Transcript
Speaker: One year ago, I spent $500 in order to do a comprehensive AI SEO test. I wanted to figure out if you could automate the whole SEO process, get tons of traffic, and then monetize it.
Speaker: And today, i finally want to reveal the results of it and share the good, the bad, and the ugly with you. Welcome back to the AI Driven Marketer. I'm Dan Sanchez. My friends call me Danchez.
Speaker: And today it's an exciting day because I literally did this last February. Last February where I spent all the money, set up all the processes and had it go crazy with this AI tool that I invested in and had been running articles, hundreds and hundreds of articles ever since then.
Speaker: And what I found out is that it's Kinda got mixed results. In fact, it's probably more bad than good. But let me show you what I did in order for you not to fall for the same trap I did and now what I'm doing instead. So I bought a tool because it was cheap on AppSumo. You know, you can get these lifetime deals on AppSumo and I fell for this one.
Speaker: It's called the SEO Writer with Auto Posting or WP Auto Blog. And it's still a pretty prominent tool. And I think it's still an AppSumo deal if you want to test it out yourself. But here's what I found. This tool works directly with WordPress, uses a yeah an AI connection tool in order to write hundreds and hundreds of articles. And I upped it to like, I bought i double id triple stacked the code so I could write up to 500 blogs.
Speaker: a month so let me log in and I'll kind of show you how it works you can see here's all the different articles I was writing in addition to buying this tool I bought a whole different web host so I could have five WordPress sites and that were fast and on WPX dot net so is is not the cheapest web host but it wasn't the most expensive either 20 bucks a month and I just bought a whole year's worth just so I could test this thing out So I was like half of the 500 right there. The other half went to triple stacking the codes for WP auto blog.
Speaker: And, and then of course I had to register five domains because I wasn't going to test this out on my main site. Are you crazy? Didn't want Google to like flag my site as spam. I had to test this out on fresh domains and see if it worked there.
Speaker: And here's the tool. You come to come in here and you can do some keyword research with it. But I was doing the keyword research more manually because I wanted it to hit the right keywords, right? And then you tie in your keywords here or what I did was put in just the titles instead of letting it just make the titles up for me. I was pretty good at title writing. So I masqueraded all the titles and kind of just use the same template that I knew would work over and over again.
Speaker: for the long tail keywords. You come in here, you put in 100, 200, 300 titles, right? And usually trying to do this as fast as possible in Google Sheets with the titles. And then you get to have some settings, you know, it's like is it English, do you want it to publish immediately? No, you usually schedule it out every six to three hours. So that doesn't look like an AI bot just slammed and threw down, you know, four or 500 articles all at once. You want to spread it out over the month is spread it out every three to six hours, depending how fast you're posting.
Speaker: You get a group going and you can post it onto a website. You can see here's all the websites I did. I did multiple categories. I did one on AI marketing, one on video studios, one on audience growth, one on prayer, and one on Bible verses. Because these are two – prayer and prayer Bible verses are things that I've written extensively for. I've done a lot of content marketing for so I knew the market.
Speaker: So I was trying a bunch of different things that were analogous to things that I'd done before or things that I was known for that I could kind of see and vet more easily. And all these websites are gone now. And you'll find out why in a second.
Speaker: You could pick the style. You could pick the point of view that you want to write in the article length, which I always went for the maximum article length, of course. And then you can pick out some of these different sections in there.
Speaker: You could give it a but personal prompt, which was really key. this was This wasn't in here at first, and then it became in here about halfway through the year. I started utilizing it because before you couldn't really prompt it. You couldn't give it special instructions here.
Speaker: You could at least load in some special instructions. You could have it give it some instructions on what types of images to use, buttons, setting up other links inside of it. So you could do some cross-linking between these articles.
Speaker: Overall, not a bad setup, right? But what I found over and over again is that it was good at writing out, knocking out all these articles, but the articles were just horrible. They were so bad.
Speaker: She would even pull YouTube videos into it that were tried try to be related, but would oftentimes fall far, far away from the actual topic you were writing for. If you went into these articles, they they were just bad. In fact, let's take a look. And I haven't even previewed this, but let's go ahead and take a look at one of these articles. so So you can see this one, I was trying to go for a lot of like long tail AI search. So ultimate guide to AI for shoe repair. I was just like going down the list of all the different local businesses, right? Things that nobody else was going to be writing for. I'm like, I'm going to hit them all. So I went like, you know, laundromats, like salons, like all the different, I just came up, I had ChatGPD come up with a list of like a hundred different local businesses. And then I wrote the ultimate guide to AI for insert local business. And that was kind of the approach.
Speaker: And it writes an okay article. But it goes in and it's just not really helpful. It's like machine learning algorithm algorithms can analyze customer data and predict trends in shoe repairs. Wait, what?
Speaker: By examining historical data on repairs, these algorithms can help you identify which types of shoes are the most frequently bought for repairs, although you stock although you allowing you to stock up on necessary materials and parts. You're like, okay, that's not really that good.
Speaker: Broken images often. and I just found over and over again that the results that it was generating were sometimes just at best, at best it was man At worst, it was actually wrong, like like way wrong. And of course it is.
Speaker: like So say what what do we got here? Like here's an embed, this and this was very common for the embedded YouTube videos. it would puff It would find of course it's about shoe repair. So it goes and finds a video about shoe repair.
Speaker: Does this have anything to do with AI for shoe repair? No, if anything, you wanted a video about AI in here, not about shoe repair. So that would throw it off the images and it would also be off.
Speaker: In this case, it picked images that I think probably would have worked. It probably would have had pictures of shoes like this one. But ultimately, it's just not that helpful of an article.
Speaker: The funny thing is,
Speaker: It still worked. People, these blog posts were written. They got posted to the website. Even even though they were kind of garbage posts, they still ranked because it was the only freaking thing that, it was the only post being written about AI for shoe repair, right? So naturally, Google's seeing all these articles come out and they're like, well, it' it's something.
Speaker: it's it's It's not nothing. And they would rank. The issue is none of them, all I did these 100 times and I ranked for hundreds of different articles. They would rank, but they would get no traffic across hundreds of posts.
Speaker: Anything with any volume at all wouldn't rank. So it became just something that just didn't work over and over again because I had heard of people writing articles like this just to test and see what got traffic and did rank and then start pulling in traffic and they would come back in here and then rewrite it in order to make it actually good.
Speaker: Because at best it was C content. But at worst it was often DNF content. pretty Pretty often. This one would be a C. And it's still meh at the same time. But it they would rank. Some of them would rank. This would be an example of one that would rank.
Speaker: And it just would rank and get... almost no traffic, even ranking number one or two or three for some of these search results, because there's just no one searching for it. Even with lots of them, these websites would get almost no traffic coming to them.
Speaker: So at the end of the day, i ended up pulling the plug. You might be looking at this and dan saying, Dan, great. Thank you for figuring this out. Thank you for spending the $500. So I don't have to, and you're welcome.
Speaker: But I wanted to give one piece of advice before we go. While this was a failure of a test, And do not try to automate all your SEO content, even if it looks like it could be a something you put on autopilot and then reap the rewards a year later. It's not going to work that way.
Speaker: And if you do try it, I think the few people who are doing this are doing this in a much more sophisticated way. I'm still putting a ton of time into it. I try to put the least amount of time into it, set it, forget it, and just kind of review it so that I can pick apart the good ones and double down on it.
Speaker: But it ended up not working as well. I still think AI is actually a good tool for writing articles, but the process still just takes longer. You can't expect a tool like this in order to come up with good content. In order to come up with good content with AI, you still have to work on all the research you have to do up front and truly understanding what someone's looking for what with the search intent, what do they actually need, what it takes to outrank the top three articles that are currently ranking.
Speaker: and then actually crafting the content collaboratively with AI, injecting your expertise so that it actually has something good to go off of. Now, I think there might be some opportunities with Google's deep research or any of the deep research tools that are coming out now in order to speed up that process and help do get really good sources and stuff for your articles. But I still wouldn't copy and paste those directly into it without collaborating with AI in order to do it.
Speaker: Because ai can essentially just speed up your writing. Now, if you're a writer, you might be able to write faster than AI and get your ideas on onto onto the screen quick. But me, i'm I'm not that fast of a writer. It takes me a long time to knock out a 3,000 to 5,000 word article.
Speaker: So now what I do is just organize, outline it with AI, dictate to fill in that outline, have AI ask me some questions in order for me to answer to give it more context so it can then fill it in and write it for me.
Speaker: But is that AI writing for me? Yes, but it's not thinking for me. It's my ideas, my thoughts, and my examples that I'm giving to it. Sure, I might use AI too to go and find some more examples that I might be able to illustrate with it.
Speaker: So that's why I'm saying it's a collaborative process for creating good articles and good content that are actually worth ranking, that you've actually looked through and read and thought, huh, yes, this is this article is 10 times better than the current articles that are out there.
Speaker: And therefore, it's worth publishing. It's worth people's time because it's a good answer. And I think this is the way SEO will move going forward over the next year or two. Until AI is smart enough to actually understand the whole picture of what someone's looking for, actually understand what images are actually going to be helpful in the article, what videos are actually going to be helpful in the article.
Speaker: Like we're not going to see AI automating the SEO game. at least for another year or two. Right now, you need to collaborate with AI in order to rank. And of course, SEO is switching very quickly to GEO, Generative Engine Optimization.
Speaker: So that's going to be a thing. And this game will continue to change and evolve over the next couple of years. So I'll be back with more SEO and GEO information. But I wanted to give an update on this larger project that I started a year ago.
