Internal Linking Anchor Text

Joined
Aug 16, 2022
Messages
5
Likes
4
Degree
0
I've done next to no work on my internal linking for years now and have always been advised to just use exact match for all internal links. I've become a bit more SEO savvy now and think that it weren't necessarily great advice.

How do you guys handle internal linking and anchor texts?

When looking online, I'm essentially finding a load of Neil Patel style "Find out how I Increased Traffic by 40% through internal linking" that doesn't give any detail at all about the stuff that actually matters!

Any advice on this would be massively appreciated!
 
You definitely want to use descriptive anchor texts relevant to the page. Whether that's exact match or partial match... there lies the question.

I think that Penguin came around and destroyed sites that were manipulating their backlink anchor text profile. It's a clear sign that you're either building links or buying links. But there's zero rules about building links within your own website to pages within your own website. But the paranoia carried over to internal anchor texts.

Probably the smartest thing to do is use variations of keywords and phrase matched anchors. Like, if the keyword was "exercise for dogs" you might use:
  • exercise for dogs
  • dog exercise
  • exercising your dog
  • the kind of exercise a dog should perform
  • what kind of exercise is best for a dog?
And so forth. That's probably the most future proof thing you could do.

However, I'll be frank. Google already tells us a lot of things we need or must do (or must not do) within our own websites to see success in their rankings. But them telling us how to perform internal linking is over-stepping the boundary.

The frankness is that I only use two types of internal links in my posts: exact match in the content and at the bottom in a static related posts section I use the post title. Otherwise it's menus in the header, sidebar, and footer. I push a little hard here because it's my damn site, you know. A line must be drawn somewhere. But I don't mess around with external anchor texts. Per post I'll use one single exact match anchor and no more. Everything else in natty or phrase match.
 
Patel is right, internal linking on a site that was previously lacking can be massively impactful. I've seen newer pages rank for competitive terms with large volumes with concerted interlinking alone.

My interlinking strategy is simple: compose the piece naturally and then during the editing process of formatting and fixing any types I add relevant internal links as well. Doing it this way helps ensure I don't get to carried away because I'm not writing for the link placement, the link placement fits naturally within the writing. The existing content will dictate the anchor based off the flow of the article itself.
 
After getting my head wrapped around the backlink convo in the Core Update thread, I think the next project I need to tackle will be my internal anchors.

I've personally used a combination of exact and partial match keywords across my site, stitching together a three-tier silo structure. My interlinking structure is extremely intentional. But, anchor text is something I've never really understood.

So, I found this thread and thought I'd ask @Ryuzaki if you are still following this approach regarding your internal anchors...
The frankness is that I only use two types of internal links in my posts: exact match in the content and at the bottom in a static related posts section I use the post title.

Also, in terms of the keywords variations that you mentioned, is this still sound advice?
Probably the smartest thing to do is use variations of keywords and phrase matched anchors. Like, if the keyword was "exercise for dogs" you might use:
  • exercise for dogs
  • dog exercise
  • exercising your dog
  • the kind of exercise a dog should perform
  • what kind of exercise is best for a dog?
And so forth. That's probably the most future proof thing you could do.

Basically, I'm looking for suggestions on how to think about internal anchor text as I get ready to tackle this project. So, any suggestions would be welcomed!
 
I'm always looking for info on anchors as well.

Personally I'm not using optimized anchors at all anymore, especially externally. All NAT from now on (specifically, @Grind 's approach - natural language anchor very near your targeted keywords, but not actually containing it.)

For internal anchors I've used exacts exclusively and did see a pretty significant jump in rankings, but a quick reversion to original levels pretty soon afterwards.

I've now de-optimized all my internal anchors and alt-tags across my sites. Did not see any significant downward movement from this.

My hypothesis on the best approach to SEO nowadays is that everything has to be under-optimized in the old sense of concept.

Your keyword in page title, URL and H1 - that's it, full stop. Obviously your broad/phrase/related keywords will naturally appear within your content, so I don't ever worry about that.

But even for internal link building - I'll try to build links where the keyword I'm targeting is close in proximity to the natural anchor'd link. For external links I'll try to have the keywords I'm targeting in the title/url/h1, but never put any keywords within the anchor.

I feel like the less you do "best practice" SEO, the less chance you'll trip some wire that blows you up.

So no more keywords in anchors is my current approach for all links, internal and external. But I'm very curious on other people's approaches.
 
natural language anchor very near your targeted keywords
Can you give an example of how you do this? I see/hear lots of people say this and then wonder "What does that mean in practice?"

I've now de-optimized all my internal anchors and alt-tags across my sites.

That's really interesting - basically what I'm planning on doing - a few follow-up questions if you don't mind:

1. Do you mind sharing when you did this (month/year) roughly?
2. Was this across all existing content or only new content?
3. If you updated existing content, did you only update anchors or each piece of content more broadly?
4. And what was the catalyst for this change? Following advice or did something compel the change?

But I'm very curious on other people's approaches.
I implemented exact match anchors across topical clusters in '21-'22. My best guess is that 80% of my internal links are exact or partial matches (~1500 pages). I still have all of these existing anchors in place and have just started to use natural anchors for new content. So, now I'm trying to figure out the next steps and how to go about updating effectively.
 
Can you give an example of how you do this?
"The best way to learn the ins and outs of dog training is to think like one."
if your keyword is dog training, the anchor text could be "think like one" or "is to think like one" or "learn the ins and outs"

1. Do you mind sharing when you did this (month/year) roughly?
2. Was this across all existing content or only new content?
3. If you updated existing content, did you only update anchors or each piece of content more broadly?
4. And what was the catalyst for this change? Following advice or did something compel the change?
1. late last year
2. all content
3. updated whatever i needed to depending on situation.
4. like I said, wanted to under-optimize
 
Thank you for the feedback - very helpful!

wanted to under-optimize
Sounds like you were ahead of the curve!

"The best way to learn the ins and outs of dog training is to think like one."
if your keyword is dog training, the anchor text could be "think like one" or "is to think like one" or "learn the ins and outs"
Okay, that's interesting.

I always thought if you were linking to an article targeting the keyword "dog training" you would have looked to insert a synonym or closely related terms like:

- canine discipline program
- doggy boot camp
- puppy development course

I didn't realize that you would use the terms surrounding the keywords that didn't reflect the target keyword.

I always worked on the premise that I should include words inside the open and close anchor tags that reflect the content of the target page, since that would signal to google what the target page is about, and then, in turn, would build relevance and authority.... though that's me over-optimizing.

And you've changed all the anchors on your site to follow a similar pattern?

Do you have any general rules you follow in terms of anchors? Like not placing them at the end or beginning of sentences, only so many in content links per page, etc?
 
@Smith, at this point I still don't see any evidence of internal linking being a problem, but for the sake of future-proofing from anything Google may use in their less-than-fair agenda against SEO's, I would definitely stray away from only doing exact and partial match anchors internally. I would do what you're doing with synonyms, most certainly.

I would also do what @Politico is suggesting simply for the word and phrase variation in the anchor text profile. But I must admit that I've yet to see evidence of this "proximity anchoring" working, where your phrase is near the anchor text but not a part of it. I also never saw evidence that "unlinked citations", meaning dropping your brand name or your raw URL without a link really provided any benefit either. Google may crawl through those unlinked URLs, but I never saw them work. There was a lot of theory crafting at a certain time regarding all this kind of stuff and it never gained broad adoption because it never worked. If it did, you can be certain it would be exploited.

I'll say that same thing about other very commonly known things that are now being touted as the sole, silver bullet solution to current ranking issues. If it worked, everyone would be doing it, especially since it's not new and everyone has already done it to no avail. There's agendas coming from Google (the gold mine) and agendas coming from the pick axe sellers (service providers) and even from the gold diggers themselves (SEO's looking for the secondary benefit of clout and reputation enhancement). The last thing we all need to be doing is spinning our wheels with ABC busy work when there's a large enough sample population that's already done it to no effect, despite singular loud voices saying it works.
 
@Smith, at this point I still don't see any evidence of internal linking being a problem, but for the sake of future-proofing from anything Google may use in their less-than-fair agenda against SEO's, I would definitely stray away from only doing exact and partial match anchors internally. I would do what you're doing with synonyms, most certainly.

I would also do what @Politico is suggesting simply for the word and phrase variation in the anchor text profile. But I must admit that I've yet to see evidence of this "proximity anchoring" working, where your phrase is near the anchor text but not a part of it. I also never saw evidence that "unlinked citations", meaning dropping your brand name or your raw URL without a link really provided any benefit either. Google may crawl through those unlinked URLs, but I never saw them work. There was a lot of theory crafting at a certain time regarding all this kind of stuff and it never gained broad adoption because it never worked. If it did, you can be certain it would be exploited.

I'll say that same thing about other very commonly known things that are now being touted as the sole, silver bullet solution to current ranking issues. If it worked, everyone would be doing it, especially since it's not new and everyone has already done it to no avail. There's agendas coming from Google (the gold mine) and agendas coming from the pick axe sellers (service providers) and even from the gold diggers themselves (SEO's looking for the secondary benefit of clout and reputation enhancement). The last thing we all need to be doing is spinning our wheels with ABC busy work when there's a large enough sample population that's already done it to no effect, despite singular loud voices saying it works.
Thank you for the feedback.

I'll be looking at incorporating more synonyms and related terms as I update existing content moving forward. It's obviously hard to test on a live site as there's no feasible way to run single variable tests for anchors. So, I think aiming for under-optimized anchors like you said (straying away from exact and partial matches) makes sense.

Also, it seems like you have something particular in mind here:
The last thing we all need to be doing is spinning our wheels with ABC busy work when there's a large enough sample population that's already done it to no effect

And here:
I'll say that same thing about other very commonly known things that are now being touted as the sole, silver bullet solution to current ranking issues

Can you be more specific about the "ABC busy work" and "commonly known things" you're referring to?
 
"The best way to learn the ins and outs of dog training is to think like one."
if your keyword is dog training, the anchor text could be "think like one" or "is to think like one" or "learn the ins and outs"


1. late last year
2. all content
3. updated whatever i needed to depending on situation.
4. like I said, wanted to under-optimize
How did your sites do in these current updates after implementing these changes?
 
though that's me over-optimizing.
Yea my mind set right now is to not act like an SEO at all (as much as possible)

How did your sites do in these current updates after implementing these changes?
No drastic changes tbh, up or down, for the most part. March core update hasn't affected me really. One of the sites had a quick 25% dip in Oct, but has been on a slightly upward trajectory since. October was BEFORE I changed up all my anchors.
 
Yea my mind set right now is to not act like an SEO at all (as much as possible)
How far are you going though? For instance, are your Schema, CWV, and backlinks still clean? And what about the content itself? Have you de-optimized your style of writing and started inserting grammatical and spelling mistakes to signal "human error"?
 
How far are you going though? For instance, are your Schema, CWV, and backlinks still clean? And what about the content itself? Have you de-optimized your style of writing and started inserting grammatical and spelling mistakes to signal "human error"?
None of that shit matters in my experience.

Schema won't help you rank (but apparently you can rank a page with just proper schema and no on-page text content, but that's beside the point)

I've spent an extraordinary amount of time making sure my page speed was under 2 seconds and my CWV's were 100 on every single page on a site. IT DOES NOTHING FOR RANKINGS.

Is it good to have a fast page? Absolutely, for user engagement. If you're doing it to try and rank your page, you're focusing on the wrong thing.

De-optimized style of writing? I've never even contemplated what that would even mean. And "signalling human error" is, in my humble opinion, missing the forest for the trees.

I think you're over-thinking shit (and I'm often guilty of the same). You're thinking like an optimizer. Try to stop.

Just be. And do.

Try to be the guy on the left of the midwit meme.

KuRkj0w.jpeg


My entire strategy at this point I already gave you:

- Keyword in title, url and h1.
- get links (no keyword anchors)

That's literally it.

The only thing I'd add to that is this: write an article you'd actually want to read, and format it so people can skim through it quickly.

I'm ignoring everything else.

But also remember one thing - I'm the guy on the left, so take what I say with a few thousand grains of salt.

Way smarter people are making way more money than me. I'm just sick of all the nonsense in our space that doesn't move the needle.
 

It's like Nascar - if everyone has a 900 HorsePower car, but you come in with your Nissan Altima, you're in trouble.

I was doing some consulting work for a major agency that represents some big brands - at some point.

This girl I was tasked to work with was out of control with Mobile Page Speed needing to be above 90 for this weird CMS system, this was like 4-6 years ago. Thinking back I believe it was WebFlow. So I did my VooDoo and got it to like 95+ on Mobile and Desktop.

So we goto lunch afterwards and I asked her why she insisted on 90+, playing dumb. The industry her client was in all had page speeds above the 85+ mark for Mobile and to get ranking you had to be above 80+ - in Google search. I thought that was odd or she was insane - more than me.

I obviously had a look on my face like I didn't believe her. She then proceeded to show me a ton of evidence, experiments, and data the big players have done and low and beyond all their mobile speed scores are 80+.

That was the first time in my life I saw the top 20 results all with 80+ mobile, so they already knew that they needed the mobile score to be at 90+. The only problem was this WebFlow crap was still new and not known for SEO or pagespeed back then.

After I did my Voodoo SEO magic I gave them a breakdown of what I did and what to look for. It was the first time this agency had a step by step process for the Voodoo PageSpeed magic. I got my money and got out of there.

Later on I ran some tests against SERPWoo data, and there were almost no other industries that had taken page-speed seriously like that one. But I noticed pages with 90+ don't have the "emergency brakes on" so to speak.

It's not a ranking factor, but it's used to slow you down and cripple you and could hinder your upward trajectory with everything else being equal.

Look at your competition and see what their Core Web Vitality Score, jump into the 90+ area and you'll beat 95% of them. Even sticking to the 80+ area and you are good. After 90+ you are wasting your time with diminishing returns because a slow website with more relevant backlinks is still going to out rank you.

Hopefully that makes sense.

Same thing with Schema - having it is great, but not having it doesn't matter unless your competitors are ALL using it.

Watch the SERPs and see what they are doing for your target industry.
 
I'm just sick of all the nonsense in our space that doesn't move the needle.
Amen, that's why I said what I said below, which wasn't cryptic so there's no extra explanation needed to the person that was asking:

The last thing we all need to be doing is spinning our wheels with ABC busy work when there's a large enough sample population that's already done it to no effect, despite singular loud voices saying it works.
As I also said in that chunk of text, there's no silver bullets, magic button, singular solutions anymore. It's not 2010 anymore. And I promise that if anyone is saying there is, they're willing to do the work for you for money.

Way smarter people are making way more money than me.
Agreed, but what really made me aware of this problem is how many people who are not even just a little bit less smart, but astronomically dumber can end up making a ton more money. It's because they stick to the basics and then get to work.

I come from an quantum and nuclear engineering background, and when I got into SEO I treated it as if it was that complex. It was an easy assumption since Google was hiring the best in the world. Turned out all that was, was a hindrance.

I learned the ins and outs of everything you can learn in this industry, from coding up my own full websites, customizing CMS's, scripting bots, optimizing page speeds, NLP/Tf*IDF/LSI/NERD, all of it. All of the stuff that should give you a serious competitive edge. At the end of the day, it might provide the tiniest edge, but it's not something you can apply at scale, and scale was necessary. Now that scale of content seems to be less viable, I'm still not convinced any of that stuff is worth eating up your time when you can be doing marketing instead.

"ABC busy work", "spinning our wheels", "all the nonsense that doesn't move the needle". Amen to all of it.

Build a site, make it reasonably fast but don't worry about diminishing returns, publish great content for real people, use your keyword in the Title, H1, (not even convinced slug matters) and maybe once in the body. Then do marketing. And at this point, informational queries seem to be a dead end, which means we can move our scaling efforts over to marketing. Marketing marketing marketing.
 
Agreed, but what really made me aware of this problem is how many people who are not even just a little bit less smart, but astronomically dumber can end up making a ton more money. It's because they stick to the basics and then get to work.
This 100%.

Just knowing the basics, rolling up your sleeves and doing the grunt work that one is often avoiding while looking for said silver bullets.

Ive done (and do) that too much. Which is why I really just want to strip everything superfluous away and do the basic bitch grunt work I'm avoiding in search for the secret magic unicorn fart that'll somehow solve all my problems.

Build a site, make it reasonably fast but don't worry about diminishing returns, publish great content for real people, use your keyword in the Title, H1, (not even convinced slug matters) and maybe once in the body. Then do marketing.
Agreed. Like you said - marketing, marketing, marketing. Also the thing we're (read: I'm) avoiding while searching for unicorn farts.

It's like Nascar - if everyone has a 900 HorsePower car, but you come in with your Nissan Altima, you're in trouble.
This is also a mindset I'm fully on board with. Match and exceed the competition - on every level you can.
 
Quick follow up for @Politico and @Ryuzaki on the approach to internal links...

I've been digging into my internal links to remove exact match anchors. But I've realized I have ToCs on many articles... and these ToCs all include exact match anchors to H2s (some H3s). Do you count these as concerning exact match anchors that need to be changed or do these not matter in your opinion?
 
Last edited:
Quick follow up for @Politico and @Ryuzaki on the approach to internal links...

I've been digging into my internal links to remove exact match anchors. But I've realized I have ToCs on many articles... and these ToCs all include exact match anchors to H2s (some H3s). Do you count these as concerning exact match anchors that need to be changed or do these not matter in your opinion?
My initial inclination is to say no. They're #jumplinks rather than links from one page to another. Google at one point was talking about indexing portions of pages separately (like SEO pebbles or scribbles or whatever stupid names people were trying to coin for clout).

In that case it might have mattered but I'd say 99% of "jump links" work exactly like you're doing, exactly like Wikipedia does it. Google's hubris sometimes has them try to create rules and punish people for doing things the way they've always done it, in order to get them to do it how Google wants them to do it. I doubt this is one of those cases. If Google has any sense they'll see that this is a very useful method of improving user experience and leave it the hell alone.

An example of what I mean was when they tried to do the "Above the Fold" update and tried to get people to stop creating splash images for featured images. And to at least have the title and first paragraph above that image, etc. This was while they were trying to get us to not put ads above the fold. And not a single person made the change. We all collectively told them to kick rocks, and they soon decided that "images count as content" as far as the Above The Fold algo goes. That's what would happen with the ToC links. And I'm betting they learned their lesson on fighting about crap that doesn't matter.
 
My initial inclination is to say no. They're #jumplinks rather than links from one page to another.
Incredibly helpful, thank you.

This is what I thought but I was worried after re-reading the comments about exact match anchors - but yes, I agree that jump links like this enhance UX and should be a positive for Google.
 
Hoping for some advice on internal linking when you have a lot of similar articles.

Our lead gen site has a lot of templated articles. We're culling non-performers but we'll still have 30-40+ of very similar articles.

For example, our most popular template is a review of a company in our niche, where we've done 50-100 individual reviews. It's all human written so the content is unique, but the structure and focus of each article is 90% the same.

Many of our current internal links are using the same anchor text for each article. Templated content using templated anchors linking to templated content... forgive me for my sins

I'm going to wipe out all the duplicates and rebuild. It'll be like needing 30-40 unique but relevant anchors for basically one article. Been trying to figure a way to speed that up, but I guess shortcuts are what got us into trouble in the first place.

Just going to have to grind through it and come up with each unique non-keyword anchor one by one. Which will be hundreds, if not thousands in total... fuck.
 
Hoping for some advice on internal linking when you have a lot of similar articles.

Our lead gen site has a lot of templated articles. We're culling non-performers but we'll still have 30-40+ of very similar articles.

For example, our most popular template is a review of a company in our niche, where we've done 50-100 individual reviews. It's all human written so the content is unique, but the structure and focus of each article is 90% the same.

Many of our current internal links are using the same anchor text for each article. Templated content using templated anchors linking to templated content... forgive me for my sins

I'm going to wipe out all the duplicates and rebuild. It'll be like needing 30-40 unique but relevant anchors for basically one article. Been trying to figure a way to speed that up, but I guess shortcuts are what got us into trouble in the first place.

Just going to have to grind through it and come up with each unique non-keyword anchor one by one. Which will be hundreds, if not thousands in total... fuck.
As I said over here in the Massive Links thread, I'm leveraging Chat GPT to help me generate different variations and synonyms of my ideal anchor text as fast as possible... the key word is "help". I'm manually going through and updating each one and making sure the text makes sense but Chat GPT has been extremely helpful in quickly brainstorming alternatives. If you come up with a faster way please share!
 
Back