It’s just over a week since I launched another one of our mass AI sites, right in the midst of the Google update which is specifically going after scaled content; and so far, things are going really well.
The Pareto Principle in Action: Finding Your Content Winners
Many people are always debating about the easiest solution for creating AI content. They ask: “What’s the easiest solution where I can just load up my keywords, press go, and have a website?”
Well, that’s exactly what I’ve done here but if this were a site I really cared about, that wouldn’t be the end of it. It’d purely be a testing ground. It’s the classic Pareto principle: 20% of your content provides 80% of your traffic and revenue. I’m always telling this to clients: ‘You don’t know when you first start out, which of the content you create would be winners.
So if you’ve got a keyword list of hundreds of pages of content you need to write, most websites can normally get away with just 50 to 100 but if you are getting from more of a bigger site, then the chance is you’re going to have to write 300 articles in order to identify maybe 50 or so that are going to give you serious traffic. Every site I audit, it’s always the same: a couple of pages dominate and the rest are not doing anything.
That is exactly the joy of AI. In this case, I used autoblogging.ai which has a sophisticated prompt. The actual content we get is pretty good. not just plain ChatGPT. It works with the OpenAI API and also provides ‘key takeaways’, which is very good for making the content machine-readable.
Google does seem to prefer that and then you can see we’ve got this long comprehensive article with FAQs. This one is doing very well because it’s structured in the right way for Google. It uses entities, FAQs, and has key takeaways. So we’re already a decent step beyond the majority of websites just spewing out plain ChatGPT or plain Open AI API. We also have images.
Low Competition Keywords: The Secret Weapon for Ranking AI Sites
The point is we’ve now identified that this keyword is actually low competition. In the ideal world, we’d be able to look at our keyword list and filter it by keyword difficulty. Unfortunately, there’s always nuance there.
I don’t use the keyword difficulty, I always recommend actually looking at the competitors and what their domain rating is. In our experiment, we had two DR zero websites, and we also had DR 25. Generally, that’s a good sign that this is low competition. Note that you have to do that for each individual keyword, but even with that, there’s no guarantee. Even if you find a low-competition keyword like that, it’s not guaranteed that you’re going to get it and rank for it.
So the only real and reliable way to find out is just to do it and that’s exactly what the AI can achieve. So in this case, we did something like 400 pages in just a matter of days. Like I often say, I’m not doing true mass content where I’m doing thousands of articles in a week because that seems to be a clear indicator of the sites that are getting hammered at the moment. Instead, all my posts are scheduled.
So for this exercise, I basically went into the dashboard, and into the ‘posts’ section where we’ve got 323 in total; of which 200 are published, and 100 are scheduled. I went to the ‘schedule’ section and skipped to the end of the scheduled archive because when we did Autoblogging.ai, we went in order of volume. This means I’ll be publishing the higher-volume keywords first.
I’m also going to go from the oldest to the earliest posts that are not currently published but are scheduled. From 100 scheduled, we’ve now got 80 scheduled, which means by following the steps we followed, you publish 20 posts at a time. I think maybe 40 or so per day in total should be fine. I’m being quite gentle and conservative with this one, so I’m not getting them indexed or anything. I’m just letting Google naturally pick them up over time; and then in a few weeks, we probably will go through and prune the content.
Content Pruning: A Crucial Step for SEO Success
This means that we’ll take out any pages that just aren’t getting indexed, or aren’t getting attention. That’s basically the content pruning theory.
Now there’s this theory about crawl budget that the top SEOs have been debating. Certain top folks have been messaging me about it, saying: ‘Are you doing this?’ The theory is that with content pruning, Google wants to limit its crawl budget with sites; therefore you want a smaller site with lots of good content that’s getting lots of traffic, rather than a massive website with lots of pages that aren’t really doing anything.
There is quite a strong theory that if you are doing this kind of mass AI content, it is worth going through your archives using a search console so as to work out which pages are not getting any Impressions over 3 months. Take all those out because that way, Google will see, as a percentage, that the majority of the content is good and getting traffic. It can therefore crawl it more efficiently. Generally, you also win more favor with Google.
From AI-Generated Content to Revenue Generation
Like I said, we’ve now published 240 posts and from that, we’ve been able to identify keywords that we’re ranking for. If this were a higher ticket niche, or if this were actually a site I cared about, I now probably would actually go in and work on improving the page mentioned earlier in this blog post (Vegas Matt Net Worth).
So I’d be thinking ‘How can I bring in more unique insights, more value, more images, and things like that? However, you don’t want to do that from the start. Most people really get hung up on this, thinking every page needs to be perfect when we publish it.
I don’t do that because I believe you’re going to publish 100 or 200 pages and you want every one of them to be perfect, you may be wasting time because perfection does get in the way of action. I’m a big believer in taking radical, imperfect action and then just improving as you go.
In this case, I’ve done plenty of mass AI websites that don’t even get this kind of traffic this quickly, so already we’ve validated that the site is looking good and we’re going to get traffic. Now I can invest more resources into it. Having done this, I might improve some of these pages that are starting to rank.
How Link Building Helps AI Sites
I’ll probably elevate it to the next level of investment which is link building. So, first of all, foundation links, just citations and press releases; and if it’s still looking good, maybe we’ll actually start building some backlinks. I probably won’t do that until we’ve validated some idea of revenue. The fact we’re getting this much traffic this early gives hope that we may be able to keep scaling that and work out what happens when we put banner ads on it — whether it’s AdSense or Media Vine.
We should also get more clarity on how much revenue we make for every thousand clicks we get. Therefore, what are we willing to spend to get that number of clicks? And how do we do that?
Data-Driven Decisions: Optimizing Your Low-CPC Niche
We assess the backlink gap. In this case, we got a DR of 49 here, at number three, and an estimated 520 visitors per month. Everything else had a much higher DR, which meant it was a tough one. I know that there’s significant investment required to actually get the kind of DR that would enable us to rank in the top five for ‘Liz Cheney Networth.’
Now that’s not prohibitive because if we do get up to a DR of 49, not only can we potentially rank at number three for this keyword, but loads of other avenues open up for us. Just look at what celebrity networth.com has done: they’ve got a domain rating of 77 and therefore they’re appearing on pretty much every single one of these kinds of searches; not forgetting that as you go aggressive, you push yourself higher up and just generally get more visibility, and more traffic.
These are the kinds of keywords that attract backlinks naturally, but you’ve got to be in it to win it. So if we do some link building, push those pages up, and get more traffic, not only are we going to be making money in traffic, but we’ll also get more opportunities. So for a site like this (low CPC niches), I am dependent on the low competition keywords and getting those organic backlinks rather than more of a high ticket commercial website or an affiliate website, where generally, I’d be looking to rank in the top five for just a few key commercial keywords.
For that, I’d be much more aggressive; but in this case, it’s all about volume, and banner ads; so I’m really not looking to invest too heavily too soon.
