Jump to content
Sign in to follow this  
mki

Anchor Text Diversity, You're Probably not Diverse Enough

Recommended Posts

So, I was asked to analyze a few sites today that somebody is having some trouble ranking and it's the same problem I see over and over again:

Lack of anchor text diversity.

When looking at a domain that earns links naturally where I can safely assume that they are not doing any kind of aggressive link building I observe the following:

I almost never see any single keyword anchor texts above 1% of the anchors : links ratio. Usually they're all lower, or there's a few oddball anchors (Check netflix.com in majestic.)

Also, generally speaking. Regarding the main brand anchor text : linking domains ratio, this number is generally much higher than you think it might be. 20-40% is common, but the ratio of the main brand anchor to links, is almost always less than 25% (5% or less is common.)

This means that it is fairly common for a domain to have multiple links to another domain, using the main brand as an anchor and also use another anchor texts.

If you never noticed that, that likely means that your anchor texts are nowhere near the diversity level they need to be at, since this is going to force you to build even more diversification.

Also, using URLs as anchor texts, is generally pretty uncommon and I usually see generic anchor texts like "here" show up before the URL or variations of the URL.

Many people are building their branded anchor texts like this: 20% big brand - 20% bigbrand (no spaces) - 20% www.bigbrand.com - 20% bigbrand.com - 20% http://www.bigbrand.com

That is actually, highly unnatural.

For those specific types of branded anchors, a ratio of something like 65-20-5-5-5 is much closer in line to what naturally occurs. And the 65 is less than 4%-5% in many cases, compared to the total links.

Also, in most cases (again at domain level) as I go down the list, I see branded anchors, then many generic anchors, before I see a single keyword anchor text. I usually see 40-50 anchors before I find a keyword anchor text.

Now, when the first specific keyword anchor text appears, it's generally less than 1% of the domains total anchor text links (sometimes as low as .01%), but the ratio of keyword anchor texts pointed to that page, can actually be quite high. I've seen as high as 50% getting a page to #1. (Or 100% in the case of a single link, but again, compared to the links the domain has, it's a minuscule amount of links.)

I recommend: if you want to avoid penalties, don't use exact keywords as anchors and never repeat anchors unless it's your brand.

The first occurrence of a specific keyword as an anchor text for a domain only being .01% of their total link profile, is actually common.

If you study backlink profiles of big sites, trust me, you will be surprised at how chaotic and random real anchors actually are.

The stuff I saw today is WAY OFF, compared to those sites. Granted it's known it's not necessary to be as diverse as say amazon.com's link profile. But generally speaking, the more links you have, the more diverse the anchor texts should be. There isn't a specific percentage that looks natural for any type of anchor text. For a small site with less than 50 linking domains, 50% brand anchors and variations compared to the total amount of links might look natural, but it doesn't when it has 5,000.

Also, if it's a US only site, links from foreign domains are bad and shouldn't make up anything but a tiny percentage of your links.

Share this post


Link to post
Share on other sites
SEOPress WordPress SEO plugin

2 hours ago, yukon said:

This website has 62% URL anchors and ranks just fine.

http://openlinkprofiler.org

Not according to majestic.

And those keywords are like random spam.

Edit: I hope you don't use that tool, I tossed a few domains into it and it's not very accurate. One has 100+ RD and it found like 10.

1.jpg
2.jpg
3.jpg

Share this post


Link to post
Share on other sites
9 hours ago, mki said:

Not according to majestic.

And those keywords are like random spam.

Edit: I hope you don't use that tool, I tossed a few domains into it and it's not very accurate. One has 100+ RD and it found like 10.



 

 

You're obviously checking the URL anchor percentages wrong.

...and even Majestic says your OP theory is wrong about URL anchor percentages considering the site ranks good.

URL anchor text is completely normal and to be honest it's silly to think otherwise.

 

 

 

 

 

554.png

 

Share this post


Link to post
Share on other sites
14 hours ago, yukon said:

 

You're obviously checking the URL anchor percentages wrong.

...and even Majestic says your OP theory is wrong about URL anchor percentages considering the site ranks good.

URL anchor text is completely normal and to be honest it's silly to think otherwise.

 

 

 

 

 

554.png

 

The site is a christmas count down timer, with 11089 links and 474 unique anchor texts (Alive and Followed according to majestic) with nothing anywhere near that link profile in the rest of the SERP.

Even if Google ignored 90% of that site's backlinks, it will still have enough authority to be locked in the #1 position.

Edit: This statement is incorrect depending on what keyword. I did not actually look at the top keyword, when I wrote this. 50% sure. The days.to domain only has links 3 on pages >= 40CF, but none of those links actually point to the christmas page, where as xmasclock.com has a 49CF pointed straight at the page, with that link and ton of 20CF+ pages, it's a 46CF, which is actually impressive considering the amount of authority it takes to go from 40 to 41 and so on.

Which why would Google ignore any of them? There are about 15 url anchor text variations resulting in 1491 links or about 13.44% the total link profile. So, 86.56 % of the link profile is text or image anchors with 455+ different variations.

That also doesn't take into account sitewide links, which Cutts even said they "collapse those" and they likely only count that as one link. Granted, I made no attempt to factor that in to those numbers since I have better things to do today, then sit around and check the 11000 links of a christmas count down timer.

Where would you expect the only page of a one page website on a DA 48 domain to rank?

I hope you're not looking at the pie chart since that's an estimate and includes nofollows and deleted.

Even with those links, that chart is not consistent with the breakdown when you hit export and I have no idea how they come up with those numbers for the pie chart.

It says right in the footnote to check the full details in the anchor text tab.

Note: You get some indication that some of that is skewed by sitewide links in the SEMRush anchors report.

4.jpg

If you're going to ask me: "Why is the pie a lie?" I don't know man, it's a cruel world out there.

Share this post


Link to post
Share on other sites

I think what mki is trying to get at in this thread is that if a domain only has 7 different anchors pointing at it, that could be a problem. However a domain that had 700 different anchors regardless of what the percentages might look like, is probably going to have a much better day in the Google SERPs. 

It's not so much the percentages that matter unlike what so many people like to harp on. If one anchor made up 40-60% of the anchors in a backlink profile, but there was also another 800 different anchors mixed in, you are likely safe from any Penguins or other evil Google creatures. In contrast, if that one anchor was used in 40-60% of the links in the profile, and there were only 8 other anchors in the whole profile, that would be more suspicious. 

I've already started drinking tonight, so if any of that doesn't make sense, I'll try harder tomorrow. 

 

Share this post


Link to post
Share on other sites
24 minutes ago, JohnHemmer said:

I think what mki is trying to get at in this thread is that if a domain only has 7 different anchors pointing at it, that could be a problem. However a domain that had 700 different anchors regardless of what the percentages might look like, is probably going to have a much better day in the Google SERPs. 

It's not so much the percentages that matter unlike what so many people like to harp on. If one anchor made up 40-60% of the anchors in a backlink profile, but there was also another 800 different anchors mixed in, you are likely safe from any Penguins or other evil Google creatures. In contrast, if that one anchor was used in 40-60% of the links in the profile, and there were only 8 other anchors in the whole profile, that would be more suspicious. 

I've already started drinking tonight, so if any of that doesn't make sense, I'll try harder tomorrow. 

 

Basically. I've never personally believed or observed a "specific percentage" that "worked" since penguin.

I know that I can fire up GSA and annihilate my rankings fast. Which, unlike "losing weight fast", "losing rankings fast" is not so desirable.

As far as penguin, I'm extremely confident it's looking for a pattern of anchor text repetition + low page authority. That is what the spam links be.

Regarding making sense; either you drank too much or too little. That depends on you.

Share this post


Link to post
Share on other sites
On 6/8/2017 at 9:18 AM, yukon said:

I've already proven the OP theory is nonsense.

 

Your definition of 'ranking good' is really odd in my opinion. I analyzed each of your responses very carefully and explained both cases.

In the case of xmasclock.com, you're going off the data in the pie chart, which as I already explained, it wrong, it's an estimate. At one point in my life, I realized, that once a domain has so many links, the pie chart is wrong, so I stopped looking at it. I'm fairly confident that it doesn't factor in nofollow links or sitewide links, which Google handles both of those differently, according to Google reps.

In both examples, the sites are not competitive ...

Being #3 on the keyword " imgchili " isn't doing openlinkprofiler.org any good and nobody trying to make money is going to try to compete with xmasclock.com, it makes zero dollars. Actually, it loses money. If I go into SEMRush, pull up openlinkprofiler.org's organic keywords report, then filter by "SEO" or "link" it ranks on absolutely nothing valuable. This reminds me of another conversation I had elsewhere where I was talking about directly targeting extremely competitive keywords and pointing a bunch of links at it; will you whiff on the target keyword?, of course, but Google will "find keywords to rank your page on." Which after careful thought about this, it's silly, and a waste of time.

Also, both examples are completely irrelevant to any SEO discussion, why on earth would I, or anybody else for that matter, want to rank on any of those keywords? As far as what openlinkprofiler.org is doing, a bunch of sites do that, it's a massive database of domains and they aggregated data from a bunch of data providers, I explained how to do this in another thread. I'm really confused as to why anybody would want one of those sites, what good does a bunch of misclick/irrelevant traffic do for anybody?

I cased out a project a few days ago where a guy did about 30-50 guest posts and ended up being very competitive on keywords with commercial intent and ended up with a 20k revenue a month business. I did this very quickly since staring at other people's backlinks is a complete waste of time. I'm pretty much done with this discussion as well and I hope the people who are building links who are making this mistake got the message.

The fact that there are sites out there with a bunch of links, ranking exclusively on noncompetitive keywords and junk, is not interesting and I've spent enough time "debunking" this. The only reason it works is because they have tons of links and honestly, compared to a site like neilpatel.com, when I look at the organic report, it's obvious that they're not doing it right.

Share this post


Link to post
Share on other sites
6 hours ago, mki said:

Your definition of 'ranking good' is really odd in my opinion. I analyzed each of your responses very carefully and explained both cases.

In the case of xmasclock.com, you're going off the data in the pie chart, which as I already explained, it wrong, it's an estimate. At one point in my life, I realized, that once a domain has so many links, the pie chart is wrong, so I stopped looking at it. I'm fairly confident that it doesn't factor in nofollow links or sitewide links, which Google handles both of those differently, according to Google reps.

In both examples, the sites are not competitive ...

Being #3 on the keyword " imgchili " isn't doing openlinkprofiler.org any good and nobody trying to make money is going to try to compete with xmasclock.com, it makes zero dollars. Actually, it loses money. If I go into SEMRush, pull up openlinkprofiler.org's organic keywords report, then filter by "SEO" or "link" it ranks on absolutely nothing valuable. This reminds me of another conversation I had elsewhere where I was talking about directly targeting extremely competitive keywords and pointing a bunch of links at it; will you whiff on the target keyword?, of course, but Google will "find keywords to rank your page on." Which after careful thought about this, it's silly, and a waste of time.

Also, both examples are completely irrelevant to any SEO discussion, why on earth would I, or anybody else for that matter, want to rank on any of those keywords? As far as what openlinkprofiler.org is doing, a bunch of sites do that, it's a massive database of domains and they aggregated data from a bunch of data providers, I explained how to do this in another thread. I'm really confused as to why anybody would want one of those sites, what good does a bunch of misclick/irrelevant traffic do for anybody?

I cased out a project a few days ago where a guy did about 30-50 guest posts and ended up being very competitive on keywords with commercial intent and ended up with a 20k revenue a month business. I did this very quickly since staring at other people's backlinks is a complete waste of time. I'm pretty much done with this discussion as well and I hope the people who are building links who are making this mistake got the message.

The fact that there are sites out there with a bunch of links, ranking exclusively on noncompetitive keywords and junk, is not interesting and I've spent enough time "debunking" this. The only reason it works is because they have tons of links and honestly, compared to a site like neilpatel.com, when I look at the organic report, it's obvious that they're not doing it right.

 

 

 

 

You're only fascinated by 3rd party metrics which will never rank a page.

I've proven your OP theory is wrong, the page ranks for multiple keywords and tens of thousands of monthly traffic. Hell they even rank #1. So from what I see your definition of ranking good is weird. What more do you want?

I didn't base what I know off any pie chart, I posted the goofy pie chart because you rambled on how Majestic was so awesome with data and that data (your data source) proved you wrong.

I bet that Christmas site consistently gets more organic traffic than you do. So casting it off as not being your perfect example to fit your needs is absurd.

 

Quote

The only reason it works is because they have tons of links

No shit Sherlock, you're just now figuring out backlinks rank pages? Oh.

I got the message, you like to read SEO blogs and over analyze things that don't matter (ex: fake metrics). Here's some real news, most so called SEOs/sites are only posting crap to get you into a sales funnel.

...and Patel sucks.

 

 

 

Share this post


Link to post
Share on other sites
11 hours ago, yukon said:

You're only fascinated by 3rd party metrics which will never rank a page.

I've proven your OP theory is wrong, the page ranks for multiple keywords and tens of thousands of monthly traffic. Hell they even rank #1. So from what I see your definition of ranking good is weird. What more do you want?

I didn't base what I know off any pie chart, I posted the goofy pie chart because you rambled on how Majestic was so awesome with data and that data (your data source) proved you wrong.

I bet that Christmas site consistently gets more organic traffic than you do. So casting it off as not being your perfect example to fit your needs is absurd.

 

No shit Sherlock, you're just now figuring out backlinks rank pages? Oh.

I got the message, you like to read SEO blogs and over analyze things that don't matter (ex: fake metrics). Here's some real news, most so called SEOs/sites are only posting crap to get you into a sales funnel.

...and Patel sucks.

/eyeroll I downloaded the data and auto summed the rows.

Here's the thing dude. I have rankings on some affiliate sites, where that specific page, has a link from a guest post and that's all, and the DA is like a 14.

There is a lot more to this then "have a ton of links" ...

"No shit Sherlock, you're just now figuring out backlinks rank pages? Oh."

I literally posted a guide on how to build 50-500$ a month affiliate sites in 2-4 weeks and posted how long it takes them to rank. Am I suppose to post keywords as well? Oh wait I did along with information on how to find them yourself... And yes, you have to rely on the metrics to sort through thousands of rows of data and have experience, or this is a waste of time. Yet you continue to result to what I can only interpret as childish insults.

Regarding Neil Patel: He sucks? He has a DA 77... WTF are you talking about? He's a marketer, I've been doing this since for a long time dude. I literally sign up to mailing lists of people like Alex Becker to steal their ideas/email copy/email titles... Neil Patel sucks, LMAO... That is like saying that Penn and Teller are shitty magicians because they don't actually do magic on their show "Fool Us." Also, I don't buy their stuff man, why the hell would I? The guy who literally wrote the book on SEO and owns a legitimate SEO agency that works with fortune 500 companies most likely does consultation calls for less then the marketers charge for a course. He also gives away the same information those marketers sell.

Also, what's with the trash talking?

In the private section, you're trashing my interest in a domain because it was revealed how they built 90% of their links.

If your goal is to get me to leave the forum, I am considering it.

I'm not the kind of person that hangs around places very long.

And if you haven't figured it out yet, I didn't make the OP for you...

I was recently muted on a different forum, so if you would like me to go, just let me know. I'm kind of done with posting my life story up for people to read.

Edit: Also, regarding my awareness of courses, dude I market courses, I would hope that I was aware of what's out there and what it teaches. How am I suppose to market them if I don't know about them?

Share this post


Link to post
Share on other sites

I know there are a lot of arguments over metrics. I don't put much stock in them, but here is what I do use them for. I use them to save myself time. If I see a page has zeroes across the board, I'm not going to waste my time investigating it any further. If I see a page has a higher PA or TF, I'll take a closer look at it then.

I do not make any decisions based on the metrics on their own though. Well, sometimes if they are exceptionally high I might. Moz, for all their faults, is pretty decent at judging pages on the high end of the scale. If you see a page with a PA over 65, it is probably a solid page. 

As for anchor text diversity, I'm not one to really pay attention to any specific percentages. One reason is if you try to study it looking at other sites, there is no way to know what the actual percentages are anyhow. Majestic doesn't find all of a page's links. It might say one anchor's overall percentage is 20%, but in reality it might only be 2% because Majestic missed a lot of links. On the other hand, it might also be higher than that. 

What I do believe in is using a big variety of anchors. One reason is something I saw Mike say one time that stuck in my head. You are giving Google more LSI information about your page by using different related anchors. I don't want Google to just think my page should rank for 2-3 specific keywords. I want Google to think my page should rank for ALL related keywords. 

The second reason is that Google has said that something like 20% of the search queries they get on a daily basis are searches they have never seen before. The more variety I use, the better my chances of ranking pages for those unique search queries. 

 

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.