Clint set this test up to check to see if he can confirm two things at the same time. One, he wanted to see if the Google Indexing API was still working as his previous tests outlined. Second, he wanted to see if the latest Google update had changed the pattern of pages indexed and ranked for the same keyword from the same domain. In the past, multiple pages were filtered out and deindexed. However, we lacked any insight into how it was selecting the pages that got to stay.
Test Downloads
Transcription below:
Clint: Google filters indexing by density. So if you all remember, we had in the last 2020, it seemed Google's indexing system was less than ideal. During the course of that, there's a lot of things that happened, One, got harder to get regular pages, much less test pages indexed. And then two, Google released a noticed that saying multiple pages from a single domain will start to be filtered out.
One, in the search results, we didn't see that a whole lot because we saw 1 or 2 and we assumed as SEO that filtered out means you'd only have one listing in there. But it turns out that was wrong. And two, the testing pages we just figured. Hey, well maybe Google's kind of caught on to lorem ipsum tests. Let's try random alpha, nope, that didn't fix it. Maybe random alphanumerics. That didn't fix it. And then finally, we came to the conclusion that we just need to use multiple domains.
The updates came out, was it December 2020? Updates came outs, and I wanted to see 2 things which I hadn't tested. Because I mean, for some reason, I didn't even think about it, it was one, does my Google indexing methods still work? And then two, is the filter still there now. And so I set this test up on the 20th of January, and we ran it all the way through the 15th, or the 16th of March. My indexing rate was at 60%, which is typical with my testing domain just based off of the random domain ranking factor, and I kind of expect those numbers. But you can see I got a 82 92 85 78 and 96. All using Google Indexing API.
There's no indexing service or anything like that. It's just the Google Indexing API ranking our indexing methods. And so 60%, I was actually pretty happy with it. It was exciting, especially considering that was only 24 hours and typically I had to wait 72 hours to get these greens on all of the sites anyway. So 24 hours, I was happy.
Dori: It could be one of the sites that were red.
Clint: Yeah, well, you know I think it's a random domain ranking factor that Terry Collins talked about ages ago. And sometimes that affects those multi test sites. But you'd like to set in 72 hours, that all kicks in, and they're all typically red across the board or green across the board. I mean this test was also just free information, it was a redo of the keyword density test. It's kind of a basic template that's inside of the software.
However, I modified it, so tests that control are at 10%, and everything else would go up until 5, so 15 20 25 30, all the way to 95%. And based on the indexing and the sorting, 50% indexing one, based on the average rank. And then 95 65 65 60 and 90. So part one, what does this do? Well, this tells us the keyword stuffing can most possibly still work and still work quite effectively, depending on how you leverage it and what your keyword is that the average percentage at the top 10.
So, let's say your average is percentages of 5%, you probably don't want to go in there with 95% keyword density Google's gonna think about it smart enough to figure that out as being crazy at normal, and it's not going to give it any benefit. But if you went in there with the 7%, and everyone else had a 5 and you had to do a little bit of keyword stuffing to do it, you might win when all things are considered equal. So something to consider there. If you're just having a little bit of trouble, bounce it up, play with it. Alright, so go ahead.
Dori: That's really interesting. That's the 50% and the 95% was like, wow, another wow one.
Clint: Alright, so that again, that was after 24 hours and on February 24, I did another check. Remember I set this up Jerry. February 16. So 24th is roughly a week later. I didn't check to see the results. And as you can see, most of my test pages are gone. So I went from 60 to 13%. Instantly, we knew that the filter was back or it was still there it was still applying.
So, that's lesson number 2 out of this test is that you can indeed cannibalize your content. And what was even more interesting was actually they sorted that by keyword density. So if you look at the average rank here, if you sorted by average rank 20% percent and a piece content still a lot.
However, when you look at it a little bit closer and how many pages actually remained in the index, and what Google was selecting and leaving in the index, you can see that the numbers are 85 95 90 80 60 75 65 50. The controls are set at 10. So there's a woodpile, there's a randomness, and then we get to 25, and 55. And then the rest of those all the 10s, 35, everything that would be considered probably normal in content are all gone.
It was all sorted by keyword density. So 5 95, etc. 85% was actually the winner of all the test pages, eight of them had a 5%. And that's the one selected. Okay, so filters are there, and the sorting is there. So if it doesn't, Google didn't have anything else to grade our web page by other than keywords. So minus all of the factors that really didn't have anything left to go to put keyword density. So keyword density is obviously a factor.
Now, the keywords stuffing part is just makes it weird. Like, why would Google do that? Like, why would intentionally or be taught that the highest wins? If it doesn't matter, and I think it has to do with that the averages of the search results. Again, if your average is 5%, you don't want to be at 50%, you want to be more like at 6.5. And you're going to win that keyword density measure. But in this case, Google has nothing else to go by so it just said alright, whoever the highest wins which would be great if you can find this kind of search results.
Dori: Will the programmers at Google deleted that code that searched and destroyed keyword stuffing sites.
Clint: Yep, for sure. So we're back again, the 13.2, a little bit of a drop, but again, I sorted by samples, but you see, 90% was winning in February. And then in March, where 11.4 85 was winning again. 90% drop, but 80 85 65 and 60 are all winners. And then you see everything that stuck. So pretty much everything with greater than 50% keyword indexing, the filter said, oh, we're gonna keep you with exception to these two control pages right here that we're at 10 is just kind of in the middle.
So, let's recap a little bit. First, the indexing methods still work. So if you're using my methods that I thought you guys, or you're using Google or mega indexer, you're okay, the Google Indexing API still works. And so I recommend that you continue doing that, especially if you want to see effective results, typically 24 hours, you can expect it but it's not unheard of to wait 72 hours for those to kick in. So just be patient with it.
Second, the page filter still there and Google still working. And now we just kind of know a little bit more about what it's doing. Finding the most optimized page for the keyword. In our case, in our test case, we did keyword density, so found the largest and then I think as it got used to seeing the largest between 50 and 95 80 85 kind of turned out to be the average content should be. And so we're just gonna keep in a 5. I think that's why that worked out. It was the average is the law, the average is what we expected to see. This is way too much. That's not enough. 80 85 as we want.
Third, we had a whole bunch of test pages set up this week from the SIA team, he sets up pages, gets them index for us. And then the testers can go in there and use them as a desire. We looked at it, and most of those who that maintain top 10, up 10 pages index stayed for a while, and then they all disappeared. Oh, and he removed a lot of these. And I wish I hadn't thought of this before, but the domains uses some of them .com, some are .net, some are .orgs. I think there's .info in there as well.
And as I was clicking through there, I noticed that some of those other tlds were not getting filtered, whereas the .coms all were. So I want to set up another test to see if that's actually the case, because that would be pretty interesting to know if it only applies to the .com.
So, then if you have a .net, and keyword cannibalization is much less of an issue for you. Alright, so on the topic of keyword cannibalization, this also means that since they're grading on the number of factors and the level of optimization of your pages for the filter, if you are trying to rank a page, and let's say you made it 3 months ago, then you went through Dori's on-page course and you're learning a whole bunch of stuff, and then you made a new page, then you optimize it thinking you are creating a silo page for your money page. But your SEO skills were much better than they were in 3 months ago.
And let's say you did that five times. And those five new brand new SEO pages are really, really good SEO, they're all pointing at your money page. And then you notice your money page deindexed. And I believe based on the results of this test is you just created a keyword cannibalization and not only did you do that, but you got four pages ranking for the same keyword because they're all really really tightly niched, and you should do niched silos.
And Google just went through your silo, found the three that were the most valuable for that particular keyword and just deleted from the index, the ones that were not. And so that's why people could be seeing their money pages not ranking. But their silos. My silos are doing great, but my money pages are not is because your silos may actually be entirely too niched, to your main topic, and your might be cannibalizing your own stuff. Let's see. Talk about that.
Dori: Webinar if it's available where you teach your indexing methods
Clint: It's in the members area. Alright, let's see. Oh, and then finally, keyword density is a ranking factor. When all things are equal keyword density wins. And as you see here, 85% won the day. So one and two, every once in a while they tested the 30% page in there, but went into 80-85% was the winner.
So again, I think they found the average of the pages that were ranking. And then when everything was indexed, right? So we'll go all the way back up here to make sure everything was indexed. And these ones all ranked higher. This created the average for what the keyword was doing. And then Google filtered some stuff out. And then Google did it again. And then finally again, and it settled on the the numbers here in the May 16th in 85-80. These are all the winners.
That's a really really good information here. I think you guys can apply this. So One, I am not telling you to go out and make pages with 85% indexing, it's not gonna work. And two, I'm not telling you to use the old keyword stuffing method of adding a whole bunch of instances to your content and then making it display, none. That is probably not going to work either. What I am telling you is a keyword density is a factor. And all things related being equal. If you are a little bit higher than the average, within reasonable expectations, you can beat the target pages with keyword density.
So, do not discount that as an idea when you're writing your content. We did the content optimization video on SEO this week, to show you kind of how to get through that, just write it normally and then do the optimizations and stuff. And then ultimately, your keyword density is going to be just about right.
So, do that and you're gonna be okay. But now, you know the keyword density is a thing. I will also tell you that most of the stuff that we're testing, when it comes into regard to place word placements and stuff, we're actually testing keyword density.
So, there's another test, the HTML tags test 10 revisited that and actually tested it, not to see if it was a ranking factor, but to see if Google would read that those tags and he tested 154 of them. And each one of those tags either index, the keyword that was in the tag or it didn't. And I will say index that it ranks the keyword that was in that tag, or did the page was indexed, but it was missing the rankings, right on some and versus others.
So there, you could be manipulating keyword density when you're actually talking about what's in your HTML at the same time. So you don't necessarily always have to be thinking about keyword placement and keyword density. When you're writing your content, you can completely ignore it and think about boosting it up a little bit. When you're applying ID tags to your h1s or the alt tag and your images and all that stuff. This all could be potentially driven to increase the keyword density of the page.
Google is obviously reading them, we know that now we know which ones are reading. So we just have to test and see Is there a fine line that you just don't want to go across? Ted suggested maybe that is probably 12%. That makes sense to me. And my recommendation is still the same. So don't go nuts just go a little bit higher than everyone else. And that's when you're gonna get the benefit out of it. And you don't even have to stress too much over it when you're just doing a little bit higher, versus trying to gain this number.
Dori: Did Tod say where he got tha 12%?
Clint: No, he didn't. And I didn't ask anybody. I had doctor's appointment this morning. And I had to run out. I should have asked, that's a good question. I didn't get to be the problem with testing that right is, we know that all of these work. So let's just assume that we did 1 to 15%. And then maybe 12% wins, does that really the magic number? And if we bounced 12% against 95%, which one is going to win? Based off all the tests we've seen, 95% would win? Right?
So, then that leaves you with what we gotta do it in the live search results where you're competing against all these other factors too. So how do you know that? Maybe in my test page, I just had a better title or Google like. The keyword was in the right place in the title versus what? You know, the other one, and maybe that's why the 12% won, because you got to do it in against real competition. So I think to me, that's a little bit hard number. Ted doesn't save too many things without a test though.
So, as definitively as he did. So I'd expect that there's probably a test somewhere sitting there somewhere that he's got that shows 12%. Thank you all for showing up.
Dori: Another question. Does the density test? How does this relate to surfer pop and cora results?
Clint: So that's actually right at the last minute there.
Clint: That's a great question guys because there have been a lot of talk inside of the slack community about over optimization and those tools. So the the number one argument is, I use pop and I use surfer or I used whatever. And I over optimize my page. Well, clearly, unless you went full crazy with your keyword density, which Google's looking at. You didn't over optimize your page. There might be something else in there.
You might be the application of your words and stuff in formulating a topic, but by you adding words in regards to the recommendations from any of those tools. Unless you add in 1000 of them you didn't over optimize your page. So It's a good case in this situation where you need to look at, did you follow a recommendation in there, not in regards to adding a keyword. But in regards to the format of your page that may or may not have messed things up, ie, having to create and these are famous is like, Oh, you have to add 51 new H tags. Clearly you don't, right?
That doesn't make any sense in any planet. But the tools are all set lower, you have to add 30 entities to this, or you have to add 5 exact match keywords to this. And typically, what I think we are doing by doing that is increasing our word count, and reducing our keyword density. And that's why you're not seeing results.
So, how do you avoid that? Number one, I gave all of you an SEO this week of content optimization video, I am going to redo it again for SIA only. And we're going to do it from start to finish. Like think of a piece of content, we're going to write it together. And this is going to be a long, crazy video, we're going to optimize it in phrase surfer in links. Those are the three tools that I use right now phrase surfer in links, notice I use three tools, I don't depend on one. And then we're going to do hit it with grammerly.
So, I write it in the short version of this whole video would be write your article in Word, put it in phrase, optimize it to 100% put in surfer, optimize it to 75 to 85%, put in links 85 to 90% and then go back to word and put it in Grammarly. Every time I do that, I usually end up like 3 or 400 words less than my original document after I optimize everything. Which actually increasing my keyword density. So I think that's where all these problems lie with these tools like they all should end up giving you decent results. Because most of them are all based off of the same thing. Either using the Google NLP API directly.
They're using a clear scope is using Google NLP API and IBM Watson or they're using text razor which uses the Google NLP API. So you should be getting similar results for each one of the tools which makes me laugh when people compare them because you should be getting similar results to them.
And if you look at when we go through the video with phrase and surfer and stuff, some of those are completely off and and some of those are completely different. And it's because of the way competition is selected phrase by default, top 20. Surfer, you can kind of pick your competition, pop you definitely pick your competition. Cora definitely pick your competition.
So, when you change the competition, you change the recommendations and that's how they become different. So you're messing yourself up from one tool to another because of the ability to or not the ability to identify your competition a little bit so yeah. I think honestly everyone who's having issues with any of the tools having a keyword density issue. And I think that the test proved it and I owe you all a content optimization video, the walk through how to use these multiple tools.
If you have a decent budget, then surfers what? 89 bucks that's what my level is. All the other people that SIA use it. Inlinks is 39 I think thats what I'm paying, phrase I got the lifetime but I was paying the monthly when it was a beta is like 40 bucks. And then Grammarly is like 100 bucks a year. I didn't mention cora at all. And I love cora. I don't need that for content optimization.
If you have any doubts that this stuff works with just content optimization. Keyword analysis, this page is just content optimization. And there's no schema. There's no magic stuff. There it is number two, so number five, number two. For a while, there was actually ranking for keyword analysis and keyword research. But Google got smart is like not that good. But it's just using those tools and putting that stuff together.
So think about it this way. You get it up this high, or you get it up to number five, and you do your schema, maybe that knocks you in the three or four, and then you build four or five backlinks, you've just spent significantly less money than what this company probably did buy all the backlinks or do whatever it is they're doing, just by optimizing for keyword using the content writing process. So I owe you guys that. Hopefully, you will take it and apply it and use it.
Dori: Yeah, that's what it's what we're doing for any site, say as well. So get a sample example on display as we continue to publish articles. Yep, Garth, to understand surfer. That is not. Do I understand that surfers under reporting and pop maybe reporting? Yeah, it could be the other way around to depends on how you selected your competition. There's those are really user input based. And but I didn't say that at all.
I don't really get that from phrase surfer and get where do you get your competitors from Google search results? competitors are all there, you got to look smart. You know, if you're local, your competitors are not homeadvisor. And Yep, oh, they're never there. They're above me. Well, they're above you because of directories. And Google's favoring them and pushing down local local sites so that they've forced local business to pay for ads, because local businesses are the primary revenue driver.
So, your competition is the other local businesses in your area. So you have to pick those out and select those depending on the tool you want. And think, pop now actually used to recommend like force recommend your competition, and then now they allow you to pick or vice versa. I think there's two options. When I left pop, there was two options, I could pick or I can just let it do its thing.
So, it depends on how you're using that based on the results you get. And it's the same with surfer but I believe surfer just recently implemented Lang either to pick your competition that used to be forced. Everyone is your competition, and we base your number off everyone that we saw. So they're all certainly making improvements.
For me, it's a UI thing. I just like using surfer better than I like using Pop. Pop up this new UI. It makes it more user friendly. I might go back. We'll see. I'm kind of a genius. That way when it comes to tools, I really don't care as long as it's easy for me to use and it works for me.