Many people say that AI will take our jobs, and other people are saying, "No, it's those people who use AI that's going to take our jobs." Me, as a business owner, I really don't care about any side of this debate. What I care about is what can I use AI and automation in within my business operations so I make more money, generate more revenue, save more time, and cut more expenses. That's all I care about as a business. So, in this video, I want to show you exactly how I use AI to one of the most important processes that I have, which is creating proposals and offer documents, because that's the money maker within our business.
And I take you through the framework where I thought about this is it worth it or not, and should I use AI and automation over here? And I'm going to give you that framework by the end of this video. I'll show you how exactly it works so you can implement it as well within your business. That being said, let's get into the screen.
So, the $83,000 per month offer document, by the way, this works for proposals as well. So, we're going to cover three things in this video. The first one is my framework for seeing any business operation or task, whether it's worth the ROI in terms of reducing the cost or increasing the revenue or saving us time and adding AI or automation to it, and where does it fit within the whole business model that we run. Point number two is my thinking in step-by-step process on what do I consider, what are the challenges, what are the design factors over here, whether I'm going to do this or not, and finally, how can I automate this?
And I'm going to show you a full demo on N8N of how I've done this, and we're going to just run it live when you guys see exactly how to do it for yourself by the end of this video. So, just here very quickly showing you that I've had many versions of the offer document. For example, version one looked something like this. It was plain text in a Google Doc.
Then I added some colors to it and changed some words from 5,000, for example, to 8 10,000, made it bold, and so on. So, each iteration I've done throughout the years has been based on the market feedback. What are people telling me about this offer document? And that's exactly why I made this iterations for it to convert more people.
So, more eyeballs needed to be converted to sales, and this is the iterations or the feedback I got. Then I started changing things and steps within the offer. Then I added highlights. I changed the spots.
I added your private invitation over here just to make it more exclusive. I changed it from 12 weeks to 16 weeks. I just designed it better. Then it became more complex even for me, but simpler for the client and converting better, is that I added more, let's say, labels over here.
I added numbers. I added sections. And I added more words that based on the questions I was getting. For example, when I was sending this offer doc, some people asked me questions about the pricing.
Some people asked me questions about where does this fit in exactly within my model. Every feedback was added within this offer document until I got to a final offer document, as you can see, adding pictures, highlighting things, adding codes here and there, adding a certain message, a very personal message at the beginning, and so on, until we went to from Google Docs until Gamma Docs, which I'm what I'm using right now, gamma.doc app, if you know that application, it's good, but I'm not affiliated affiliated with them in any way, but this is what we I use right now. The main point is this one through a lot of iteration and process and during the years of feedback, getting what the clients were telling me, and so on, and we've got to this final form. So, what I want to do in this video is take that final form and show you how I would AI-ify it.
How would I would use AI to do this, and why is this important? I'm going to tell you in a bit. Beca- Before I want to explain to you how does this model work? How does my model work, and why is an offer in this form very important for us?
So, the model that I'm running right now, the world-building model, works as follows. We people are right now in the world, they're strangers, they're moving around, they don't know who you are, who you exist, and so on. So, you need to find a way to get those strangers to become leads. And this is the very first part of the model.
And a lead basically is someone telling you that you have permission to pitch me, to talk to me, to tell me anything you want. To basically, I give you permission to talk to me. Tell me what you have. That's a lead.
Then, we need to take that lead into a customer, which is we need that person to transact in any small amount just to show buying intent, whether they're serious in paying us money or they're just freebie seekers. They're just here for the free PDFs or whatever you offer, the lead magnets in in particular. So, the lead becomes a customer. A customer then becomes a client here once they see an offer document which they like, and then they become a client.
And then a client then become a case study, which is the results you get which you're going to feed back to strangers so they become leads and customers, clients, and case studies. So, where does this in particular fit in? The offer document. Let me tell you.
So, an offer document or an offer in general is what takes any individual between these stages. It takes a lead to a customer, you're making an offer over here to make them a customer. You're making them another offer to take a customer to a client. And sometimes the lead is warm enough, they watched the YouTube videos of you, they know who you are, they trust you, and maybe someone referred them to you or whatever it is, then that person can see the offer document and also then become a client.
And then this is a very special case for me because I show people how to write offers. And based on what has been working with within my business, I show my clients exactly that. And within the first 7 to 14 days, we work on creating their offer so they can market it as soon as possible. That's one of the quick wins.
So, in this very particular case, if I was able to automate this, since I have the final structure based on the feedback I've I've gotten so far. If I've been able to automate this, maybe during the onboarding, when someone comes and signs up to Seolia, and during the onboarding I ask some questions, and from the onboarding form I redirect that information that they've given me and create an offer document where they have an offer document as soon as they sign up. And then on our first call, what we could do is, "Hey, now we can go market it and this is the plan." So, the time to value that quick win becomes so small because right now is 7 to 14 days until we set a meeting, and then we sit on that meeting and and this is fast, by the way. 7 to 14 days is fast, but imagine it happening in minutes. And this is the main point of this specialty.
So, the question here is is there value in what we're doing right now? There is value 100% as you can see because offer move people through the stages, and if we could do this fast, and we could get a more personalized offer for every individual, and you can do it for your clients if you're selling to your clients and teaching them how to create offers, this is a very nice use case. That being said, this is exactly what I do every single time someone tells me, "Let's add AI to this. Let's automate this." I think about what is the ROI of doing this.
Is it worth it or are we following a shiny object? And let me tell you why. Because many people in 2023, 2024 were saying, "Let's add chatbots to our websites." And you you think here, okay, chatbots are amazing. It's a very nice tool to have for your website.
People come on your website, and you get they get this pop-up at the bottom, they chat with this person, and so on. But what is a chatbot essentially? It's a glorified FAQ. So, you're just adding complexity to the system just to look cool by having a chatbot where you could put all that information in the FAQ where they can toggle by themselves.
So, that does not have much of an ROI in my opinion. What has more ROI is things that pass this test, which I want to show you exactly. So, I'm always asking this question, are we automating this or is it a shiny object? That's the main two questions I'm asking and also I have this quote always at the back of my head.
Hard human tasks are easy for machines and easy human tasks are hard for machines and I've learned this the hard way when I tried to figure out this very easy thing we do as humans to make a machine do it and it was very extremely hard for a machine to do it because we use our visuals, our head and our thinking and decision-making and so on. So, what is an offer doc essentially? It's different parts as we talked about and some parts are fixed variables and for example, the structure over here and some parts are variable which are changing. For example, this part is changing.
The title is changing for every individual but the title stays at the top. Then you have this intro statement. It's changing for every individual but it stays over here. So, the structure stays the same but the text would then change it changes.
So, this is very important because when we're doing this RY's test, we need to see our options and whether it's worth it. So, in this case, we have two options. We can just use ChatGPT, Claude or any LLM and prompt it and say, "Hey, this is an offer doc Adam created. Can you create another one similar to this?" This is fast.
This is easy. This is it it gives it flexibility to its thing but I've noticed and I've tested this out and I've tried it with the a couple of clients. It lacks consistent structure. And it also loses context after many prompts.
But what if we could make it more structured but use only because remember, this offer doc as I I just explained, it's a structure. It's a structured format but we only use AI in certain segments over here to fix the text based on the input of the user. So, the pro the pros of this the advantage of this is that it's we segregate where automation or AI is done and we get consistent outputs but the problem here is more complex, of course. And for the less tech-savvy who don't know much on technology and so on, they may find it hard if they don't have a form attached to it that's very simple to just click and you get a WhatsApp message or an email with the offer document that comes to the user interface.
And it requires a front work. Where in ChatGPT, you just drop the document and then just ask it, "Create for me this offer document." Out of these two options, the NA10 option is the one we're going to follow because one, we're following a fixed structure, we want consistent outputs, so that's point number one. And point number two is that it's a proven system, so we don't want to touch it much and I give AI the flexibility to play around and do whatever it does because that's going to reduce our conversion rate. And if we reduce our conversion rate, what happens?
We reduce sales, and that affects our business, and we're not happy because we're making less money all the all of a sudden. So, this is in terms of the options. We we checked our options and we chose this option out of everything. Then, I thought about, "Okay, is this a shiny object?
Is it Is it worth it to do it for my clients or not? The time and effort I'm going to put." First of all, I I ask myself all these questions, and I call this a shiny object net where if it doesn't pass these questions or if something looks wrong within these questions over here, I just stop doing it. So, the question number one is, "Is this task complex or easy?" Remember, easy tasks are hard for machines, hard tasks are easy for machines. So, is this task complex or not?
It's a medium complexity complexity task given that you need to know the know-how and the structure, a proven structure, and also some feedback where you need to make certain words specific, writing sales copy, and so on. So, it's a medium, let's say in terms of task complexity, it passes that. How much time does this take? It takes much more time for new individuals when they join because they have no clue how to create an offer.
They've read a book by Alex Hormozi and so on. Okay, that's the theory behind it, but how do you make this tangible and present it to the client? It is through a proposal, which looks like a corporate document or it's very professional, or do you want to make it an offer doc where it speaks directly to me, to my heart, and it converts me, and makes me want to give you money because you talk to my pain points. So, this is essentially, if you're experienced, it takes 1 to 3 hours, and I do it once or twice a month, depending if I'm I'm launching a new offer or I'm updating offer based on the feedback I got from the last launch I did at the time.
I launch offers every month, and particularly at the end of the month. So, it's me updating. But, at the very beginning, it takes time to do this if you don't know what you're doing. And task confidentiality is not confidential because this document is going to be public anyways.
You're going to be sharing it with your prospects to convert them into clients. So, it's not that confidential. And who's owner of this? You can have a team member do it, but I suggest you do it at the beginning and keep on doing it because it does not take much time, and it's a high-leverage task within your business.
So, this is important. So, founder mainly. And what if this task fails? You already have your offer document, and if it fails, you're not going to lose your business because you already have an offer document.
You just need to update it and so on, and you know the know-how. So, it's not that bad if it fails. So, it passes the shiny object test that we're not just wasting our time just to look cool and add AI to our systems, but it's actually worth creating an AI out of it. So, this is essentially the checklist I go through.
This is the business ROI checklist. Three questions I ask. Okay, now we know our option. Now we know what to do, and we know how we're going to do it, and it passed the save the shiny object net.
Now, what are we going to do? So, does it increase revenue for the business? That's the first question I ask. Does it free up time?
And does it decrease cost? So, increase revenue, more leads, more sales, more retention, more up-sells, a clear metric and outcome. And does it free up time? Does it free up my time?
Less repetitive tasks? Is it a proven task that I'm doing or is it an innovation? For example, if it's an innovation, I'm trying I'm testing something new. I'm never going to make an AI do that at the very beginning.
I need to test it out myself. If it's a proven task, okay, come let's automate certain parts of it and then add AI to other parts as well. And I'm asking, can I decrease the cost of it? Is it Is this a task someone else is doing or not?
And do I need less software licenses and so on? And does it decrease expense in any way? So, in any task, this is very generic, but in any task, this is what I'm looking at. And very, very important, the final question I ask is, how do you do it without affecting current business operations?
So, if I'm sending offer documents every single day to someone, how can I do it now? How can I add this automation and or AI automation in and not affect me sending offer documents? This is very important because you don't want to lose money in the process of integrating AI. So, if it passes all of this over here, then this task is worth creating an AI out of it or an automation out of it and so on.
But, you need to decide how much of it is going to be manual and how much of it is going to be automated. Where Where do Where does AI fit in? Where does automation fit in? What's is manual work and so on.
That's very, very important, by the way. And then you're ready to go and I'm going to show you in a bit how it works, but then you're ready to go and make this task an AI. So, the system. Once I went through that process, I asked myself this.
What does the system look like? Like, how can I now make all this analysis, theory, and so on and doesn't take time, by the way. Don't Don't think this is going to take me days to figure out if whether this task is worth it or not. Just go through that with five within 5 minutes, you have your answer.
But, now how How we do this? So, you think about it. Okay, I need a template for an offer document. Amazing, I have that.
Then, I'm going to use it on Google Docs most probably at the very beginning. So, I have Google Docs as well. I need a template on Google Docs. Okay, I need to create a template on Google Docs.
What part of the template is going to be variable? What text will be variable? And what text will be fixed? That's another part you're going to solve for.
And then you're going to start thinking, "Okay, where does the AI work? On what elements does the AI work?" And so on. That's something else you're going to talk about. And then you write these four steps.
You say, "Okay, the user will come or I'm going to come and fill up a form." After a sales call is most particularly like I I'm I'm talking with you on a sales call and we finish the sales call and as you're giving me answers, I'm filling up this form like it's behind the scenes. And as soon as we end the sales call at the end of the sales call, what I can tell you is that, "Hey, your offer document is ready. Here it is. Let me present it to you." Completely personalized to what you exactly told me.
I can even pull the transcript of the sales calls on this field and we're just giving ideas over here. Or when I want to update my offer, this is a use case for it. So, I need that form to accompany that kind of scenarios as well. Or if I want to go to a new client during the onboarding where it can also include the information from the onboarding and create an offer document for that person.
This is a special case for my case, but it may work for your case as well. So, this is just the steps that I think about. Then I say, "Okay, how can I templatize this to work for all scenarios?" And then third point is how can I use AI and which AI should I use? Which model should I use?
And so on. And how can I replace the placeholders within the document to actually do this? This These are the questions that I ask myself essentially. And then, as soon as I ask these these questions, I'm telling you live right here because I went through this process.
These are immediate questions that came up. So, the first question was what are the inputs of the form? Like what are the questions I'm going to ask to make it easy for the user to fill without feeling overwhelmed or wasting a lot of time and getting the most information. Like, what is the 90%?
I'm not going for 100% here. What is the 90% of the information I need with the least amount of questions so I can fill up this offer document. That's point number one. Point number two is what should we have as placeholder?
What's fixed and what's variable as we mentioned within the offer doc? And how much is the input output ratio? Meaning that how many inputs will I get from the user? Will I get Will I ask the user three questions for example and get 10 outputs to it?
Like the the bigger this is, the better it is because I'm telling the user, "Okay, you just give me you answer just one question. I'm going to give you an offer document." That's the best case that you can do. Or just going mind reading mode and I just read your mind and I create an offer, but that's kind of in the future, neuralink and so on. But but if you think about it, what's the And if you think about this scenario, what's the least amount of questions we can ask to get the most amount of outputs, which is useful for our case?
And then what are the prompts we're going to use? What is the user prompt? What is the system prompt? What are we going to identify the agent or whatever we're going to use as and so on?
Or are we going to use a message model or an agent? So, if in N 810, I'm going to show you in a bit, you have agents and you have message to me you can message models. So, when you message a model, what happened is that the output can be very structured compared to an agent where the agent has more creativity and more freedom to think for itself and choose the right path for you. In this case, if I was creating a content creation bot or workflow over here, I would have used an agent.
But in this case, we're going to use a message model because we want more structured outputs as we saw the offer document and so on. And what sort of platforms we're going to use is going to be Google Documents. So, essentially after you go all this through all this process, you come up with this simple workflow over here where it looks like this. I'm going to go into my screen now.
I'm going to go and show you exactly how and we're going to run this together. But, this is a simple workflow you get. The user submits a form. It goes through an AI.
It changes certain part of that form. It creates a copy of a Google Drive document because you have the template over there. And then, it just fills out that new document it just copied with the information that it creates from the AI, and then you have your offer document. That being said, let's go into N8N and see how this works.
So, here's N8N. As we mentioned these questions, what are the inputs of the forms? I'm going to show you in a bit what the inputs are. What should we have as placeholders?
I have the document over here. I'm going to show you the template exactly how it is. And what are the prompts that we're going to use? I'm going to show you the prompts in this box over here, a node we call it an N8N.
And what are platforms we use are the Google Drive and Google Docs as you can see. So, for the form question, what are the inputs? Let's look at that. This is the inputs of the form.
What is the name of your offer? This is the name. You're going to see where it's going to go in terms of the placeholder. What is the How Who do you help?
Pick you Like I put placeholders as well. Who do you help? And what results do they want? How do you deliver it?
How long is the program? How much time do you need per day or per week for this program that you're running? And any price or payment plan you have in your mind right now? Do you have any social proof or not?
This sets the tone of the offer document itself. So, in terms of inputs, these are the inputs and we get a JSON file over here as you can see. It looks something similar, but I'm going to run through it you live as well so you do it again. But, this is something similar.
You're going to get something similar in terms of the structure. It is the elements that we have. So, remember input output ratio. These are the inputs that we're getting right now.
So, this is in terms of the first question that we answered. The second question is, what is the template and what is the structure if you're going to do it for a proposal or whatever it is. This is the file that we look into which is it has the template itself. And this is the file that's going to copy every single time that's going to create an offer document for anyone.
And this is the template over here. So, it has placeholders, the program name, your private invitation, exactly how we did it, the outcome statement, and intro statement, one big idea one, the description, big idea two, description, big idea three, description, then a summary looping back to this outcome over here so the reader does not forget or the outcome or why they're reading this particular piece of offer and why they're here. It helps with conversion, by the way. Then you look at the timeline.
This is what we're going to do as point number one, point number two, point number three. Here's descriptions of it. This is a support structure. And this is how long it'll take till you get to the profit or whatever you want, the value you're here for.
And then you look at the investment, which is the payment plans or whatever you whatever it costs. Then a quick disclaimer over here if you have guarantees or not. And the next steps in terms of the summary and their CTA to join. So, this essentially is the document or the template that I've created.
And look at this. I just want you to notice this. We had around six inputs from the form. Yeah, over here. 1 2 3 4 5 6 7 8 8 inputs.
And then it outputs a couple of things. 1 2 3 4 5 6 It's more than 8. The point is it's more than 8. So, this is the input to output output ratio. The less input you have, the more outputs you get which are accurate and precise 9 If you have a 90% accuracy of the you're doing it the way you want, then you are ready to go and you're making the life of the user much, much easier in in terms of using whatever you're creating.
So, this is point number one. As soon as this happens, you click on the next node, which is the OpenAI model starts working. So, we're going to see how it works. It has a bunch of a bunch of prompts over here.
So, the system prompt just identifies what this AI is. Like you just give it an identifier. In In our case, you are an offer doc generator, a high conversion copywriting assistant trained to do Google Docs and so on and so on. The second point over here is and you give it a bunch of rules.
The second point over here is what's the task at at hand. So, you write the task, what is the section, what is the instructions, and so on. And then the third point here, you give it an example, which is the example of your offer document. For example, you take your offer document, throw it into ChatGPT, tell it, "Okay, write it in XML for me." Throw it in here as an example.
And the fourth part is just matching these or mapping these to inputs for it for the AI to use in order to create this for you. So, as you can see, we inputted this. This is the output over here. This is the output we got.
We got the program name because the input is program name, the outcome statement, which is based on these two, somewhere here. We aim to help you get to $20,000 per month with less than 4 hours, with no sales calls, and a consistent client base that builds your authority. So, this is me just mentioning what do people want. Intro, leverage the next 16 weeks with three core core pillars to redefine your business growth.
I'm going to show you how it looks like in the document. So then, big idea number one, big idea number two, big idea number three, and then phase one, this is what we're going to do. Phase two, we're going to achieve this. Phase three, we're going to scale and sustain.
And this is the time to profit. This is the pricing terms. This is the disclaimer. This is the summary, and so on.
Amazing. So, it just filled out those spots that we want. Now, what's going to happen is the next note is that we don't want to keep on replacing the same document over and over again because we can't do that to begin with because of the placeholders. So, the best option you could do is to copy this document, right?
So, you come here, click on this, it should copy the document that's over there. So, you can see the document ID over here. And then you just rename it as the name of the document, private invite, and put the date next to it. And if you go here, we refresh the screen, it's over there.
As you can see, see our private invite April 15 and right now the timing. So, if you open this, it looks like something as you saw in the template with the placeholders ready to be replaced. And then for the fourth point over here, you just click on this, you have the ID, you just pull out the ID of this document, put it here, then you put test up. Just 1 second, it's loading the field.
This is taking time because of the placeholders basically. We had a lot of placeholders, and I just added all of them in. So, now it's done. Let's go check the document.
There it is. Señor, your private invitation. We aim to do this, the second statement. This is point number one, point number two, point number three.
It has an outcome at the end of each one. Results consistent, client flow, and increased authority positioning you as the go-to ex- expert. This framework guarantees efficient operations. And then this is what gets you to 20,000 per month.
And then this is what we're going to do in in point number one is that we're going to first do your client attraction system. We're going to do whatever we're going to do together. Then we're going to jump to systemizing everything. Then we're going to scale everything.
And within weeks you're going to hit 20,000 per month consistently. To ensure that you are enrolled and so on, we do a 20 $250 per week for 16 weeks. This is the guarantees, and this is the next step. So, phase one, phase two, phase three.
You just remind them of the outcome that they're here for again, and then tell us what to do, and email us at this email. By the way, this is not my email, but just I'm telling you this is how it's done. Amazing. So, this is a quick demo of how this runs very, very simply.
And let me tell you my thoughts about it. So, first we started by having our business hat on and thinking whether this is worth it to add AI to an automate or not. And then we had our builders hat on and saying, "Okay, how are we going to do this? What are the challenges?
How can we make it easy for the user?" and so on. And now let's analyze it again as a business owner and sitting in the shoes of our clients as well. How do they see everything? So, a couple of points.
Did it save us time in this scenario? It saved us a lot of time. Did it save the client that's going to come to me during the onboarding process time? It saved that person a lot of time as well and headache and thinking about the right offer and so on.
It at least gave them a guideline where during our first kickoff call or our first call or consultation call together, we can talk about okay, this is the offer we got. Let's do some tweaks here and there and we're going to get into the results. So, it did save time and it was highly valuable for the other person and for myself as well when I want to create new offers or I'm thinking about okay, I want to run a customer offer, let's say I have a bunch of leads. I want to turn those leads into customers.
How can I do that fast? I have this idea in mind. Let me try this. Fill out this form.
Let's see what AI gives me and so on. I think about it. A couple of things that I would fix about this model over here is one, I would test different types of LLMs. So, right now I'm just testing it on OpenAI, but Claude 3.7 as we know is much better for copywriting and having that tone and emotional brand voice that you have within the copy.
So, that's one thing I would test 100%. Another thing I would test is how easy it is for the users to actually use it. And how can I reduce the number of questions the users are asked to get this offer document? And what else can I create out of this offer document?
And then I'd ask okay, can I create a presentation out of it? Can I use transcripts on call out of it? Can I make it look nicer? So, these are all secondary things.
But, the main primary goal of it being a high ROI and worth it to do so is exactly what I told you. And then on top of this, what I'm planning to do next is how can I make a nice interface, a front-end interface since we're talking about vibe coding and all these trends going on right now and vibe marketing right now. So, if you have this all these things going on with Bolt as well and lovable dev and Replit and all these nice tools that are coming out, maybe you could test and web hook it to this or connected, let's say in less technical terms, connected to this N8N workflow, where you just give them that onboarding experience of as soon as they click check or buy on on Stripe, they get redirected to a page where it gives us this nice animation that's personalized, takes them into space because I like space themed stuff, and then gives them this form which looks really nice with a space themed UI UX thing. They fill out the form, they click next, they get to an offer document in their email and so on.
So, imagine this onboarding experience and for the next 7 days. They will this and this is a business question. Will this increase retention? Will this increase uh let's say that customer staying for longer, the lifetime value of the customer?
Will it increase the customer actually loving your product and referring you, advocating about your product and service to other people? I think it will because I'm doing it right now manually, but what if we could do more things behind the scenes and leveraging whatever information we have in terms of inputs and giving them more outputs they really want and in different directions. So, this is my thought process right now. It is 100% worth it and I think you should apply the same framework to every task that you want to add AI or automate.