The Truth Behind Selling AI Automations To Small Businesses! #n8n #agents

Hi everybody, welcome to this YouTube channel. In this video, I want to discuss something with you that I recently shared over uh Reddit community. I shared a post that went viral.

It got almost 70,000 views and a lot of people were commenting on it. It was a really controversial kind of topic that nobody's trying to talk about, but because it went so viral, it made me believe that a lot of people are resonating with what I said, and I just wanted to share it on YouTube. If somebody else was in the same space as me and fell into the same trap, I would love for you guys to know this.

Before we move forward, my name is Sup. I've been building and selling AI solutions, part of it as a full-time job in a corporate and recently also as a freelancer and then as an AI automation agency. So, I've seen everything.

I've seen how AI should be developed, how AI should be sold. And I have to say recently we just entered the nonsense area where you can build an AI agent that took you five minutes and you charge people €2,000 for. And this is obviously is not sustainable and is not part of the truth.

So I would love here to discuss with you what I said in that post. Why it got so much attention. Why did people resonate with this post so much and also tell you what is actually happening in real life because I feel everybody's just lying and everybody is just posting some things on social media just so they can increase their YouTube channel or whatever.

And all these statements like 100k a month AI automation agency. I would love to clarify what that means and how these people in my opinion are actually getting their money. and it's not through AI automations.

Let's go to the post. In the post, I mentioned like is anybody making actually money using NN and AI automations or am I the only one who's not getting it? So that was basically the post I wrote that in my experience I've been trying to sell for small businesses.

So this dream that they sell you like yeah create an AI automation and start selling it to restaurants or local businesses and in my experience it has been a very challenging thing to do. First of all, small businesses are not so tech-savvy. So, it's really hard to integrate such a high sophisticated tech stack into their original tech stack.

I mean, with NN and it has some integrations with Gmail, with Google Cloud, Google Calendar, Google Drve. So, it makes it much easier to do like a plug-and-play solution, but some of them really don't care. And of course, they tell you like don't sell with AI, try to sell with the results.

So that's what everybody should do and that's what I've also been doing. But I just saw that the readiness wasn't there. So even if you sell with the results like hey this would get you more clients blah blah blah people find it still very difficult to trust what is what is it doing to see how they can add it to their business.

So we always tried to add like some human in the loop approach give them the control the feeling of the control and that sometimes it did help do a cell but it was never these results that these people are mentioning and it's never so easy as everybody's saying of course everything worthwhile is going to be very difficult to to do and I'm not saying that you shouldn't do this but I just want you to know that it's going to be very difficult to first of all create something and try to sell it or doing like some sort of AI consultancy especially if you're just starting like nobody knows you, nobody trusts you yet. Of course, you need testimonials and all this kind of stuff. But I just want to clarify that yeah, getting 100K a month or these kind of results that people keep trying to sell to you is just not so realistic in my opinion.

And I want to share here with you what is actually realistic, what I managed to achieve and what I know other actual agency owners here in Europe or in the US have been able to achieve and what kind of stuff actually improved their performance and what stuff didn't work out. So this post since it got like 70k views there was a lot of opinions but so far I would say 99% of the people agreed with that statement that most of these content creators actually are not making their money through AI. So I got a lot of comments on Reddit that said people who make money in a gold rush are the ones who are selling the shovels.

And I think that is very correct especially in this case. I think the people who are making money with these AI automations are the AI influencers and content creators. So when somebody tells you I'm making 100K a month, it's not through some sort of building solutions and selling that.

It's just not so realistic. Especially that I know that they don't even have such a large team. Realistically, maybe you can service four, five clients that have a very small ticket.

But like a small ticket, if you want to like do four clients, realistically, you can make maybe $40,000 a month, which is still a lot of money. But I I guess this as a oneman person agency, this is going to be like the limit. And then you need to start hiring and then scale linearly.

Even if you do that, then it would require you so much time or just have such an insane lead magnet or lead generation approach that you have so many people in your lead funnel and like customers that are waiting to be serviced to be able to scale to like 100k. But what I saw is that most of these people are making the money either through YouTube school communities which is technically like you build a product and you sell. If you did like a school community um if you don't I'm familiar with the school.

uh school is like a platform where you build communities and then you can monetize it. So some of them are charging around $80 per seat per month and they usually do like the first 500 for free and then after that you have to pay and then the price keep increasing. So that's actually how you can reach 100k or 50k easily bec not easily because obviously building such a community takes a long time but it's a much more efficient way to generate money and cash flow than servicing and dealing with clients.

And I think a lot of people and influencers noticed that yeah the content game and the educational game is much more profitable and much less headache than the actual AI and software development game because also most of them actually don't talk anything about what actually goes on in real life. So all of these influencers say oh I built this AI automation. They don't even talk about contracts.

They don't even talk about what happens under the hood. They don't even talk about like these data agreements that you sign with a contract because they want to share data with you or if you're dealing with personal data of a client. This is what really fascinated me.

Like when I started selling to local businesses, I was confronted with about 10 20 different topics that I didn't even hear about and I didn't even know existed. And I was like, okay, well, I can't do this because of in Europe we have GDPR. I'm sure whenever you're hearing this, you also have something similar.

So this is data protection laws that you need to make sure that you're not handling personal data in a wrong way or at least be upfront about how you're handling the personal data. We also have the AU AI act that you have to take into consideration where the data is being processed. For example, if you're developing in Europe, you cannot just send data to the US.

So all these tools that are US-based like make. com, Vapi, all this kind of stuff, it makes it really difficult to install to European clients. Of course, this video I'm not going to make it just Europeanbased, but this is my point.

Like all these technical details are not really being uh mentioned. The only guy that mentioned something interesting was Liam. Some people say he's the founder of the AAA model who actually at one of these videos posted the contract.

He said, "This is what you need. You need a contract to do like a service for a client. " And he posted like the template that looks quite good.

And he was the only person who actually did that. And I think I believe that he actually had clients and he's been developing with his team a lot of stuff but he was also very upfront that a lot of chunk of his money is coming from the AI education and the community. I mean his community is still for free but these courses he's doing all this attention he's getting through YouTube.

This is where the most money is coming. So I just want you to know like the next time you see someone posting about their gains and 100k a month and all this crap just know that it is possible that they making this money but not through this AI automations not what they preaching that they are doing. So just I want you to know that that being said I'm not saying there's no money to be made in NAT and automation and building AI agents.

there is and I've made some money doing that and I want to tell you how you can do this money specifically for AI agents for example. So the problem I see in AI agents is that they're still very unpredictable. They are super powerful.

They're getting even more powerful by the day. But you need to build a lot of trust in these systems and make sure that these systems perform reliably because the moment they do something very stupid, your clients are going to lose the trust and they're going to maybe stop working with you or have such a negative feeling about the AI. You as a developer or you as a system designer or as an agency, you need to make sure that you empower the humans to intervene at any point where they notice the AI is going crazy or it's just saying some really really stupid stuff.

that is possible and we need to tell people like hey this is possible that you might have some challenges recently I was having a consultation with a client where they said like I don't trust AI image generators anymore because the person was an influencer and trying to create like thumbnails and stuff like this and at some point the AI just went crazy and started giving them such a freakish large hand in every thumbnail and her experience was like I would never use AI again it was it was absolutely horrible and I spent money on it and I don't trust anymore. So these things can happen and you need to make sure as a designer to really minimize these events and try to have like some fallbacks systems that take that into account. That being said, over the past one year, I made some money selling these AI solutions.

But the main way I've managed to get this money is either by doing a consultation where I tell people how they can use AI in their business, do the implementation myself. So it was usually trying to solve a very specific problem for the business. Some sort of automation maybe LinkedIn based or outreach based or email sending.

It was more like in the marketing getting more clients based than just a fancy chatbot that does this and this and that. In my experience also automating customer support is not such a massive sell as everybody trying to say. I spoke to a lot of business owners and they told me for the same money I can just hire a virtual assistant somewhere and then they can just handle the whole inquiries and I don't need AI and I at least I can trust this person because it's an actual human sitting there and it turned out to be similar cost.

So until the point where the AI cost gets reduced dramatically that it makes sense to replace these virtual assistants with AI. It's still not as such a massive use case to use AI chatbots for automating customer support. The other place where it actually got a lot of attention was some process that can either automate some things.

For example, automating content generation process. This is personally has helped me save around 10 hours. So my team was super grateful when I built this for them.

And I was super grateful because now the content has been automatized and with one work I can have three times the reach without investing any time and effort or paying anybody. So that's actually one use case where you can see that yeah I create content once for example this YouTube video the content and the video gets transcribed automatically it gets generated different posts for LinkedIn Reddit even medium I'm now implementing an AI workflow that creates a medium article and then I can unfortunately you cannot post on medium anymore they close their API but you can at least copy this and paste it so maybe takes you 10 more minutes but then you have an actual article on medium you increase your reach your authority your brand whatever I think this is super helpful because it led to increase more traffic towards my channel especially in early stages where you don't have so many views you need to just post everywhere just add some awareness about what you're doing and then once you have decent following then maybe you don't need to do all these things or maybe you do cuz you don't know where your next client going to come from so that has been ways where I saw like okay AI does make sense I built also AI automations where somebody came with me with the idea and they said hey we want to build this And you can then choose either you build it with low code tools or N8N or workflows. Usually I started doing everything with N8N because it's European based as I mentioned in earlier videos.

It's super quick to build stuff prototype and actually go in production with them rather than just build it the usual way which is Python or custom code. You can now build it with nan or make. com and save a lot of time and then you get paid by the hour.

And usually the disadvantage is that you can finish much much quicker to build such a thing. So you cannot build such a massive number of hours. So maybe sometimes I finish a project within 20 hours even if it's a complicated project that previously it would have taken me 50 to 60 hours.

So that means it's good you're more efficient but also you're getting less hours. So what you can do ideally is to sell by the project. But that is the point that everybody's trying to sell you is like yeah just build this or copy this and sell it to other businesses.

Even me in some of the videos, I created this AI automation that I sold one to a wedding venue. And the automation was saving them time by responding to potential clients much much quicker. And there has been some studies that the quicker you respond to a client with an offer, the the most more likely it is that they going to convert.

So I use that technique to try to build these automations for wedding venue industries and try to sell it to them. And some wedding venues were interested. I had a lot of consultations, spoke to them.

One of them actually said, "Yes, would love to do this. " Of course, because it was the first client. I had to build it and test it for free.

That's just to gather some feedback. But then the problem wasn't selling to these businesses. Whenever I got them on a call, I managed to sell them.

I convinced them how important this is for their business, how it can generate them more clients. The problem with that specific use case, it was finding those clients. So, the the client finding process was very complicated.

the database of wedding venue owners wasn't so accessible. So you had to scrape a lot of websites, scrape a lot of blogs, trying to find the emails and send these emails. There was just a very finite number of wedding venues in Germany and nearby countries that even with like maybe 5% selling rate, it wouldn't have been such a big success like everybody's saying like yeah, sell to restaurants and you can make 10K a month.

It's not that simple. It's not that easy. And also the implementation usually takes a long time because you're doing a personal solution for each person that the system fronts has to change the integration the Google calendar actually one of the biggest bottleneck that I found was the onboarding how to tell them how are they going to be using your solution so usually you can either do it yourself so you have one account NN which is has to be either self- deployed or a corporate membership not just a normal membership but you need to go and talk to them and then say Hey, I want to be in control of all my clients data and then sell my client's data over my membership and then automate stuff for them.

So, this is possible, but then it's so much headache because you're going to be a data controller. You're going to be falling into some legal aspects that you need to do a lot of contracts with these people. So in my opinion the best way is let them create an NN account, do the workflow, build it for them and then you need to still onboard them how they use it, how they connect their Gmails, how they do everything.

Usually they give you access so you enter there and then you do everything and then you leave. Maybe you keep the credentials in case something breaks and also add some notification channels maybe to your Slack that if something fail that you get notified as an agency and then you go back and fix it. But it was not so scalable.

You know, it was it was a little bit difficult to see how can I scale this after 100 clients. Of course, you can charge a retainer and then it makes sense because you have monthly income. But a lot of things that I just saw like how do people are not mentioning how do you install the software to clients like nobody mentioned what is the process?

Do they need to create an account? Do I create an account? Do I sell under my name?

Also another big point was like you cannot use OpenAI in Europe because it sends the data to the US. So you need to create an Asia account and deploy these AI models in Europe zone and then do I give my client an API key? Do I keep the API key?

Do I manage a lot of really really a lot of headache that wasn't really mentioned? The last point that I also want to clarify is the idea of building these workflows and selling them. Here I see an advantage is that the time to build a product has been reduced by a lot of time.

So you can build something really really quickly. But then the problem of selling and marketing, it's still the same problem as any other SAS or any other products out there. So it's not as easy as like okay, I created this AI automation workflow that automate emails and then I'm going to call everybody and everybody's going to start saying yes, yes, yes, we want this.

We want to please install it for us. The process is still the same. Did you strike a need?

Do they care about it? Are they willing to pay for it? How much are they willing to pay for it?

This whole process of validating an idea is the same anywhere. If you start a startup, if you start a product, if you start something with anything, it's the same process. So you need to talk to customers first.

See what is their pain points. See if you can help them with the AI and then potentially maybe sell them after you get maybe three, four testimonials that were maybe for free because they are testing your bad product because it's still bad in the early phase. And then you have like a go to market strategy.

So the same process is any other startup you have to do here with anything or with any make. com workflows if you want to sell these and actually like scale it up as a like oh I just sold or you can do the developer so you just sit there until somebody come to you with an idea and then tell you build this for 10 20 30 hours I will pay you 50 to$100 or dollars an hour and then build it and that's it and then they manage it and then you leave but these are the only two options that I see and the second one is obviously more lucrative because you can technically scale. There's still a lot of customization for each client, but it's not so easy because it takes so much time until you find your market fit.

So, until you find who are your potential customers, are they interested? How do I talk to them? I've sent so many cold emails and a lot of them just went into spam.

A lot of them didn't even work. You need to learn about all these techniques of how to warm up your email before you start sending. So there's a lot of education that goes into creating and selling a company or selling a product that is not just building an end.

But that's what the online world just focus on is like I built this AI agent in half an hour and then people expect I'm going to be selling it for $2,000 or euros and it just doesn't work like this. And I wish it was. It would have been made my life easier.

I definitely fell for it because I thought like yeah like I I I can see this working out. And I've been now for 4 months trying to sell these things for different companies. And my honest honest opinion is that if you don't know people who have a business problem, it is so difficult to enter this field.

It's so difficult to do the number two to build something and selling it because you have no idea what is their challenges. You literally need to sit right next to somebody, see how they doing their work and then say, "Oh, you see that you're losing six hours doing this. I can automate it for you.

" Or if you are some marketing guru that you know how to get them more clients but now you can do it easier with AI then it makes sense that okay like I'm I have so much knowledge. I've worked in marketing for six years and I think now AI can do this easier than my manual process then okay build and sell this product. If you're just starting I see all a lot of people who are 18 years old or 20.

Majority of these AI automation agency owners are around 20 with no business experience. no business exposure, no contacts. I'm not judging you guys because it's me too.

I have been working for 5 years, but I haven't identified business problems that I can solve with AI. So, I was always relying on partnering up with people who might have some more industry insights than me so we can build together. This has worked, but also not so great because you're also relying on their execution.

So, can they sell these products quickly? Can they actually whatever they identified is that actually true and if you're not from that industry you don't know you just rely on their knowledge and I've had a lot of cases where I trusted people for like maybe 3 months we build a prototype we tried to sell and it turned out that it was a bust so like nobody cared about this product and most importantly these guys weren't able to sell it. So you also need to evaluate their skills.

And usually if you're entering as a technical person, like you're going to be building the tech, you rely that the other person is going to be doing the marketing and the sales. Can they actually do that? And if you don't know how to evaluate them, then it's not going to work out as well or you're going to trust the wrong kind of people.

So my experience has been not bad, but it has been different than what these people are saying. It's a definitely a thing that looks like starting a startup with all these challenges, especially if you want to go the selling route. If you want to go the just development route, the problems are a little bit different, which is the client acquisition.

How do clients know that you're doing this? And here my experience has been actually very awful because you can do things much much quicker and everybody can do it. So the competition is way high, the time is way low and the barrier to entry is also super low.

The only differentiator is that if you actually worked on production systems that you know how to build these systems effectively and then people would come to you like hey I've saw that you've added this human in the loop and you're doing logging and you're doing error capturing so you know what you're doing then I would love to work with you compared to someone who just copied some template and trying to push it onto people. This would be your advantage but it still doesn't justify you putting such a high label on it compared to all these other people and platforms like Upwork. It really breaks my heart when I see people saying like, "Yeah, I'm making 100K a month on Upwork.

" Like, this is absolutely nonsense. I have been working with Upwork. I've hired people from Upwork.

I went into consultation work with Upwork. Upwork really attracts the kind of people who don't care about quality. And this is just a generalization.

I'm not talking about everybody, but in my experience, people go there because they have a limited budget. They just want to get the job done. They don't care about the quality.

and they are willing to sacrifice some things just because they are limited on the budget. So from a client perspective when you go there you just want maybe reasonable price for maybe a reasonable communication because you just want to get or test or get your things up and running. From a developer perspective, when you apply, you see that there's like 50 other people applying and definitely maybe 90% of them are have ridiculous prices that it works for them because maybe in that country their cost of living is way cheaper so they can afford to do that and some of them have like maybe 30 40 clients at the same time.

So they don't give you as a client like the same attention, but it works for them, you know, and it's just really hard to compete with them, especially if you're like European based or somewhere where it doesn't make sense to you to do something for $20 an hour, you know. So Upwork has been absolutely a nightmare for me. I I definitely stopped looking for any gigs on Upwork.

And I think when I speak to my friends who have like a 40 person agency and they tell me Upwork is never have been a client source that we take seriously because people and corporates and like Fortune 500 companies, they don't go to Upwork to find people to do their work. They they have different strategies. So if you think that Upwork is the solver and the the way to go, I I'm really sorry to break it to you.

My experience has been super bad. That being said, I know some people who are also in Europe who managed to do Upwork work and they made some money with Upwork, like good money that they started their agency full-time, but they are not doing AI automations. They are doing typical data management, typical like hardcore work that requires a lot of skills and years to develop.

That's the difference. If you want to do just like a a chatbot or something like this, Upwork is not the great place. You're really competing with much cheaper and equally skilled people.

And in my experience, you would just rather spend your time doing something else. So that's about Upwork because I saw one influencer saying, "I made 100K on Upwork. " And I was like, "Dude, like this is absolutely nonsense.

Like how can you say that? " And it really broke my heart because all of the comments people were like really cheering for this person and saying, "Yeah, like you're great. " And me too.

I was cheering for for him and I was really like really excited about what he was doing. And then when I when he started talking about results that I was like I tried myself and I know these things don't work because I hired people from there. I know also the quality what to expect from people from Upwork.

It just doesn't work like that. This was in a nutshell my experience. You can still make money with any as a developer.

You can try to build and sell things with Nitn or AI or agents in general. I believe you can build SAS products but then the challenge is selling and marketing these products is it's not the building anymore and the people just focusing on hey you can build this for 5 minute and sell it for a,000 yes you can build this for a five minutes but then it's going to take you maybe two months to 6 months to sell whatever you built and this is what have been always the case like the selling and the marketing is always has been the difficult thing to do previously was also building a product was super difficult but now it's getting easier so I just wanted to share share this with you. If you're in a bad patch where maybe you're not getting clients or you keep seeing that everybody's posting some things online that makes you insecure and makes you feel like, "Oh, what am I doing wrong?

" My opinion is that you're not doing wrong. There's nothing you're doing wrong. You're you're on the right track.

Whether you're validating an idea, whether you're building something, I believe that you're going to make it and you're going to be there at some point. It just it takes time. These things take a long time and people just don't talk about that because it's not so popular and nobody want to listen like, "Oh yeah, I built this in 10 minutes and it took me a year to get my first client.

" It's not so catchy, right? So people don't really like to talk about this. It's very unfortunate.

But I don't want you to feel discouraged and maybe give up before just because somebody is telling you that you should be making 100K by now and you're not doing that. My opinion is that these guys have much different source of income than whatever they're trying to sell. That being said, thank you for listening.

Of course, this was a very maybe not controversial, but a very opinionated video just because I got this massive feedback from this our Reddit community that it made me believe that hey, a lot of people may be struggling with this and maybe they would like to hear some opinion of somebody who has been really hands-on in the in the space over the past one year and just to feel like it's okay to feel whatever you're feeling right now and whatever you're going through. I'm really sorry about the quality of this video. I I'm traveling right now and I'm just doing this from my laptop.

So, maybe it's not the best video quality for this video, but I it's been on my mind and I just wanted to let it out. And I hope you found value in this. And please let me know in the comments what do you think uh where is part of my mental thoughts have been incorrect.

So, feel free to challenge me. Feel free to to chat.