Dan Martell - November 18, 2026


How to Build a 1-Person Business with AI Agents

It took me 30 harsh years to learn what I’ll teach you in just 27 minutes

Episode Stats


Length

31 minutes

Words per minute

226.51

Word count

7,023

Sentence count

281

Harmful content

Misogyny

1

sentences flagged

Toxicity

3

sentences flagged


Transcript

Transcript generated with Whisper (turbo).
Misogyny classifications generated with MilaNLProc/bert-base-uncased-ear-misogyny .
Toxicity classifications generated with s-nlp/roberta_toxicity_classifier .
00:00:00 The people running the world's largest AI companies all agree on one thing.
00:00:04 The biggest companies that will be built over the next 10 years won't be built with big teams.
00:00:08 They'll be built by one person with a team of agents.
00:00:12 But most of us are still thinking that if I want to grow, I got to hire a bunch of people.
00:00:15 No, you don't.
00:00:16 Back when I was in my 20s, I started my first company.
00:00:18 We were doing 1.6 million in revenue with 20 people on payroll.
00:00:22 I was stressed out all the time.
00:00:23 Earlier this year, one of my companies at Martell Ventures hit 1.2 million in reoccurring revenue
00:00:28 with just one founder, two part-time contractors,
00:00:31 and AI agents doing 99% of the work.
00:00:33 So today, I'm gonna show you
00:00:35 how to build a one-person company with AI agents,
00:00:38 what to build, and which agents to create.
00:00:40 Then I'm gonna show you how to connect them all
00:00:43 so that you don't have to be involved with any of it.
00:00:45 But first, we need to avoid the biggest mistake
00:00:47 everyone out there is making.
00:00:49 Before you build anything,
00:00:50 we need to know what you're going to build.
00:00:53 If you've seen any of my videos, you've heard me say this.
00:00:55 How do you find a problem that people will actually pay to solve?
00:00:59 So what we need is a painkiller.
00:01:02 What we don't need is a vitamin.
00:01:03 If I've got a migraine, I will pay almost anything to have a pill to solve that migraine.
00:01:09 If I'm not in pain, if I don't have a problem, then paying for a vitamin is way harder.
00:01:15 In business, we need to find people that have big problems,
00:01:18 that have massive pain around it.
00:01:19 Because the bigger the problem, the bigger the money.
00:01:21 So we have to start with the problem, not the idea.
00:01:25 I know your idea sounds awesome, love it,
00:01:27 but what is the problem it's trying to solve?
00:01:30 Because if we double down on that,
00:01:31 then the sale becomes so much easier.
00:01:33 So before I teach you how to buy back your time using AI,
00:01:36 let's first figure out what is the actual problem
00:01:39 that you can get paid for to solve.
00:01:40 Then we need to use the AI to give us insights
00:01:44 to how to speak to the market.
00:01:45 If you own a gym, for example, right in there,
00:01:47 I've run a gym for the last 10 years,
00:01:49 I need you to find 50 nuance and observable challenges
00:01:52 gym owners face right now
00:01:54 because you wanna solve problems for gym owners.
00:01:56 Now look at all of these problems and then ask yourself,
00:01:59 are gym owners already paying to solve those?
00:02:01 Pick the top three and those become the problems
00:02:04 that your service or offer will solve for them.
00:02:06 Then you have it write a message based on those problems
00:02:09 to 10 people, gym owners in this case,
00:02:11 and then you don't ask for the sale, you ask for advice.
00:02:15 If you ask for the sale, you get advice.
00:02:17 If you ask for advice, you get the sale.
00:02:19 When you reach out to those 10 gym owners,
00:02:21 you just ask them,
00:02:21 have you ever experienced these three pains?
00:02:23 And if so, do you have time to give me advice
00:02:25 on my new company?
00:02:26 And they'll be like, oh yeah, I actually have.
00:02:28 I'd love to get on a quick call.
00:02:29 And at the end, they might say,
00:02:30 well, what are you thinking of building?
00:02:32 And then you present them your offer,
00:02:33 the actual business that you're in
00:02:35 that solves those problems for them
00:02:37 and ask them if they wanna be a first customer.
00:02:39 If you already have a business,
00:02:40 you already know who the customer is,
00:02:41 run that exact same search
00:02:43 to help you really hone in your messaging
00:02:44 for identifying the painkiller,
00:02:47 not the vitamin that you solve for them.
00:02:48 So now we know what we're solving.
00:02:50 You've already reached out to them, asked for advice.
00:02:52 maybe you've already gotten some clients but this is where your time starts to pile up fast because
00:02:56 you're on phone calls maybe you got those clients you're delivering your time and you need to learn
00:03:00 how to get your time back by using agents but where do we start most people pick the wrong
00:03:06 thing and end up just busier than ever before without increasing their profit solving the
00:03:11 right problem in the wrong order will just cost you a lot of money so in my book buy back your
00:03:15 time i teach people this process called the replacement ladder this is that for the age of ai
00:03:21 replacement ladder with a i the first one is admin essentially your inbox in your calendar all
00:03:31 communications all information in one spot so we call this the inbox agent i used to teach people
00:03:38 to hire somebody to do this but in the world of ai we don't need to have a person do this anymore
00:03:44 we just have to configure the agent and what's crazy even a couple weeks ago when i hosted a
00:03:49 a thousand person event, teaching them all how to use AI.
00:03:52 Most of the room had still not set up their inbox
00:03:55 in their calendar with their agents.
00:03:56 They were still triaging all the emails
00:03:58 and replying to everybody
00:04:00 and managing their calendar themselves
00:04:01 because they didn't wanna let go, they didn't trust.
00:04:03 There was even this woman on my customer success team
00:04:05 and she was buried in email, support emails, 1.00
00:04:08 customers asking for stuff and she wanted to over deliver.
00:04:11 She was spending all of her time in her inbox.
00:04:13 Then I sat down with her and I just said,
00:04:15 hey, let's just give an agent all the instructions
00:04:17 things you're doing by it learning off of your existing inbox and then it wrote all the emails
00:04:23 and replies it pulled all the information that it had from her existing inbox and created a whole
00:04:28 process to answer all the emails by the next day 10 a.m her whole support email was done
00:04:34 it was such a big thing for her she actually got emotional about it she literally cried because it
00:04:40 was the first time she felt like she got caught up i still love you jen the goal is to give the agent
00:04:45 the whole inbox the whole calendar and then let it decide the rules based on how we shape it and
00:04:51 give it instructions it can decide on its own and look we're gonna use review mode only meaning
00:04:56 everything it does we're gonna review it first make sure it's good before we approve and send
00:05:00 so you still have an opportunity to review it and refine it but over time you're gonna let it go
00:05:05 you're gonna let it rip because it's gonna do an incredible job thinking like you because it learned
00:05:09 from you and here's what i know if you've got a business or starting a business is you didn't
00:05:13 start it to just become the bottleneck. You want to allow revenue to flow through you, people to
00:05:18 reach out to you, opportunities to come in your inbox, and then boom, reply to them. Boom, get back
00:05:23 to them. Boom, get them scheduled. Boom, qualify them. To make this super simple, we're going to
00:05:28 build the agents the exact same way so that it can run on its own every day. And the whole point
00:05:33 of an agent is so that you can go do other stuff, human stuff. You can use any AI tool out there
00:05:39 today. You just have to connect the things it needs, like your email and your calendar. So you
00:05:43 can use ChadGPT, you can use Claude, you can use GrokBot, you can use whatever. But to make it work
00:05:48 really well, you got to give it these four things. First, you have to understand data equity, which
00:05:52 is the data it uses to train itself. Second, you have to give it the blueprint that it follows so
00:05:56 that it always nails the task. You have to define the schedule, how often and when should it check
00:06:00 and do the work. And then fourth, let's create a sensor so that we know it's doing the right thing
00:06:05 and if it goes off track, we can jump in and fix it. Here's what most people miss when we talk
00:06:10 about this. See, every company out there today is using AI. So AI can't be the edge. Having the data
00:06:16 is. Data equity is the unique way that you think about doing something. And the data, the templates
00:06:23 that you feed the AI so that it can nail it every time. Think about the things that you have that
00:06:28 nobody else has access to. We're talking sales calls, scripts, processes, templates, all the
00:06:33 information about past customers. Your inbox is potentially data equity for your inbox agent. So
00:06:39 for the inbox agent, here's the prompt. Scan my inbox and look for every reply I've sent. Come up
00:06:45 with the rules I follow, then set up a filtered inbox using those rules. So if you get an email
00:06:50 that comes in and you usually forward it off to the finance team, guess what? It'll see that. It'll
00:06:54 do that every time. Then we have to feed it the blueprint, essentially the step-by-step rules that
00:06:58 it should follow as items come in. If you don't have them, you can have it create some, or my
00:07:04 favorite is find people that are world-class at it and just buy them or have it research the best
00:07:09 ones out there and pull those in. For example, your calendar, you can tell it the work hours you
00:07:13 normally work when you like to be home, your structure and your flow when you like to do deep
00:07:17 work and then have it book every meeting as the requests come in into the right slot. So you get
00:07:22 an upgrade in your life. Same thing for inbox. Maybe you're not the best at replying with details
00:07:27 and structure. So you ask it to do research to find the blueprint for how you should reply to
00:07:31 emails that is professional and direct and has all the information. So people aren't hitting reply
00:07:35 asking for more information and having an email create three or four other emails because the
00:07:39 first email wasn't clear. And then for your inbox, just have it sort your emails into the right
00:07:43 labels based on my GPS label system from my book. Again, if you don't know what that is, ask it. Go
00:07:49 to Dan's buybacker time book, pull out the GPS system for his email labels and apply those labels
00:07:54 to my emails. So then just have the inbox agent scan your emails, write draft replies, review them,
00:08:01 edit them and hit send. Then tell the agent to go look at all the emails you hit send and figure
00:08:06 out what it wrote versus what's different and update its rules to write better emails the second
00:08:10 time. That is what creates an agent. If you don't have a loop where it's learning because it wrote
00:08:15 a draft, you edit it and you hit send, then it doesn't have a way to learn. And without a learning
00:08:20 loop, it's not an agent. Next, we got to talk about schedule. How often do we want this process
00:08:24 to go? Now, if you want to make sure you're on top of your emails in real time, you can set it up
00:08:28 every 30 minutes. I like to wait at 11 a.m. and 4 p.m. to process the whole inbox at once so that
00:08:34 anybody that emails me knows they can't expect an instant reply like I am busy doing client work but
00:08:40 at least twice a day I'm clearing my inbox I'm reviewing my drafts editing those drafts and I'm
00:08:45 hitting sent. Look if you reply way too fast especially at three in the morning either they're
00:08:49 gonna think it's AI or you don't sleep and you don't know how to manage your time. So for now
00:08:53 that's my recommendation but let's think about this now that you've built this agent how do you 1.00
00:08:58 ensure that it doesn't just go off the rails and make you look like an idiot. That's where we have 1.00
00:09:01 to build a sensor. A sensor is telling you that it's running. So what I like to do is I have it
00:09:06 message me on Slack as a DM and tell me every time it processes all the emails that I left for a
00:09:12 draft, a link to my draft folder. So I have to click on it, go review them all. And then the
00:09:15 status of how many emails I've gotten or didn't reply or how many replies are outstanding or who
00:09:20 did I send stuff to that didn't reply? That's a sensor. Now I have a dashboard of all my inbox
00:09:25 messages what's been replied to what i'm waiting on and then i can just tell my ai to follow up
00:09:30 it'll write the follow-up email to nudge the person to get my reply think about how much better
00:09:35 that would be for sales emails reaching out to podcasts to get on structuring your life
00:09:40 coordinating events with other people like this process will not only keep you sane it will give
00:09:46 you so much more bandwidth but again pro tip day one just have it sort your email write some draft
00:09:52 and check the work before you hit send.
00:09:54 And then by all means, have it scan your calendar,
00:09:57 figure out ways you can be more productive
00:09:58 and reorganize your time blocks,
00:10:00 hopefully you have some,
00:10:01 so that you hit the work running when you've got to work.
00:10:04 And every agent's gonna follow that process.
00:10:06 So now think about it.
00:10:08 Every day, let's say a new client emails you,
00:10:10 the agent picks it up, it writes the draft,
00:10:12 you review at 11, you hit send,
00:10:14 and before the end of the day,
00:10:15 maybe they are scheduled in your calendar
00:10:17 for a 15-minute discovery call and you weren't involved.
00:10:20 This is business at the speed of thought.
00:10:22 Just so that we're clear,
00:10:23 here's the actual prompt you wanna give it
00:10:25 so it has the full learning loop.
00:10:27 Look at my email and process my inbox.
00:10:29 Draft a response as me in my voice
00:10:32 pulling from emails I've already written.
00:10:34 Now don't send it and leave placeholders for me
00:10:36 to fill in the details you might not have.
00:10:38 The inbox agent is just the first job.
00:10:41 If you're a business owner
00:10:42 and you want AI running in every part of your company,
00:10:45 I got you covered.
00:10:46 That's why I built my AI company operating system,
00:10:48 which AI tools to use in each department,
00:10:51 what to hand off first, and what to keep human.
00:10:53 If you wanna just scan the QR code on the screen
00:10:55 or click the link in the description
00:10:56 and I'll send it over for free.
00:10:58 The best part about this for you
00:10:59 is you now have your mornings back.
00:11:01 So it means you have time to go get more clients.
00:11:03 But here's the thing,
00:11:04 if getting new clients still means
00:11:06 that it adds way more work into your calendar
00:11:08 and getting more and more new clients
00:11:10 means that you don't have time to serve the ones you have
00:11:12 and get them a quick win fast,
00:11:14 then the people aren't gonna refer more people to you
00:11:16 because they're gonna think you're overloaded.
00:11:17 That's why we need to go to the next level
00:11:19 of the replacement ladder.
00:11:20 Next up is delivery.
00:11:22 We need to build a delivery agent.
00:11:25 The best part is this agent can run everything for you.
00:11:30 I'm talking about onboarding new clients,
00:11:32 answering every question they might have,
00:11:33 creating documents for you to review with them,
00:11:35 and in many times, doing the work
00:11:37 that you actually sold to the person.
00:11:39 Whatever problem they had
00:11:40 that you're offering a service around,
00:11:42 on the backend, they don't need to know
00:11:44 that you're using AI.
00:11:45 What's funny is some people still hold on
00:11:47 to this part and not even realize it. Even in one of my buddy's software companies, he was selling
00:11:52 a bunch of licenses, but part of the onboarding required the person that bought the software to
00:11:57 sit with somebody on his team to make sure they connected everything so they could activate. He
00:12:01 didn't realize that work needed to be scheduled. The customer had to come to the call with the
00:12:05 right information. And for a lot of the customers, it took 10 hours of time to get the customer
00:12:10 onboarding. So as he grew the business and did better marketing and had better sales calls,
00:12:15 all of a sudden, all the customers started backing up.
00:12:18 And when a customer has to wait 10 days, 20 days
00:12:20 to finally get the thing they bought,
00:12:22 they're gonna decide to cancel and refund.
00:12:25 And when I showed him what I'm about to show you,
00:12:27 he's like, oh my gosh, I can't believe I'm still using people
00:12:30 for the thing that AI can do for me.
00:12:32 So then he implemented this
00:12:34 and he has AI actually onboarding the customers,
00:12:37 calling them, doing intake forms,
00:12:39 making sure they're scheduled,
00:12:41 doing follow-ups, text messages,
00:12:42 all managed by the delivery agent.
00:12:44 The key to doing this at scale if you're just starting
00:12:46 or if you have an existing business
00:12:48 is to simplify and really ask yourself,
00:12:50 what's the one thing we sell?
00:12:52 What is the one process to take from somebody
00:12:54 from giving us money all the way to being very happy with us?
00:12:57 And those steps in the process have to be documented
00:13:00 because that's how we're gonna train our delivery agent.
00:13:03 In my world, as soon as somebody becomes a customer
00:13:06 and they gave me money,
00:13:07 I wanna move as fast as I can to get them their first win.
00:13:11 As soon as I get them their first win,
00:13:12 And then I can ask them if they know anybody else just like them that they might want to refer to
00:13:16 me so I can help them the same way I just got them that way. That is how this thing makes you a lot
00:13:21 of money and gets you your time back. The goal is not becoming better at doing tasks. The goal is
00:13:27 actually to stop doing the tasks that the technology can already do for you. And in today's
00:13:32 world of agents, that's pretty much everything. So here's your goal. Have your delivery agent
00:13:37 onboard every customer so you don't have to. Meaning that as soon as you get a customer,
00:13:41 you don't have your calendar turn into chaos
00:13:43 trying to get them activated.
00:13:45 So we gotta start with the first part, which is data equity.
00:13:48 What do you already have in regards to your internal systems
00:13:50 for what great looks like?
00:13:52 That analysis is your data equity.
00:13:55 So on that step, questions I'm gonna ask myself
00:13:57 to start looking for the data part is,
00:14:00 what do I log into to get this person results?
00:14:02 What will I see when they get results?
00:14:05 What do they need from me to get results?
00:14:07 All of that becomes the collection of data that I have
00:14:10 so that I can guide the agent?
00:14:12 The second is the blueprint.
00:14:13 What's the path or the process
00:14:15 that every agent should follow to guarantee those results?
00:14:18 One of the easiest ways to do this
00:14:20 is actually use any recording software.
00:14:22 I use Plod, but you can use Pocket, you can use GrokBot.
00:14:25 You can talk to any AI and pretty much just talk to it
00:14:27 about how you work with a customer
00:14:29 and have it create an SOP or a process
00:14:31 on what you just told it.
00:14:33 And it's funny because the other day
00:14:34 I was sitting out with my brother and he's like,
00:14:35 well, how would I get it to do this?
00:14:37 And I said, well, how do you do it today?
00:14:38 And then I recorded him and I was like, keep talking.
00:14:40 and what does great look like where's the data equity we grabbed all that information now we
00:14:43 have everything we need to start to build the agent so here's a prompt to actually get the ai
00:14:48 to reverse engineer the process out of you interview me one question at a time about how
00:14:54 i deliver my work from the moment a client pays to the first wing then list every step mark which
00:15:00 ones need me and build a plan that can automate the rest see for my brother the only thing that
00:15:05 he was doing in that relationship was doing the conversation with the potential buyer
00:15:09 the prep work before the call all the intake form and the work that was done after they signed the
00:15:14 sales agreement was done by the bot the ai will ask you all these questions you'll answer it and
00:15:19 then it'll create a process and tell you what tools they would need access to to automate this
00:15:24 so think about this you sell anything to anybody doesn't matter local service business a marketing
00:15:28 agency or even a business coach the moment the payment clears it sends a welcome email with your
00:15:33 intake form the customer fills out the intake form and it collects everything their goals their
00:15:37 preferences how they like to work and the tone they like to work in it now uses those answers
00:15:43 to set up the client folder that the agent has access to and it builds the faq about the client
00:15:49 and answers all of their questions and it stores everything in one place just for them and one of
00:15:53 the pro moves is using any of these ai tools and you can actually use the teach a task tool where
00:15:58 you can do something on your computer to teach the ai how to do a task save that as a skill and then
00:16:04 in the future when you're wiring up all these different steps that skill drives that outcome
00:16:09 for your delivery agent then we've got the schedule and i think this one is pretty simple
00:16:13 it should run every time a deal closes and every time a question comes in from the client the goal
00:16:18 for the delivery agent should always be get the client a win as fast as possible ideally on the
00:16:24 day they gave you money so now we got the last part which is a sensor and for me the easiest one
00:16:28 is a dashboard of every one of your clients and all the different phases of your onboarding or
00:16:33 activation within your service, your business, and showing you what stages they're at and how
00:16:38 long they've been stuck in that stage. Now, if you set up the delivery agent and you keep automating
00:16:42 each part, they should work through that really fast. Now, you're always on the mercy of them
00:16:46 replying to an email or scheduling something, but this way you have a dashboard where you can see,
00:16:50 is everybody green? Is there any yellow? Is there any red? Let's fix the red ASAP so that we know
00:16:55 our agent is working properly. And then maybe there's questions the client asks that the agent
00:17:00 can't answer. It'll just queue those up for you to respond to and then learn from your reply and
00:17:05 then use that in the future if any other new clients ask that same question. Now you have
00:17:10 a delivery agent that's running your onboarding and supporting customers and getting from new
00:17:15 to win as fast as possible. Now I'm going to give you a pro tip. One of my favorite things to do
00:17:20 with a friend is have the AI review their support email, find all the reoccurring questions that
00:17:25 people are asking and suggest fixes in the onboarding process to remove those questions
00:17:31 from the client's mind. I hope you can feel the excitement I have for you and your business
00:17:36 because this is the kind of thing that's going to change the game for how you scale. So now we know
00:17:41 that new customers are going to be wildly taken care of, but how do you start getting new clients
00:17:46 coming through the door every single month? Next up is marketing. How do we build a marketing
00:17:53 agent. We had to start with inbox and delivery so that we didn't get overwhelmed if we get you
00:17:59 a bunch of new customers. And what's funny is most people start at marketing, but then get
00:18:03 overwhelmed. We do it on the third level of the ladder so that we get our house in order to be
00:18:07 able to receive and deliver for those clients because there's no point in getting customers
00:18:11 that are not happy because then they won't refer. And if you have to keep getting new and new
00:18:14 customers every time, there's no point of building the marketing agent. The problem with most
00:18:18 marketing is that if it depends on you, you're only going to do it when you have the time. But
00:18:22 you get busy in other areas doing sales calls or trying to hire more people then the marketing ends
00:18:27 up taking a back seat so we have to build a marketing agent so it's running every single
00:18:31 day finding you leads so that it never stops when it comes to marketing consistency is the
00:18:36 foundation for winning write that one down so data equity remember that intake form we were asking
00:18:42 people to fill up for the delivery agent we can use that to guide the new marketing agent to go
00:18:47 find us people just like the ones that just bought from us because that's the perfect fit customers
00:18:52 somebody already understands they have a problem they like the solution that we offered are the
00:18:56 best people like them that would buy from us so that's the data equities if you have existing
00:19:02 customers any information you want to feed that to your existing marketing agent so now the blueprint
00:19:06 the best part is we can just tell the ai because it's got a lot of marketing training and processes
00:19:11 to just say every day go find me new clients just like me i need their email and their phone number
00:19:17 that have this problem that live in this city
00:19:19 or that have this size of business.
00:19:22 For example, search Instagram, search Facebook groups,
00:19:25 search X, find people in this city.
00:19:28 All these AI agents can run these long running tasks
00:19:31 to search people that are complaining
00:19:33 about the problem that you solve
00:19:34 and recommend those people and use the data
00:19:36 that it finds on their website, et cetera,
00:19:38 to give it to you so that you can reach out to them.
00:19:41 Pro tip, if you sell to consumers,
00:19:43 like you have a gym or a local med spa,
00:19:46 you can have it search social media finding people currently complaining about that problem
00:19:51 and offer them a solution or at least give them something of value. Next we go to schedule.
00:19:55 Easiest one is ask it every day to go find you 10 new leads every morning delivered to your inbox.
00:20:01 Now you might say well those leads aren't very good. Tell it what's not good about the leads and
00:20:05 it'll learn. The last one is the sensor. With all marketing we want to look at the top of funnel
00:20:09 activity and the quality of the bottom of the funnel. So did it find us leads? Were they qualified?
00:20:14 Did they get on a call?
00:20:15 Did they buy?
00:20:16 Once we find the process the marketing agent's been running
00:20:19 to find the leads that bought,
00:20:20 we tell it to learn to go find us more like that.
00:20:23 But we have to monitor that and create a dashboard
00:20:25 that shows us every day the activity and the process
00:20:29 it's running to generate leads and quantify it
00:20:31 and measure it so that we can feed it back to the agent
00:20:34 so it can do the job better.
00:20:36 Like my brother who sells homes,
00:20:38 scans Facebook in his city
00:20:39 for anybody talking about their home projects
00:20:42 and then just cleverly reaches out to them
00:20:45 and says, if you're ever looking to sit down
00:20:46 to design your dream home, just let me know.
00:20:49 If you want your marketing agent
00:20:50 to also help you create content
00:20:52 so that you can add value to the market
00:20:54 and also go viral so that you can get
00:20:56 more potential new leads, just give it this prompt.
00:20:59 Act as a viral content topic finder.
00:21:02 My niche is, tell it what your niche is,
00:21:04 then ask it, give me 10 topics built on real,
00:21:07 observable, nuanced problems for people in my niche.
00:21:10 Make them feel novel, not cliche,
00:21:12 and easy to explain in seventh grade English.
00:21:14 Then build scripts for these in my voice as short videos.
00:21:18 Now here's a pro tip.
00:21:19 If it gives you a list of things
00:21:20 that just feel broad or familiar,
00:21:22 ask it for a more specific observable problem.
00:21:25 That little iteration at the end
00:21:27 will allow you to create very cool, very viral content.
00:21:31 You just gotta pull out your camera,
00:21:32 talk to it, and publish.
00:21:34 Once we've got marketing running every single day,
00:21:37 the next agent makes sure that those leads
00:21:39 don't slip through the cracks. Next up is sales agent. This is just where a lot of people get
00:21:44 stuck because buyers still buy from a person. And the best part about this is I'm saying the sale
00:21:51 still stays human. What your agent does is everything else around it and it can even
00:21:56 coach you. It's my favorite because it makes me money. And across all my companies that have
00:22:01 incredible salespeople, it allows them to increase their bandwidth and also get really good at sales
00:22:06 in a learning loop. And I'll teach you how. It's like one of my company's CEO was like spending
00:22:10 50 hours a week on sales. Awesome. Lots of leads. His inbox was managed, but he's on a bunch of
00:22:16 calls. So what we did is we built an AI voice agent that would call every new lead first and
00:22:22 just tell it, hey, this is an AI voice agent. I'm just calling on behalf of Bob. And then it would
00:22:26 ask the questions to qualify them, touch on their pain, essentially following their internal process,
00:22:30 data equity, so that the AI agent could qualify the leads and cancel the calendar invite so that
00:22:35 he didn't spend time with people that were never gonna buy.
00:22:38 And it said it was a sales agent
00:22:40 and the person still answered the question,
00:22:42 still talked to the AI bot.
00:22:43 He got 20 hours of his week back
00:22:45 and those 30 hours of sales call,
00:22:47 people showed up and they wanted to buy.
00:22:49 A lot of people still say to me,
00:22:51 I don't think AI can touch sales.
00:22:53 Watch this.
00:22:54 So the close stays human,
00:22:56 but the qualifying from the leads,
00:22:58 let the AI take care of that.
00:23:00 And I'm saying, how about we just figure out
00:23:01 what you're doing in the sales process
00:23:03 and give it to your sales agent.
00:23:04 And if you don't know how to sell, we're going to design the sales agent to actually call you,
00:23:08 act like a buyer, you try to sell it, and after the call, it'll rate the sales call and give you
00:23:13 tips to get better the next time. I build sales agents for my friends all the time. When they're
00:23:18 super busy, I'm just like, hey, why don't we look at your past calls, figure out who are the perfect
00:23:22 buyers, have the sales agent look at all the questions and objections to bring in, give you
00:23:26 the talk tracks to deal with those objections. Now we have the bot monitoring your calls and
00:23:31 giving you feedback in real time every time you get off a call. So now we got to build the sales
00:23:35 agent. This is the agent that's going to help you do more sales and help you get better at sales.
00:23:39 So the first part is data equity. I know there's companies out there that just don't record their
00:23:43 sales call. That is a wild misstep because having that data, those calls from your top reps, from
00:23:50 your best calls where people bought quickly, that information is the data equity that is unique to
00:23:55 you that makes your business valuable. That's why it's called equity. If you don't have any data for
00:24:00 the sales call, just have it scan all of my sales videos and build what you need from that. Look,
00:24:06 I give away everything for free. My mission in life is to die empty. That's why I produce
00:24:10 literally hundreds of videos every week. And in those, there's a lot of sales stuff. Just have
00:24:15 your AI analyze my stuff to then create the stuff you need. So for the blueprint, we want to give it
00:24:20 two jobs and specifically in this order. Again, if you're just starting off, I want you to first
00:24:26 have it coach you have it call you and run mock sales calls with you acting as your ideal client
00:24:31 and then review the call afterwards against your sales steps and tell you where you could have done
00:24:36 better the second part is for it to take over i'm talking about qualifying before anybody books in
00:24:41 your calendar following up the same day calling back-to-back no-shows and rebooking essentially
00:24:46 doing everything other than the clothes putting data in your crm tracking all the information the
00:24:51 buyer told you is their key problems and storing that all the data stuff all the process stuff all
00:24:56 the calendar stuff just give it to your sales agent so here's the prompt to create your sales
00:25:01 agent call me and run a sales call as my ideal client then review the transcript what percentage
00:25:06 did I talk versus listen which of my sales steps did I skip and how should I handle objections I
00:25:12 fumble it's like waving a magic wand and designing the perfect sales manager or sales assistant for
00:25:18 you and have the ai just run it then we go to schedule and that for me is review after every
00:25:23 sales call if you record it it can plug into any software that you're using to record and then
00:25:27 analyze and give you feedback in real time so if you have another call in a half hour you'll get
00:25:32 that tweak before the next call and you can try it on the next one for the sensor and a sales agent
00:25:37 it's two things one is your win rate as you do sales calls are you improving your ability to
00:25:43 close and two is data hygiene is the information it's identifying in your sales calls being saved
00:25:49 transcribed and added to your crm so that it tracks everything so that when you get a customer and you
00:25:54 hand it over to your delivery agent all the information that it needs to onboard the person
00:25:58 from the stuff they already told you as a sales person doesn't have to be repeated by the client
00:26:03 i'm telling you if you do that it will be a game changer so that's how we get better at sales
00:26:07 without losing your whole freaking week by having the agent coach you first so that you can get
00:26:12 better without having to practice on real opportunities and then take every part of the 0.82
00:26:16 sale that isn't human and give it to the AI. So here's another prompt that'll help you just dial
00:26:21 this in. You can say, I need to close more deals from sales calls, analyze these call transcripts
00:26:27 and pull out all the objections that the prospect raises, group common ones together, then propose
00:26:32 the best way to handle each objection, build the script for me. And here's a pro tip, you want to
00:26:37 give it both close one and close lost calls so the script reflects what actually moved the deal
00:26:42 forward. So now we built each agent individually, but the challenge is, is that there's some things
00:26:47 that end up in the inbox that is not used in the marketing. And then you have your marketing agents
00:26:51 going out there, finding leads, and it might be able to find some pockets of problems that it
00:26:55 should be feeding to your salesperson to use in the sales call. And then when you onboard, like
00:26:59 I mentioned, the sales conversation should be brought forward to the person onboarding the new
00:27:03 customer. But right now they don't really talk to each other. To solve this, we need all the agents
00:27:08 to share one brain each agent knows how to do its job incredibly well but it's learning about your
00:27:16 business and we have to share it in one place as we create other agents in our business it doesn't
00:27:21 have to relearn every time the most valuable thing in your business is the data equity that you're
00:27:28 creating and collecting on how you do things different than everybody else but right now
00:27:32 it's like individual brains we need that one brain imagine if every doctor in the world that works on
00:27:38 a specific type of surgery took everything they learned around how they do that surgery and it
00:27:43 was available to every other doctor through a click of a button that is what we're talking
00:27:48 about one shared brain for all doctors doing a specific type of surgery and the truth is all
00:27:54 these agents and new agents you create are only as good as the data that it knows about your business
00:27:59 and that data equity has to live in that brain that's why we want to create the one brain so
00:28:03 here's how we do it first we have to put your whole business in one place i'm talking your
00:28:07 email your calendar your calls your crm all of your docs that's your data equity second
00:28:12 we want to connect every agent to that brain whatever tool you use or run on it could be
00:28:17 claude chadipetti or grok this way those bots immediately get context on your business now i
00:28:22 use apex which is our own platform but you can use whatever you want but here's how you structure it
00:28:28 you have to start with one handoff and add more once it works so for example every friday you
00:28:33 have your sales agent save all the top objections to the brain marketing agent reads those and turns
00:28:38 those into next week's topics for you to create content around and on monday you check the list
00:28:43 that it created to drive your marketing that's what every sales call tells you essentially what
00:28:47 your buyer is worried about and now it turns all that into next week's videos instead of getting
00:28:51 lost when you hang up i use it for having my agents do follow-up emails because now the inbox
00:28:57 agent has my voice and it can learn how I like to write. And then it can use that to write follow up
00:29:01 emails for the sales agent. I can have my delivery agent monitor some of the sales things that I've
00:29:06 been saying and commitments I'm making to the customers to create a punch list of things we
00:29:10 got to make sure we deliver on. So then when the new customer comes on board, we don't drop the
00:29:13 ball, right? These are simple yet powerful things to create a unified and complete experience for
00:29:19 your new customers. And look, you can build that brain yourself. And I did a whole video on how to
00:29:23 do that, but it takes time and also it can break. That's actually why I built Apex. It's one of my
00:29:28 companies at Martell Ventures because it holds every meeting, every email, every decision all
00:29:32 in one place and it extracts the insights into what I'm doing and allows it to be shared across
00:29:36 all my agents. When people see me move from Claude now to ChadGBT, now to GrokBot, the reason I can
00:29:42 do that really easily is because my brain, my Apex system just connects to it and it has everything
00:29:47 that I've ever given it from past systems.
00:29:50 Like all these AI tools that are launching every single week,
00:29:53 those can change, but my brain, it stays the same.
00:29:57 And look, if you wanna check it out,
00:29:58 just scan the QR code on this page
00:29:59 or click the link in the description.
00:30:01 So this is how we use a replacement ladder.
00:30:03 Once we find a problem that is a painkiller
00:30:05 to slowly buy back our time as we get new customers
00:30:09 so that we can over deliver.
00:30:10 The one thing the AI will never be able to do
00:30:13 is have a vision for what you wanna create.
00:30:15 Having a vision big enough for all your dreams
00:30:18 and goals to fit inside of, that is the thing.
00:30:20 To the degree that you can do that,
00:30:23 you will pull that future into your life
00:30:25 and this replacement ladder AI edition
00:30:27 will just help you get there faster.
00:30:29 That's why the future will be full of people,
00:30:31 one person with a whole lot of agents
00:30:33 that are able to run multi-million dollar companies
00:30:35 without a bunch of people.
00:30:37 So leave me a comment below
00:30:38 and let me know what stood out for you today.
00:30:40 What did you need to hear most?
00:30:41 What got you most excited?
00:30:43 I need everybody to leave me a comment below.
00:30:45 And remember, if you want my AI company operating system
00:30:47 to roll out AI across every part of your business,
00:30:50 just scan the QR code on the screen
00:30:52 or click the link in the description
00:30:53 and I'll send it over for free.
00:30:55 And if you wanna build AI agents
00:30:56 for every other part of your life,
00:30:58 just click this video and I'll see you on the other side.