'll I'll pull all the information from that. 00:23:53.039 --> 00:24:14.799 So as a business owner, before you start this, is there value in creating you know that brand document that maybe includes your core values, your key customers, your your top customers, your average sale, your average, like just as much information as you can put in there and tell it like analyze and learn this document. 00:24:16.000 --> 00:24:16.480 Right. 00:24:16.640 --> 00:24:17.200 Exactly. 00:24:17.359 --> 00:24:20.720 It's here's all of my resources that you need to refer to. 00:24:20.880 --> 00:24:22.480 This is how my business operates. 00:24:22.640 --> 00:24:29.119 This is like you said, my mission statement, or whatever, whatever's

important to your business and your customers, more importantly. 00:24:29.599 --> 00:24:34.400 And that like the brand kit is a great idea or a great uh use case. 00:24:34.559 --> 00:24:38.400 It's it's simple because most people have a website that it can refer to. 00:24:38.559 --> 00:24:51.359 And so that's what's important is uh what uh again with Claude, it's a Claude MD file, which is basically the the Bible essentially of what your AI assistant or agent can refer to. 00:24:51.680 --> 00:24:58.720 And within that are called skills, where you know it's specific to one agent or specific to another workflow. 00:24:58.799 --> 00:25:04.960 So you can build like your brand kit skills of you know, we want to have these kinds of colors all the time, these fonts. 00:25:05.200 --> 00:25:18.160 And uh again, you can you can turn that into a different process, not just branding, but here's our rules for how we follow up or how we how we close deals, right? 00:25:18.480 --> 00:25:19.920 Yeah, that's exactly what I was saying. 00:25:20.000 --> 00:25:25.680 Like once you do this brand kit, you put everything in, then you get it to analyze your inbound process. 00:25:25.839 --> 00:25:37.279 And then once you kind of flush it out and say, okay, here's what it is, okay, now add it to the brand kit for me, or create me a new inbound sales brand kit, almost like, hey, you're new here, read this, right? 00:25:37.359 --> 00:25:38.640 This has everything in it. 00:25:38.960 --> 00:25:39.279 Right. 00:25:39.440 --> 00:25:50.640 And and it also uh with that brand kit, like if you say, hey, write me up a document, a proposal, and it's going to build that proposal with your brand, with your, you know, with all that stuff. 00:25:50.720 --> 00:25:52.400 So, like you said, it's uh yeah. 00:25:52.480 --> 00:26:00.400 So now when it gets to that end thing, so so yeah, I don't want to stop at okay, we've linked everything. 00:26:01.200 --> 00:26:04.160 Now, how could like how could it look? 00:26:04.319 --> 00:26:12.079 An email could come in and you could have it trained to recognize an email that looks for the word proposal or bid or whatever. 00:26:12.400 --> 00:26:12.880 A job number. 00:26:14.160 --> 00:26:28.880 A job number, it reads that email, looks at the requirements, then goes into back into the ERP because we're linked, starts researching have we done business with this person before, looking at the margins maybe you've done or the costing you've done in the past with them. 00:26:29.119 --> 00:26:40.079 But you said I think in the beginning you can set it up to either say, hey, you have a new bid, or it could ping you and say, You got a bid, here it is, I prepared it for you. 00:26:40.319 --> 00:26:44.880 Look at it, or a bid came in, what do you want me to do? 00:26:45.599 --> 00:26:47.839 So now we have everything linked. 00:26:47.920 --> 00:26:54.160 How can we start push pushing this all together to really create some efficiencies? 00:26:54.559 --> 00:26:59.440 That's where I think starting small, you have that one use case and build off of it. 00:26:59.519 --> 00:27:04.880 So, really kind of hone in that one use case, usually in the front end, like we're like we're discussing now. 00:27:05.119 --> 00:27:07.200 And then, all right, what happens after that? 00:27:07.359 --> 00:27:09.440 Or refine it, you know, watch it. 00:27:09.839 --> 00:27:14.240 You can actually even build self-learning uh AI agents. 00:27:14.319 --> 00:27:19.920 So they're gonna learn if you're if you're constantly saying, well, this doesn't work, you need to adjust it like that. 00:27:20.160 --> 00:27:24.799 It needs to learn that it got it wrong that time and improve then. 00:27:24.880 --> 00:27:27.519 So the next time it's more consistent output. 00:27:27.920 --> 00:27:32.799 Um, so that's really that next level is the self-learning AI, is what they call it. 00:27:33.119 --> 00:27:34.960 And I think you touch on a good point. 00:27:35.039 --> 00:27:41.599 A lot of people try this the first time the first quote comes out and it's not close, and they're like, well, this doesn't work. 00:27:41.839 --> 00:27:43.519 Yeah, AI is terrible, right? 00:27:43.839 --> 00:27:49.519 Yeah, but you have to go, that's where you have to keep talking to it and saying, okay, next time do this. 00:27:49.599 --> 00:27:51.039 Don't do that anymore. 00:27:51.200 --> 00:27:52.160 Stop doing that. 00:27:52.319 --> 00:27:53.519 Do this instead. 00:27:53.759 --> 00:27:55.200 Here, look at this, too. 00:27:55.279 --> 00:27:56.480 And you have to hone it. 00:27:56.559 --> 00:28:05.279 We did a show way back where it was a guy who owned a roofing company, and he talked about how he had a full-time 40-hour week guy just doing quotes. 00:28:05.519 --> 00:28:05.759 Right. 00:28:06.079 --> 00:28:12.240 And he he went in and and created his own quoting system with AI on his own. 00:28:12.400 --> 00:28:15.440 He said it took them about two weeks of fine-tuning. 00:28:15.680 --> 00:28:24.640 He said, now his guys, when they finished, they used to take all their pictures and then give it to this poor guy who would just like pour over it and do all the quotes. 00:28:25.359 --> 00:28:29.920 He said they would get in their truck, upload their photos and their notes to this AI program. 00:28:30.559 --> 00:28:34.319 And before they got to the next job, the quote would be done for the job they just left. 00:28:34.400 --> 00:28:40.079 And that that person that was doing all the quoting did zero quoting and was in another part of the business. 00:28:40.400 --> 00:28:40.799 Insane. 00:28:40.880 --> 00:28:41.759 And that's exactly what it is. 00:28:41.920 --> 00:28:49.200 It's not, we're not trying to eliminate people, we're trying to replace the the work that doesn't make you money. 00:28:49.359 --> 00:28:52.079 So then you can focus on what does make you money. 00:28:52.400 --> 00:29:00.160 Yeah, and but you need to be, it takes you a long time to train a

salesperson. 00:29:00.640 --> 00:29:00.960 Right. 00:29:01.200 --> 00:29:04.400 To train this AI program is gonna take you a lot less time. 00:29:04.559 --> 00:29:06.319 You just have to be patient with it. 00:29:06.480 --> 00:29:09.839 You have to understand that it's not gonna work the first time. 00:29:10.000 --> 00:29:12.480 It might be way off, it might not even be close. 00:29:12.640 --> 00:29:19.359 It might get some things right, but you got to go back and talk to it because there's so many variables and elements in here. 00:29:19.680 --> 00:29:23.920 Yeah, I was gonna say you have to understand the complexity of what you're quoting too. 00:29:24.000 --> 00:29:37.200 I mean, if you have a very complex uh product or or service or something that's extremely variable, well, you your AI doesn't have much data or history to uh to use to be completely effective. 00:29:37.359 --> 00:29:45.759 But if you have something that's pretty repeatable and something that's very predictable, it's gonna be, you know, it's gonna be a lot easier to implement and quicker. 00:29:46.079 --> 00:29:47.680 And Craig, you brought up another good point. 00:29:47.839 --> 00:29:55.200 If there are variables, you could even train it to say, you know, give me an here, here's your rough quote. 00:29:55.359 --> 00:30:00.240 Tell me how many man hours do you think we're gonna need to to invest in this job? 00:30:00.400 --> 00:30:03.119 What's the our average hourly rate we're gonna use? 00:30:03.279 --> 00:30:05.279 What what are these z, right? 00:30:05.680 --> 00:30:11.519 Um, so then there is still a person who's going, okay, this is you know, maybe you have a senior tech and a junior tech. 00:30:11.599 --> 00:30:12.480 It's a basic job. 00:30:12.559 --> 00:30:16.480 I could use a junior tech on tech on this job, maybe it's 40 bucks an hour. 00:30:16.640 --> 00:30:18.319 I think it'll take them 10 hours. 00:30:18.559 --> 00:30:23.680 You just tell them that, and they're like, okay, and then they factor it in and give you the information. 00:30:24.000 --> 00:30:24.319 Right. 00:30:24.480 --> 00:30:25.920 That's exactly right. 00:30:27.119 --> 00:30:29.680 I don't I I love AI. 00:30:29.759 --> 00:30:32.319 I think it was built for me because I'm such a scatter brain. 00:30:32.400 --> 00:30:43.279 Like I can work on uh 10 different projects at a time and like jump from project to project very easily because it remembers everything we've talked about, right? 00:30:43.440 --> 00:30:43.759 Right. 00:30:44.000 --> 00:30:46.720 Um, but I I do go ahead. 00:30:47.039 --> 00:30:48.559 It can also be agreeable too. 00:30:48.640 --> 00:30:49.440 That's what I've noticed. 00:30:49.519 --> 00:30:53.920 You know, it's like you you start having these ideas, and it's like, oh, wait a minute, I don't know if that is actually a good idea. 00:30:54.960 --> 00:31:01.440 I start I start every project by saying uh always operate in truth mode. 00:31:01.599 --> 00:31:08.400 Never make up any info because it will make up stuff, and if it doesn't have info, sometimes it'll make it up and throw it in there. 00:31:08.480 --> 00:31:18.000 So I say, never use any inf only use information that's in the document, in the brand document, or we've discussed in this conversation, and always use red team. 00:31:18.240 --> 00:31:22.559 If you, if an idea I'm giving you, if you don't like it, let me know and tell me why. 00:31:22.799 --> 00:31:23.839 Yep, exactly. 00:31:24.000 --> 00:31:28.160 I I totally agree, and that's where we were talking about the MD file or the skills files. 00:31:28.480 --> 00:31:40.240 Like, this is what it needs to refer to in order to get that consistent output and not go out of bounds and try to solve it for yourself or try and solve it based off of its own reasoning. 00:31:40.640 --> 00:31:49.759 Um so that's important, and then uh also I think what we what we're starting to get to with AI is like, oh, what how much can AI do for me? 00:31:49.920 --> 00:31:59.759 But really, what I what I've seen a self-driving sales system being successful is is it aligns your people with efficient processes using the right tools. 00:31:59.920 --> 00:32:05.519 And that right now, we just have to treat AI as a tool for those people and processes. 00:32:05.680 --> 00:32:10.559 If those people and processes aren't aren't aligned, that tool's gonna do nothing for you. 00:32:10.640 --> 00:32:18.880 So combining all three of them, writing it out to where AI can understand the processes and people is where it's gonna be most effectively implemented. 00:32:19.200 --> 00:32:24.319 And I think I don't know about you, but the biggest miss I see is people not talking to it. 00:32:24.480 --> 00:32:27.680 They try to type everything, and that feels like a burden, right? 00:32:27.920 --> 00:32:35.440 Just talk to it, and you can stutter and stammer and change your mind and go back, and you don't have to go back and edit, it'll figure it out, right? 00:32:35.759 --> 00:32:38.079 And then there's there's a really cool AI. 00:32:38.160 --> 00:32:39.839 Uh do you use Whisperflow? 00:32:40.079 --> 00:32:41.279 Yeah, Whisperflow. 00:32:42.000 --> 00:32:43.519 Best things in sliced bread. 00:32:43.599 --> 00:32:46.240 I don't know that I I mean I I think I've forgotten how to type. 00:32:46.319 --> 00:32:47.680 I just everything. 00:32:47.759 --> 00:32:48.240 It's amazing. 00:32:48.880 --> 00:32:58.960 Yeah, it takes out the stuttering or uh even uh language barriers, or you know, if you're if it's not your first language, um, it can pick up your broken English or whatever pretty pretty well. 00:32:59.359 --> 00:32:59.599 Crazy. 00:32:59.759 --> 00:33:05.519 And if you change your mind mid-thought, it'll recognize that and just go with the second thought and eliminate the first one. 00:33:05.680 --> 00:33:15.759 So if like don't think when you use AI, and I'm talking to the listeners now, talk to it because it takes way less of your time. 00:33:15.920 -->

00:33:23.680 And even you know, if you have salespeople out on the road, now I'm switching going to outbound that hate using the CRM, shit, man. 00:33:23.839 --> 00:33:30.079 Just tell them to get in their car and pull it up on Apple uh Connect or whatever, what's it called? 00:33:30.240 --> 00:33:31.759 Uh uh Siri. 00:33:32.400 --> 00:33:32.720 Right? 00:33:32.880 --> 00:33:34.000 No, but what's the Apple? 00:33:34.079 --> 00:33:37.440 Most people are on Apple phones and there's the app in your car to connect your Apple. 00:33:38.559 --> 00:33:46.880 But pull up the app if there's an app for your CRM, pull it up or pull it up on your phone, hit record, and just talk into it. 00:33:47.599 --> 00:33:49.359 We don't need to be typing notes anymore. 00:33:49.440 --> 00:33:56.240 There's there's so many options to be talking to your which feels really weird in the beginning, but you get used to it. 00:33:56.480 --> 00:33:58.319 Yeah, and that's a good point, too. 00:33:58.640 --> 00:34:07.920 Like if you're driving and you need to log an interaction into your CRM, say, hey, this customer just said we're you know actually two months out instead of one month out, whatever the case is. 00:34:08.159 --> 00:34:12.800 Let's say they're on a phone call or it came in via email, this update. 00:34:12.960 --> 00:34:26.880 That's where uh your calls, you can have uh granola or um WhisperFlow or Fireflies or whatever you're using to just transcribe meetings, can actually be on your phone as well and pick up that call. 00:34:27.360 --> 00:34:32.400 And the agent can put that automatically, that interaction into your CRM. 00:34:32.559 --> 00:34:39.760 It can it can pick up from that conversation, you know, the specific bid request or project that it's that it's regarding and throw it in there. 00:34:39.920 --> 00:34:42.320 Connect it to the right company, the the contact. 00:34:42.480 --> 00:34:50.559 Um so you can either have it transcribe that or get off the phone and say, Hey, Claude or ChatGPT, log this into my CRM. 00:34:50.639 --> 00:35:00.480 I just talked with uh Joe Smith about um this specific bid request, and it's gonna be actually delivering in two months, and then it'll do it all for you. 00:35:00.719 --> 00:35:06.719 So I I I I'm not a big admin guy, so CRM uh data input is not my favorite. 00:35:06.800 --> 00:35:14.960 So that's exactly and it's it's funny when you know you're recording it for that to be transcribed. 00:35:15.519 --> 00:35:26.800 You ask, I find it affects the way you talk and ask questions because you'll clear really try to talk to get it clear so your AI note taker will like not miss it, right? 00:35:26.960 --> 00:35:27.840 So I can do it. 00:35:28.079 --> 00:35:32.000 So we're talking about October 27th, right? 00:35:32.159 --> 00:35:33.519 Yes, October 27th. 00:35:33.599 --> 00:35:34.239 Great, okay. 00:35:34.480 --> 00:35:45.360 And the scope of the work was which actually helps you because then you're repeating it back to the

individual, and they're like, okay, you get it, you understand, you heard me, but you're really doing it so they pick it up. 00:35:45.440 --> 00:35:48.639 So when you go back to your notes, you have all the information you need. 00:35:48.880 --> 00:35:53.599 Yep, but it yeah, it adds clarity for both parties, and now a third party that's taking your notes. 00:35:53.760 --> 00:35:55.039 So super interesting. 00:35:55.360 --> 00:35:55.679 Yep. 00:35:55.920 --> 00:35:58.719 Well, Craig, super interesting conversation. 00:35:59.199 --> 00:36:07.440 We've talked about a lot, but if you want, if if you could, you know, if a business owner's listening to this and they want to think, okay, and let's talk about inbound sales. 00:36:07.599 --> 00:36:12.159 I'm ready to uplift my inbound sales program. 00:36:12.400 --> 00:36:17.280 How can I take the information you talked about here today and scale from here? 00:36:18.079 --> 00:36:29.679 First, uh, I think I discussed it a little bit, but first and foremost is identify the pain points, identify the repeatable processes, and that's where you need to map it out. 00:36:29.840 --> 00:36:39.119 Again, record it, map it, put it in a document, SOP, and provide that document to your uh LLM or your AI of choice. 00:36:39.440 --> 00:36:44.239 And that's where you can really take that next step of what happens next. 00:36:44.400 --> 00:36:52.000 Um, provide your goals, and uh again, just analyze that specific use case and uh and really map it out. 00:36:52.159 --> 00:36:59.519 And then say, again, with your goals, like here's what I'd like to automate, and then it'll help you find that that scenario to make it to make it happen. 00:36:59.840 --> 00:37:02.159 And is that that's the R in race? 00:37:02.400 --> 00:37:06.960 Yeah, R in race is review, and then architect is really your goals and what okay. 00:37:07.039 --> 00:37:08.400 So that's the R and the A. 00:37:08.719 --> 00:37:11.280 All right, you remind me what the C and the E? 00:37:11.599 --> 00:37:15.360 Uh confidently launch and enhance. 00:37:15.519 --> 00:37:20.559 So I had to start with C with confidence because you don't want to just launch, you want to be confident you're launching the right way, right? 00:37:20.960 --> 00:37:25.360 And that includes the AI playbook or the AI handbook. 00:37:25.599 --> 00:37:34.400 That that's I think your really first step before even entering this is how how do we keep this under control and uh and with a good strategy. 00:37:34.719 --> 00:37:35.039 Right. 00:37:35.280 --> 00:37:36.559 Okay, very cool, man. 00:37:36.639 --> 00:37:37.119 Well, thanks. 00:37:37.280 --> 00:37:38.480 I appreciate the conversation. 00:37:38.559 --> 00:37:46.960 It's it's it's been I I I like talking about AI, especially when you talk about something really specific, because I think we talked about this before. 00:37:47.039 --> 00:37:58.800 When you talk about AI in general, it gets very overwhelming because there are a billion people out there using it for their own personal uh use cases, and it can feel like a lot sometimes. 00:37:58.880 --> 00:38:04.320 So I appreciate you breaking this down and being able to talk to talk uh specifically about sales. 00:38:04.639 --> 00:38:05.440 Absolutely, yeah. 00:38:05.519 --> 00:38:06.000 I agree. 00:38:06.079 --> 00:38:10.719 It's uh the more specific, the better, uh the better output you're gonna get too. 00:38:11.039 --> 00:38:17.039 So if people want to connect with you, Craig, and learn more about what you do or pick your brain, how can we, how can they get in touch with you? 00:38:17.280 --> 00:38:22.400 Just find me on LinkedIn, um, Craig Mooney and uh Auto Build Solutions is my company. 00:38:22.480 --> 00:38:23.119 So you'll be able to find. 00:38:23.280 --> 00:38:25.039 I don't think there's too many Craig Mooneys on LinkedIn. 00:38:25.119 --> 00:38:27.760 So you should be able to should be able to find me there. 00:38:28.079 --> 00:38:29.280 Very cool. 00:38:29.760 --> 00:38:30.559 All right, man. 00:38:30.639 --> 00:38:31.039 Thanks so much. 00:38:31.119 --> 00:38:32.320 I've enjoyed the conversation. 00:38:32.559 --> 00:38:32.800 Cool. 00:38:32.960 --> 00:38:33.199 All right. 00:38:33.280 --> 00:38:34.239 Well, there's Craig Mooney. 00:38:34.320 --> 00:38:35.760 That was that was an incredible conversation. 00:38:35.920 --> 00:38:39.199 We are going to be jumping into the members area to continue our conversation. 00:38:39.440 --> 00:38:42.239 If you are not a member yet, you can sign up for free. 00:38:42.320 --> 00:38:46.079 And we like to jump back there and have conversations with our guests. 00:38:46.239 --> 00:38:51.039 Um, continue the conversation, and you can get access to all of that good stuff for free. 00:38:51.199 --> 00:38:58.880 So go to the website, uh scalefromhere.com, sign up, be a member for free, and join the party. 00:38:59.119 --> 00:39:05.280 Also, if you haven't yet, hit that subscribe button, help us spread the word to more business owners. 00:39:05.440 --> 00:39:12.079 We would really appreciate it if you could just reach up and select, subscribe, or follow whatever it is on your platform. 00:39:12.400 --> 00:39:15.440 We would be forever indebted to you. 00:39:15.679 --> 00:39:20.079 And uh, let me see, we'll get ready for our next episode coming up. 00:39:20.159 --> 00:39:30.960 Uh, we have Sam Palazzolo, and we're gonna be talking about uh unrealistic leaders, how to achieve real results by being unrealistic. 00:39:31.280 --> 00:39:32.800 Looking forward to that conversation. 00:39:32.880 --> 00:39:34.480 It's gonna be a lot of fun. 00:39:34.880 --> 00:39:35.760 Thanks for listening. 00:39:35.840 --> 00:39:37.039 I'm Corey Harlock. 00:39:37.199 --> 00:39:39.280 Remember, you've built a real business. 00:39:39.440 --> 00:39:42.960 The next question is how do you scale from here?