My AI Automation Agency FAILED

Two months ago, I decided to launch my AI automation agency and it didn't work out. So here is the story. So first of all, what is an AI automation agency? And so we have to think about what is an automation? So an automation is a sequence of steps that you configure for a specific task. So for example, let's say that you want to send an automated email sequence h after you receive a new lead. So someone want to apply for your service and then you want to contact this person automatically and do all sort of things. Well that could be an automation. Um there are a lot of software for doing this. So the older one is a zap year but there are other ones like meg. com or n10. Um and so those software have existed for a long time. But what has changed recently is that those software became more capable. So you can do all sort of crazy stuff like you can do to have a if statement, you can have a for loop and now you can have AI also in the steps. And so that means that we can do so much more automation and because of that we've seen on the market a new type of company that help other companies to automate the process by using uh those software. So we call this AI automation agency. So very practically what they do is that they will analyze the workflows of companies and they will create this process to uh to automate them and so they make money with uh two different ways. So first when an AI agency sells this automation there is a fixed fee. So it can be $5,000 $10,000 and then there is also supposed to be a recurring fee because uh very often those automation will be run by the automation uh company because there are all sort of things that can go wrong. So there is some monitoring to do. Uh there can be some customization. Maybe there is some API uh that you need to register for. And so just to make it simple, the AI agency just tell the customer, look, we're going to do it for you. So just pay us a fee every month. And of course, this fee is uh above what is their actual cost. So they actually have some recurring revenue. So on paper, this is a really nice business model. and some agency makes AI agency makes uh 10 20k 30k even 100k u so this is a very uh solid business model and so why did I choose to launch this AI automation agency well so for those who don't know my channel so couple of months ago I was known uh mainly for in the crypto space I used to be an education channel for blockchain developers so completely different from AI and then I decided to switch to AI because my interest have changed. Um and when I got into this space one of the first thing I wonder is okay well how can I niche down because you know AI is really big so I want to talk about one specific topic because otherwise it doesn't make any sense. So uh I studied on different YouTube channels how can we make money as an entrepreneur in AI and there are a couple of business model I saw but there's one that really got my attention it was the AAA model or AI automation agency and why it caught my attention because it seems to me that it was one of the easiest way to get into AI. um it was uh one of the fastest way also to get into AI. So for me it was very practical because I don't need to think too much for uh for months about what I want to do but right away um you know it's very fast to launch an AI automation agency really uh in you just need to study maybe for couple of weeks and then you can already offer your service. So, I thought that could be a great way to get started. And so, my long-term plan was to create this AI agency, wait to get results, at least to make $10,000 per month, and then once I reach this level, the step two would be to run a coaching business or have a private pet community where I help my student to uh achieve the same results. basically to launch and run their own AI agency. So that was the plan and uh and there are many other YouTube channels that do this. Uh so that was my inspiration. Um so after I decided uh to go for this plan, next thing I did was

to learn because I didn't know anything about uh AI automation. So I watched ton of YouTube tutorials and there's one channel in particular that I recommend, Nate AI automation. Great channel. uh he has this crazy 8 hours course and um I watch it in a uh at a 2x speed and uh yeah and that was great. So after those tutorials I also started to do um some real project because if you just follow tutorials you don't have enough real experience the experience comes when the experience and the confidence comes only after you've tried to do something and you will realize that contrary to the tutorial it it's not as smooth. Okay, maybe something is a bit outdated. Uh the graphical interface change, something change and then you have to do some research to fix the problem. And this is when you learn. But this is also when you acquire confidence. So this is what I did. And to um pick this automation, it was easy. I just picked some uh processes that I have in my business of a YouTuber. So I automated, for example, I run a job newsletter. So, and this is really boring for me to do because this is always the same step. So, I use I created an N810 automation for that. Now, it just takes me 10 minutes. Great use case. Actually, I've made a video on this if you are interested on my channel. Um, another automation I did was a tool for market research. So, I wanted to know if those private uh n uh those private school communities that we see for AI automation, I wanted to see if they're really profitable. So, I created a scrapper that every day goes on different communities, check the number of members, and then report this on a Google spreadsheet and then I can calculate the evolution of their profits and I can see if they are turning. Um and that was very interesting because some of these community they make ton of money like 100k 200k but I saw that actually a lot of them are not growing the journey is pretty important. Um and I was able to do this with uh an automation. Then um I did also some uh automation for my own community. So I have a free community for AI solopreneur by the way if you want to register follow the link down below. And so for this community basically I uh when people join my community they ask me I ask them some question and I want to record those question in a Google spreadsheet because once I've approved them I cannot see those answers anymore. So instead of doing it manually that was a great use case for an automation. So this is what I did and then the automation also add them directly to my uh email list. Um and I also did an automation to study to help me for content creation. So this automation check the some my some of my favorite YouTube channel and then it records uh some uh outlier videos or video that did pretty well. Um so couple of little automation I did and I use all of this to apply to job. So the next step once I felt confident in my skills, I wanted to find my first client and this is why I created an account on Upwork which is a freelancing platform. So I created a profile. Um it took me some time you know because you have a lot of information to fill and you don't really want to use AI too much because people can see when you use AI. So, you just want to use AI for the bullet points, but then you want to actually write yourself the text so that it looks like it was written by a human. I also created an introduction video. Um, and because that helps you to stand out a little bit because now customer can see how you look like. It's not just a random profile. And I created my portfolio on Upwork. So I used those project uh that I did for my own business and so I mentioned them in my portfolio. Um I also had to invest a little bit of money because nowadays Upwork is not free anymore. Uh if you want to apply to job you have to spend some credit. So the minimum is 20 a subscription of $20 per months. Um so I've already subscribed for two months. I spent $40. Um and uh I applied to five jobs only. I had three conversations. Um there was one good client. So that was a corporate client that possibly had there was some potential long-term collaboration uh and he was interested in uh in training. He wanted me to train their team for AI automations. And uh but the majority of the leads I had were really so and especially the last guy was uh it was just a

disaster. Basically I was uh so I didn't work with that guy. We're just talking you know potential lead and I could see the guy was completely Um I just asked him a couple of question to try to understand what he wants to do and he was completely unable to have like a normal conversation. So I just decided to ignore him and I didn't want to keep to keep working on uh on looking on Upwork really. Um so then I changed my strategy and I'm like okay well why don't I um leverage my platform YouTube. I mean I have a huge advantage here. I have this YouTube channel with 170,000 subscribers. So of course I can get some leads coming from YouTube. And I've heard from other AI agency that YouTube is one of the best source for leads. Um, so I created a couple of videos about Nitain. I did some tutorials and in this video I put what we call a CTS or call to action where I invite viewers to um to send me a message if they need an automation from me. And I got five leads from YouTube um and one meeting. And that was out of all the leads I got, it was the best one. So basically this guy was working for this uh consulting company that reviews construction plans uh from architects because you know you have to those construction plans they have to meet certain standard for construction building codes. And so what they what he wanted to do was to create an automation that helped them to speed up the process of reviewing these uh construction plans and it was really interesting but uh I was a bit scared because for first project it's pretty uh pretty scary to do that. Um so nowadays the models like Chad GPT they are multimodel multi-model which means that um they can accept different type of information in input so text but also images so potentially you can fit it a construction plan give it some criteria and maybe it can find what is wrong with this construction plan but I don't know it seems pretty complex for a first project um I made a proposal and the guy never came back to me. Um and yeah and that was it. So it shows that YouTube in a personal branding is really good for generating leads compared to Upwork. Um but after that I basically decided that I don't want to continue. So why is that? Okay. So first of all what about the market for AI automation? There is definitely some demand. Okay. Um, I saw it on Upwork. There are a lot of jobs. If companies that are looking for this service, no problem. But it's competitive. The barrier to entry is very low. You know, like if you want to use anything, make. com, you know, this is a graphical interface. So even non-coders can do it. Uh, you don't there isn't much to it really. only small difficulty would be to create AI agents and maybe know how to insert some context. Uh but even then like without coding at all you can already get some result. So what I heard from other agency owners is that and there are a few bad things I heard I heard. So first of all about the recurring fee that I mentioned before you know it's one of the main appeal of this business is that beside the initial set of fee you supposedly you can make some recurring uh you have some recurring income. Okay but the people I spoke to told me that in reality the churn is skyhigh. Okay. So maybe customer will stay with you for 1 month, 2 months and then they will cancel the automation and then try to re redo it uh themselves because they don't want to pay you anymore. Um so it's really hard to actually get this recurring income. And the few successful uh AI agency owners that I met um and not YouTube guru but real people that I met in real life so that I can really trust that what they say is true that the most successful thing they did was to sell education to businesses. So instead of saying okay I'm going to u implement this workflow for you um they taught uh those teams how to create workflows themselves and also how to use the AI tools and it makes a lot of sense because for a business owner it's very convincing if you tell him or her hey you know what I'm going to uh improve the efficiency of your team by 20% I just need um to run this four workshop

is $1,000 $2,000 And you know for business owner it's pretty good value proposition uh and also it can be a good first step to upsell some uh some more expensive services after. So, so this is what I realized is that really right now education is the thing that sells but not necessarily um I haven't met a lot of people who actually sell uh automation but they do exist because on Upwork you can see this is real but this is competitive. Another thing that bothered me with AI automation was that I really wonder how future proof it is especially when AI gets better. What is the effect for this AI agency? Uh does it help them or does it not help them? Because when AI gets better, well, it's easier for customer to do it themselves. So why would they hire some uh some agency? So I'm really not sure that this is really long-term. I think potentially with an AI agency, you have maybe a couple of years where you can make some money, but even less, but I'm not sure um if it's really long-term or not. And then I also wonder about SAS, okay? Because what happened with SAS business is that a lot of SAS, they're going to automate the most popular automations. So, think of it from the point of view of a customer. Do you want to bother to set up like a custom automation or to use a SAS service that already hide all the complexity um behind that you don't really see this? I mean most customer they just want something clean and simple. So over time I think that SAS business are going to eat the lunch of AI agency. Okay. So another thing I want to mention is that I don't think I worked hard enough to really test this business model. Uh both in terms of time and in terms of quantity of work. Um I needed to do maybe 10 times as much as what I did to really test the idea. So I want to be fair and really say this. But um the thing is I'm really didn't enjoy uh chasing clients you know like uh I could see after chasing just a few customer basically it really reminded me that this is a servicebased business and I just I hate that kind of business really like I you know before um before running my education business I used to be a freelancer and I just hated it. I just hated every second of it because you always have to uh hustle to find your next customer and I wasn't happy at all. Like that was really freelancing hell. I wanted to get out of it and I got out of it thanks to YouTube really. Um but uh so now I don't know why I thought that maybe I changed and maybe uh I would magically be interested in that kind of business again and no I haven't changed at all and I still hate it. Um so there is still an advantage to that kind of business which is it's it would be easier to get started an AI agency uh compared to let's say a SAS but it's harder to scale. Um this being said I was a bit disappointed because I thought it would be easier than this to get started. I realize it's actually also competitive. So my point is that well if you are to struggle anyway to uh get your business off the ground, why not going for something that um that you really value like a really worthy objective, something you will really enjoy, something that has even more upside like a SAS. Um and I also realized that after I started to talk about AI agency, I realized that my audience or you guys uh you guys really didn't follow. Okay. Um, so there was some initial excitement after I announced my pivot to AI, but then the videos after that when I started to talk about uh video AI automation, they didn't do so great. Okay. So, I run some polls in my community and I realized that you guys were more interested in SAS but not at all in AI automation. And um, you know what? Me too. Okay. Um, it's just that I was just a bit of a too much of a coward because I thought that AI automation is easier. But, uh, no and in any case, we just have one life. Okay? So, let's just go for the things that we really want. Okay? So, let's just go for uh AI SAS. And so, what so from now on this channel, I want to talk mainly about how to create your SAS business with AI. uh we will still have some videos where maybe I mention automation and enterain but only if this is relevant for building your SAS. Um for example, one thing you can do is you

can study in there is a marketplace on the website of Nitain where you can see all the popular workflows and you can see uh you can get some ideas for your SAS by seeing which workflows are the most popular. Um, but I don't want to talk about the business of running an AI automation agency because basically I decided not to uh not to go in this direction anymore. So from now on we go all in for SAS business. Now a quick note, I'm not alone in this AI automation agency that I launched. So there are other people in the team. So I'm not going to completely shut down the operation, but this is not going to be my main thing. So, I'm gonna uh create another channel just um to push this agency and I'm going to do videos 100% to get end clients. So, I'm not going to try to sell the dream of running an AI automation agency. Now, the goal is really just to find those clients. So, uh, if you are interested in this, I'm going to put the link down below for this, uh, this new channel, but in the future, I won't talk of this channel anymore. So, I'm kind of curious. What do you think of this new pivot? Are you interested in building your SAS? And if yes, what is stopping you? Tell me in the comments down below. All right, thanks for following this video. I will see you next time. Bye.