199 - How to Capture After-Hours Leads with AI Call Agents with Austin Gray

0:0030 % of the code right now within Facebook or Meta and Microsoft is being written by AI. Then you look at their earnings. Everybody had blowout quarters. They've grown top line revenue. They've shrunk their employee costs because they're using AI to help their engineers 10x. It is so early. People still just have no idea how to use it. And if they are using it, they're using it to 10 % of its capacity or 5 % of its capacity. They're not even using it for every use case that they could be using it for. What are some other interesting ways that you've seen people making money with AI right now?

0:34So you want to talk AI? You've been doing the series, right? I just finished. I want to know what you learned. Oh my gosh, come on, man. I haven't had time to freaking digest all that. Talk to like 35 people. What are some of the coolest things you've learned? Three things that you're implementing. I'd say is we are early. I know that's like a cliche thing to say, but when you're in a Twitter bubble and everybody's talking about AI, it may feel like, oh, dude, everybody's using AI. But it is so early. People still just have no idea how to use it. And if they are using it, they're using it to 10 % of its capacity or 5 % of its capacity.

1:11They're not even using it for every particular use case that they could be using it for. I'll give you an example. I'm not even going to talk about agents or any of that stuff. I talked to this woman. Her name was Abby. She's a writer. And she just showed me in Claude how to organize her Claude. So what's your LLM of choice? Who's GPT right now? I have a subscription to Claude. So ChatGPT has projects. Claude has projects as well. Do you use the projects at all? So in our last episode, you went through that. Oh, okay. That's right. That's right. We talked about it. Sorry. I'm talking to too many people.

1:44So that was one thing where it was just mind blowing. She didn't show me anything crazy. It wasn't like she's like, oh, this is the software that I built from scratch. It was literally just, how do you think about architecting a system that allows you to leverage AI more than you've ever leveraged it in the past? And so the way that she was thinking about it by creating reference documents, by having different projects for different aspects of her life, like even her personal life, to me was fascinating. So that's one thing I've totally implemented and stolen is I have projects for everything now, whether it's work, whether it's personal, whether it's, you know, something that I'm working on with my kids, because each one of those projects keeps context.

2:24And I think the context is important. So to me, that's, that's the first thing is like, we're early, we're barely using it. And even the things that we're trying to use it for, we haven't even, we haven't even scratched the surface on the easy stuff. Did you record an episode with her? Uh-huh. You did. Did y 'all do screen share? we did yeah i could send you a link yeah let's put that in the show notes because last time what we did is we went down a deep dive rabbit hole but if you've already recorded that content if already lives on the internet i'm going to go watch that and then i'm going to recommend that our listeners go watch that so this is actually something that listeners like if y 'all have been listening to any of my ai stuff recently then like you probably know that i just i move really fast organization is not my strong suit but i'm kind of to the point right now where i'm in gpt and i've got like a thousand different thing that i've prompted it on and i just need to organize it this is going to be really interesting to go and look at it and if you'll share that link what was her name abby abby she had a million dollar community she had a million dollar business she shut that down and kind of built out her her new business that she's running just framed around AI, which was, which is fascinating.

3:39How she hit 60 K MRR in four weeks using AI. Who is that? Oh, that is Amanda, Amanda Orson. There we go. So she has a software that is effectively trying to eliminate realtors on the buy side and the sell side of real estate transactions. And so it's a marketplace where you can go and you can list and people can buy and it cuts out the middleman and she gives you a bunch of resources and templates that, that, you know, make it easier for you to not have to use an agent. Cause most people are using an agent for the contract or for the title company or for recommendations on who's going to do the inspection, like all the stuff that you just wouldn't normally do.

4:17And they make it easy for you to have a private party off-market transaction without having to use a real estate agent. But she was able to build that using AI within her team. And I mean, at the time, this was two months ago, she's at, you know, 60 grand MRR, but it was a very quick sprint and she was able to launch it. And it wasn't that AI built it. It was that AI is now giving, it's 10Xing her team's ability to actually get stuff out quickly. This was something that was interesting. Actually, I heard Satya Nadella, do you know who he is? CEO of Microsoft. And Mark Zuckerberg were having a conversation at one of these conferences.

4:54And Mark asked Satya, he's like, hey, what percentage of your code is written by AI? And they hemmed and hawed and they were like, well, you know, we can't really totally quantify it. But the number that they got to was like 30%. Like 30 % of code right now within Facebook or Meta and Microsoft is being written by AI. It's not that it's just on its own writing code. It's doing the first pass or it's doing an audit or it's helping to QA something for one of these engineers, but 30 % of code. And that's right now, which is just nuts. And then you look at their earnings. Everybody had blowout quarters.

5:28Microsoft, Meta, because they're now leveraging AI. They've grown top line revenue. They've shrunk their employee costs because they're using AI to help their engineers 10x. It's pretty cool. Where are the biggest opportunities right now that you've seen from interviewing these people about AI? All right. Well, let me take a step back. If you're a small business owner, what I would say the biggest, the first step that you do, I think we've talked about this is just, just go pay for a premium plan. I don't care what it is. I don't care if it's perplexity, chat GPT, Claude, Gemini, whatever it is, go pay for a plan.

6:02The reason you pay for it is so that you have more ability to ask it questions. If it's free, you kind of hit a cap, but just start playing around with it. After that, then you'll start kind of seeing opportunities to implement it within your business. But I would say from what I'm seeing, the number one best use case is sales. And here are like a couple of examples after our phone calls. So people who are calling for your services after hours, nobody has the income to be able to have a 24 hour team that's just answering calls, right? So many times those people call and they'll leave a message or they'll call and nobody answers because there's no message machine.

6:38And then they call the next person, the competitor, and then the competitor gets that business. So there are a couple of things that people are doing right now. There's a company called Think, T-H-Y-N-K. I had him on. John Long is his name. And they have AI customer service representatives who will answer calls. Now, they'll do it during the day and after hours, but I like the after hours component to it. And so they're capturing business. They're capturing leads after hours, which I think is incredible. And beyond that, they're also able to capture those leads and enter them into the CRM. So it's not quite an agent where it's autonomous.

7:11But you are telling it, hey, once you capture these fields of information, populate it in the CRM. And then when you come the next day to work, you've got two or three people who have called in. It's like, oh, you've got an evaluation with John Smith and Mary Jones at 11 and 12 o 'clock and you didn't do anything fast. So like capturing leads essentially after hours, I think is a really cool use case. Have you seen anybody doing that right now? I've seen a couple of people who are building products on it. We just started using open phone and it looks like their AI assistant or answering assistant works pretty well.

7:44I've experimented with that a couple of times. I've been hesitant to implement any sort of AI phone answering. To your point, I'm open to doing the after hours because why not? They're not getting touched anyway. You would have lost it anyways. Exactly. And the second thing that I want to ask you here, so I'm running a lot of Facebook ads for several businesses right now. And we're getting a ton of leads, but some of those come the middle of the night, early, you know, 3 a.m. in the morning sometimes. So is there a way that you've seen for an AI call agent to, or call assistant, to call those leads?

8:24Yeah, there is. I mean, so John Long, I can't remember which episode number, but that's his whole company is Think AI. and I actually have a phone conversation with his agent live on that episode. This is pretty fun. Hey, I don't know if you remember this, but when we started this podcast, we entered into a social contract. I would spend time, energy, and money producing this podcast, interviewing these individuals and giving you insights into how to build, buy, start, grow your business. And you would like, subscribe, and leave me five-star review. Now, out of that, we both get to talk to really cool people and hear really cool insights.

9:00We both get a ton of value, but I just want to help you keep your word. So would you do me a favor? Will you go leave a five-star review for me on Apple or Spotify? It would really help. And if you want, even share this with a friend. And look, you still know it's an AI agent. The problem is the latency. This is the main issue. They can actually have quicker conversations with you. So if you talk to an AI agent right now, it's like, hey, how's it going, Brian? It's good. How are you doing? I'm great, man. And, you know, there's like this lag. The reason that there's a lag is because AI still hasn't perfected listening and understanding when to jump in.

9:39If they just hear a pause or like a longer breath, they think they're supposed to jump in. So these companies will actually say, you need to wait not 0.5 seconds, but 2.5 seconds before you jump in. And so that's why there feels like there's a lag. We have the technology for there to be no lag technically, but they don't understand like the context of when to jump in and when not to jump in. And so it kind of just causes them to talk over each other. So from that episode, honestly, what I found was, I think if you have an AI agent and the AI agent says, I'm an AI agent, how can I help you? Like, just let the person know upfront.

10:15There's usually not a big problem with that. I'm sure on the tail ends, you'll have people who are just buttheads and they'll hang up. But at least then the person that you're talking to is like, oh, okay, I'm in the know. I'm just going to proceed with this conversation. So the short answer to your question is yes, you can have them call in the middle of the night. You obviously have to set up automations probably through Zapier or some automation software like that. But yes, you can have them actually call outbound and have conversations with potential leads. Cool. That's great. I like your thought process there on just being straight up with the customer.

10:49Hey, this is an AI agent, but I can answer most of your questions just to let you know, like, I'm going to document these. And then somebody from our team is going to be giving you a call first thing in the morning or however you want to frame it. Right. The other thing on the sales side, I would say is like, it's not an AI call agent, but it's an AI chat bot. So anybody who's going onto your website and wants more information, you can pretty easily train up a chat bot to be able to interact with people when it pops up and they start having a conversation, et cetera, and they get to a point and then it takes the lead and shoves it in the CRM.

11:22I do think that there is a space to be utilizing AI, your call or chat agents to follow up on leads or take in inbound leads that you otherwise wouldn't have gotten to that that are time sensitive. That's awesome. What else are you seeing as opportunities? What's the CRM that you use in your business? Go high level. Go high level. Okay, cool. Yeah. So go high level would be a perfectly good CRM to use, but basically qualifying scrubbing lists that you have, right? So you, you have to export it and kind of either put it in a CSV or in a Google sheet and then have Gemini and take a look at it, but you can actually have it now.

11:59I think clay clay is a CRM where it will go and look at those prospects and you can tell it, I want you to pull its LinkedIn profile, their LinkedIn profile. I want you to pull their business information. I want you to qualify this list of people based on these attributes. So just cleaning the data in order to get it to a place where your sales team can take it and run with it and just identifying, hey, here are the top potential leads and the bottom leads, I think is a really cool time savings of AI. So like you've seen deep research on ChatGPT. That's effectively what you're doing is like a deep research-esque function on the leads that you've generated.

12:38Upstream of that, there's this software called Browse AI. Have you ever messed with Browse AI? No. Okay. So browse AI back in the day when you wanted to go and scrape data from websites. So let's say I wanted to get the phone number of every pickleball owner in the United States. I would have to essentially write script to say, go to Google business profile and pull this specific information. And if they weren't on Google business profile, you're kind of out of luck. That's actually what a system like not go high level, but out scraper does right out scraper just goes and pulls all the Google profile, Google business profile listings and nothing.

13:11that they've got them in a centralized place. But what software like Browse AI does is it allows you to actually go to every individual website and then using context, because it's AI, you tell it, go find the About Me page and find their hours of operation, phone number, address, services they offer. And so then it will go to the website and actually look for all of those things and populate it into a spreadsheet. So you're no longer dependent on whatever Browse AI, or excuse me, whatever Outscraper has pulled together from Google Business Profile. you can actually go to the websites of these companies and pull that information because you're you're now using context you're not hard coding where you know and telling the software where to go to actually pull it so i think that's a really really cool use case for businesses that are trying to find more and better leads to call on that is not in the same pond that everybody else is fishing in which is google business profile browse.ai huh yeah yeah that's super interesting yeah and all this stuff is like straightforward and inexpensive in the grand scheme of things for what you're actually getting no it totally is and if you're willing to spend a little bit of time on it you'll you can figure it out and build some pretty cool stuff in a short amount of time where are some other opportunities outside of sales okay okay so outside of sales this is tangential to sales however it overlaps to all operations and again i'm like looking at things that don't require agents.

14:35I'm sure that you listen to Greg Eisenberg and he'd be like, oh, if you just plug in these three things, you can have an autonomous robot that goes and washes cars. I'm not there yet. But I'm like the caveman who's just trying to figure out how the wheel moves and he's designing Ford F-150s. Anyways, there is a lot of information that we communicate every single day that's not recorded, but there are ways that you could be recording it. So for example, I have a limitless pendant that records all my conversations. I'm not wearing it right now because we're on this and I could get the transcript if I wanted it.

15:13So there's no point for me to record it. And throughout the day, you have many calls with your team. Maybe you have a call with the sales team. Maybe you have a call with the ops team. Maybe you have a call with finance, maybe whoever. And so I like to think about what are opportunities for me to capture the information that is on these calls in a non-invasive or at least easy way. And so the first way that I would say is, well, let me back up. Why? Sorry. Why? Why would I want to capture all this information? It's because I can then take that information in those transcripts and put it into my LLMs and say, please summarize this conversation.

15:45Please tell me who I need to follow up with, et cetera. And then if you want to get really crazy with it, you can just start creating zaps that you don't have to do anything with the LLMs. You just set up this app and it does it on its own and it sends out meeting minutes after the meeting is completed. I don't know about your listeners, but I was awesome in meetings. It's fantastic. Dude, we solved all the world's problems in every meeting we had. Do you know what the problem was? Hey, what did we say we were going to do? I can't, what? No, no, no. That's not what we said, Austin. What we said was that I was going to go and you were going to stay.

16:19No, I remember that very clearly, you said that you were going to go and out, you know what I mean? And so recording, it just takes away the mental stress of like having to write down every single note, but also the ability to go back to it and double check many different things. And I'll give you some examples of it after this. But what I would do is take a look at any meetings that you have with multiple people where Zoom is involved. And I would look at getting an AI note taker, either Otter AI, Fathom, Fireflies. I'm sure there's one or two that I'm missing, but those are really simple. They integrate with your calendar and get them on every single one of your meetings.

16:53And they're going to record and take meeting notes for you. And then you can go and take the transcripts, which is fantastic. If you then want to take it to the next level, I would get one of those pendants. And it doesn't have to be limitless AI. It can be Plod. Plod's another pendant that's out there. It's cheaper. But just record your conversations. Be upfront with people and tell them like, yeah, hey, I'm recording this. But I can't count how many times I've had like one-on-one conversations and I'm relying on my recall at this point is just not good because I have so many different conversations.

17:24So video is cool, but I have what scientists call a face for radio. And so what's even cooler is long form audio via my podcast and my newsletter, nickonomicspod.com. Go there for free, subscribe to my newsletter. It's one email per week, super tactical. And then go to my audio podcast. I do three to five episodes a week, depending on how curious I am. And it's stuff like this. It's all free, no sleazy sales pitches, nickonomicspod.com. So here's a real world example of that. I have an intern right now and I'm in the process of launching an accelerator that's focused on healthcare. So we're going to be leveraging AI.

17:55It's going to be an accelerator and it's within a healthcare organization that I'm working with. Well, I have an intern who's helping me create that. And our first couple of days, all we did was talk about what does this internship look like? What do I need from him? What are the deliverables? What are we trying to do? How do we get there? What are all the things that we test? Hours of conversation. Typically, I don't have those recorded. It's not like I have some, some chat bot, but I have my limitless pendant. And from that conversation, those conversations, I created a 12 week internship program, deliverables broken down to the day where it's like, all right, today you're doing this.

18:31This is who you're talking to. And it took me a little while after those conversations to iterate that way. But he now has a roadmap and it's not like he's calling me every day. Like, Hey, what should I work on today? It's no, we talked about it. You know what the end deliverables are just go and do. So that's an example of that. But think about every time you onboard a new employee, you have a sales meeting, you're talking strategy once a month or once a quarter with your executive team. You don't have a conversation with your bank about financing. You're talking, you're talking to your accountant about taxes.

19:02How many times like the next, every year I do taxes. I'm like, what did we decide last year? Why did we decide that? They feel like reeducate me. But if I have those conversations recorded, I can just go back to the transcripts and ask AI those questions instead of spending$350 an hour for my attorney or my CPA to tell me the answers to those questions. So that's valuable in sales. It's valuable in operations. It's valuable in onboarding and orientation, people management, the whole gamut. I think figuring out a way to efficiently capture your conversations and then be able to extract value out of them is totally underrated.

19:36This is incredible. this I have so many ideas even and I know you're you're sort of looking at it more from like you said you're doing a health care accelerator but like my head is going to man if I'm an excavation business owner with a crew of three can can you think about how many times listeners right now you've had a conversation with other people and they're still calling you back asking for questions I mean, my head is going to like, what if you recorded your morning operating meetings and you had some sort of automation set up to where, because I'm sure, I'm sure you're set in with zaps.

20:18I haven't even looked into Limitless yet, but the way I'm envisioning is you can record a conversation that you have. You probably have some sort of zap automation set up to where it sends that recording. Limitless not yet. It will. It will. I haven't figured it out yet. Yeah, it's like, it's kind of like an extra, it's harder. They don't have a bunch of API integrations. They will. Oh, really? Yeah. So right now I'm copying and pasting it. But then from there, everything's kind of automated. So let's talk about the future, like once they will. I mean, I'm envisioning this as if you, eventually you could be like, okay, we are having a morning operational plan meeting right now.

20:55And John, Trey, and Adam are present, right? And then you have some sort of automation set up to where it's recording what you're saying as the foreman of that crew. It's taking the transcript, plugging that into ChatGPT with a pre-prompted framework to summarize and create checklists and step-by-step outlines for people. And then you can voice prompt it to say, please send this to Adam, Trey, and John. Some sort of automation could set up. And now we're getting into a, I think this is what you would call an agent, but I'm trying to paint the picture for where the future of this could be and how it can be.

21:37Can I do you one better? Are you ready for this? Yeah. Cause I've been, I've been thinking about this as well. I've had many conversations about this. Think about you have a sales team, you know, maybe you have two, three, four, five sales reps out there. How do you know whether or not they're doing a good job at the end of the day? There's one of two ways. Either you just look at their sales numbers or you go on a ride along. Neither one of those tell you the whole story, especially on a ride along. You're just kind of getting the best version of them at that moment. What if they had this on every day and you could go and have AI analyze the transcripts and like, Hey, tell me how they're doing.

22:08How are they doing about closing? Are they following up the people that they're supposed to follow up with when they have conversations? Are they speaking more or less than the person? Are they asking calibrated questions? Just, you know, a whole host of things. And then you don't have to go and spend five hours right along with every single one of them to identify their pain points. Like you've already gotten like 80 % of the way there from those transcripts in and of themselves. So it allows you to manage those salespeople way better than maybe you otherwise would have in the past. Or like you said, these guys are out in the field.

22:39It could even be a liability thing where anytime they go to a client's house, they've got it on and it's recording the conversation. Now you have to be careful about two-party consent states if you live in one of those states. I don't. I live in a one-party consent state. but if it's capturing those conversations and then it's like there's a dispute like oh well let's go back to the to the transcript and even if they're like well i never said that it actually records the audio as well so you can listen to the conversation it all of a sudden kind of eliminates the need to rely on memory and you can just kind of let's let's just look at the tape let's like refer back to what we actually talked about but yeah morning meetings sales i think is a is a really, really big one.

23:18I'm with you on the sales. Cause like right now in one of my businesses, we are, we're getting the leads. We're trying to dissect at which point of the sales process is broken. And the only way that me is a remote GM right now of this business, I'm relying on the in-person salesperson to tell me what the customer is telling them in their on in-person site visits or calls with the customer. So to your point, this could be a great way to break down where the process is breaking. The other cool things that have been is like, I've gone to several conferences now with it. And at the end of the day, or even in the middle of the day, I'm like, who did I meet this morning?

24:00Oh, here are the 10 people you spoke to. Here's what their businesses were. Here's what you guys talked about. Did I say I needed to follow up with anybody? Oh yeah, here are the follow-up items you said you were going to do. I guess it's powerful from that perspective because otherwise when you're going to these conferences or you're going to places where many people are congregating, you want to remember the connections that you're making. How else do you do that? It's like, it's kind of, unless you're just taking notes the whole time, there's no way to really maximize those efforts. But with these now recording the conversations, you can go back, refer to it.

24:31When you go to follow up, when you're like, oh yeah, Austin said that something about excavating, what did he say? I think I might have something for him and they go with the transcript or you listen to it and you're like oh easy that's what he said um here's his pain point i can send him a follow-up message that's very personalized i mean just the amount of especially you as a creative entrepreneur the amount of ideas you have that on any given day this is going to unlock so many ways to document those what did we talk about scenarios and i mean this is this is going to open up the world into having a personal assistant eventually once this thing has an open api it might have an open api i just have it's it's more complicated than the typical software that's been out there for a long time it might not have a zap right so it's like i think it has API capability, but it doesn't have a zap.