We've recently gotten to do (and see results of) some testing around internal links & anchor text and have come to some interesting, if not 100% proven, conclusions. As a disclaimer, SEO testing, whether done in a controlled environment or on live sites, is both challenging to quantify and subject to fair critiques both practically and academically. That said, I feel fairly confident in stating the following findings.
NOTE: These hold true primarily (and in some cases exclusively) for Google. I don't have enough data/experience to say whether the same results would be found at Yahoo! or MSN/Live.
- Internal Anchor Text Has Very Little Impact on the Homepage
Although linking to a page like our Web 2.0 Awards with the anchor text "web 2.0," for example, appears to provide a positive nudge, the same does not hold true if I link to the homepage of SEOmoz with the anchor text "SEO." In fact, for those who put a great deal of time into optimizing their links back to their homepage to say "new york apartments" or "antique rifles" rather than "home," I'd test whether modification of the link anchor text has any adverse rankings impact. In my recent experience (on a few different sites), the answer is no.
_ - Excessive Internal Anchor Text Linking / Manipulation Can Trip An Automated Penalty on Google
I recently had my second run-in with a penalty at Google that appears to punish sites for excessive internal linking with "optimized" (or "keyword stuffed anchor text") links. When the links were removed (in both cases, they were found in the footer of the website sitewide), the rankings were restored immediately following Google's next crawl, indicating a fully automated filter (rather than a manual penalty requiring a re-consideration request).
_ - Beyond a Certain Point, Adding More Internal Links to a Page Does Not Necessarily Flow More Anchor Text Value
Adding internal links from many important pages does provide ranking benefit, but it appears not to matter whether those links contain optimized anchor text after the first few (or few dozen, depending on site size). Perhaps there's a limit to how much anchor text value can flow from a single site to itself (or to another site, which might explain the diminished value of sitewide links) or maybe Google discounts optimized internal anchor text after a certain point. Whatever the reason, my general suggestion going forward is to link to pages with optimized links to help with link juice flow, but not necessarily obsess over anchor text (at least, until I see data that says otherwise).
BTW - This doesn't mean I'm abandoning my recommendations for internal linking with good anchor text for either usability, keyword cannibalization or to provide an initial boost. It just means sitewides (or pouring links from 500 pages instead of 50) aren't going to make the "recommended list" of SEO tactics for many future clients.
I'd love to hear your experiences in both testing and live scenarios and any speculation you've got as to why these phenomena appear to hold true. As always, I recommend conducting your own tests and applying what shows the best results for your sites.
BTW - For those interested in SEO experimentation, I've described the process in the past here and here.
It's amazing how often internal linking is ignored. That said, the opposite often seems to be the next extreme... internal linking excess.
A little bit can go a long ways and I'm not surprised to hear your findings that there are diminishing returns.
Rather than using internal linking to be a magic bullet to skyrocket a page to the top of the SERPs, I think it is best to focus on general link building best practices, which will often guide you to the right balance.
Add:
As in FAQ style?
More details on the testing - it was performed on several large sites (we recently made some suggestions on client websites and saw the results of those, as well as on test domains). I also wrote this post because of some feedback from site owners in Q+A and over email which suggested the same effects were being felt.
Again - I'm suggesting that I feel comfortable with these based on my experience, not that everyone will find the same thing or that these are 100% confirmed. As to specific questions:
Sorry for my delayed response - this week is packed, so my time is at a premium (who knew January would be so busy!)
Rand,
Thanks for the great tips, I dispursed them among my peers here at 360i. We appreciate your research!
Jeremiah Smith
SEO Technologist - 360i
Good article Rand,
I found that anchor texts from internal links does not help you much in climbing up the ladder. However "anchor text links" from homepage or the most active pages does seem to have an impact.
As a safe practice, I try to use anchor text variations from internal links, and thus following no particular pattern.
Great post Rand! My only question would be how do you know when you have too much anchored text on a page? Also, you mentioned it was your footer links that caused the penalty so does that mean links within your content isn't as scrutinized as footer links are by the search engines?
Consider nytimes.com usage of an h2 link "Politics" at the top of every page under the Politics category. Now Google Politics.
I don't have data of where they ranked prior to this change of if they had other implentations that I'm unaware of. But, its interesting either way.
And what about optimized bread crumb usage. Would that be over linking and possibility trigger the penalty that Rand experienced?
https://www.nytimes.com/pages/politics/
Rand,
These results coincide with our own studies into "2nd Page Poaching". In our findings, it became clear that PageRank Flow and on-site Anchor Text usage are primarily valuable for small incremental positions among non-competitive terms in non-competitive positions (ie: moving a page from ranking #11 for a non-competitive term to #9 or #10)
Adobe.com used to rank in the top 5 on Google for terms like Government, Education, Manufacturing and everything else that was in their navigation. Over the past 2 years their rankings have been sliding for these terms and I'm happy to see that they're out of the top 10 for all of those terms and 30+ for some.
Having them in the top 5 for Government was pretty ridiculous (especially with the LinkScape toolbar showing so few links to that section) and it's another example of Google slowly devaluing internal links.
Thanks Rand,
I find it interesting that anchors back to the home page really didn't do much. It makes me wonder what else works for deeper pages that may not have the same effect on the main/home/index page.
Also, would it be possible to hear more about the test itself? The criteria you used and the data that was collected?
Thanks!
I like using anchor text but haven't been able to see a direct co-relation between using it and my SEO efforts yet. But you've tempted me to set up an experiment using a niche client with not-so-competitive keywords.
One of the reasons I do like using anchor text is to simply try to move users through the site more smoothly. If a user visits more pages when they get to your site and stay longer, those stats must mean something for your SEO in a round-about way?
Hi folks, I have a question with regards to the number of "internal links" on a site according to "Advanced Link Intelligence Report". Linkscape only indicates around 400 internal links for my website, whereas I know I have around 30.000. My competitor has around 40.000 external links or at least SEOMoz Linkscape indicates that numer? How does Linkscape pick up on those numbers?
Thank you for helping!
JC
The current Linkscape index favors broad coverage of the web, rather than deep internal link structure analysis. So any "internal link" numbers we show (currently) will be much lower than you expect. But they should still be reasonable for relative comparisons.
This will likely change before we leave beta.
I could probably add some research to this post in the next few weeks... I have a large site that currently 302s lots of URLs so no link value is passed (I hate the term "Link Juice"). I have a build scheduled to 301 URLs. The site has optimised (not over optimised or keyword stuffed) internal links for various key pages. I am running a ranking check on the pre 301 site and I will see what happens once link value starts to pass. If I find anything cool I will post it on here!
It makes sense that anchor text in navigation elements would be devalued due to their repetitive nature however incurring a penalty does not make sense. If the anchor text matches the content of the target page, which one would hope is the case, then using that pages top keywords as the link anchor text actually benefits the end user experience. Google punishing good navigation structure from an end user perspective makes no sense.
I can't argue against what you've observed but if navigation anchor text becomes a negative algorithm factor rather than just being discounted it ends up hurting well designed sites. Now if the footer anchor text was a redundant set of links to the normal navigation structure and it used keyword rich links that were different from the top navigation anchor text it would make some sense to penalize for over optimization since there really isn't much benefit to the end user and the anchor text changes are clearly intended solely for search engine rankings. Was that the case in the footer nav or where they links to pages not included in the main navigation theme?
Anchor text in boilerplate (site nav, footers, etc) has limited incremental value, but you still need to get it right in the global navigation.
Matt called out an over-optimized footer during the site clinic at pubcon and identified it as looking "pretty spammy", telling the site owner to get rid of it, so no one should be surprised by that finding.
The only significant internal anchor text value you will see is through inline citation text links. Yahoo (remember them) seems to be particularly sensitive to good inline citation.
Rand,
When you mention that internal anchor linking has very little impact on the homepage does your experience refer to navigation links, links within content or both?
Once again, any SEO tactic that is too easy to game the search engines and doesn't provide value to users will fall at some point.
Good experiment Rand.
I also recommend clients as general housekeeping it is important to ensure your links are descriptive as possible....
As well as SEO's performing internal link optimisation for clients, you can save both time and effort by providing training sessions for content authors. This always works well....
Another small piece of the SEO Jigsaw....
Shahid
Rand can you tell us if your experimental website is old enought to suffer this kind of penalties that you mention?
I'm asking this because I see many old websites doing this kind of over-optimization and Google does not punish them.
BTW nice post.
Great post and thanks for the information. Makes complete sense to me. While it's something we still recommend, it is definitely not something we focus much time on. I think it can help from a usability standpoint, but if you "try" to over optimize the site by doing this, you are gaming the system and Google just doesn't like that : ) Or at least they won't rewardyou for doing it.
Scott
Very interesting Rand. I've always wanted to test the effects of internal anchor texts.
You stated that excessive internal anchor text linking can trip an automated penalty, what do you consider an excessive amount? Thanks for opening our eyes to this.
I honestly can't say where the limit lies, but in both cases I observed, there were in excess of 30 links sitewide that were clearly anchor text optimized to push rankings.
Do the 30 links include footer links, or is it just content links?
How do you know you are excessive with internal linking ?
does 10-15 linking to home page from all internal pages considered to be punished by google ??
Rand,
Or it's a matter of making sure you don't over index too far against other sites that Google feels are similiar to yours.
100% inbound links with a targetted phrase is gonna look fishy if all your competitors have 20% targetted and the rest non-targetted but the content from the inbound link's page is similiar.
I deal with this stuff a lot. There is definitely a ceiling at which more inbound links from your own site don't do you anymore good (see 'Barack Obama' inbounds for Chicago Tribune). I will go even farther to state that there is a ceiling to the number of inbound links period that do you any good. At some point a webmaster 'nails' the Popularity aspect of the algorithm and you can't get any more points from that section of the algo. You need to then focus on the Relevancy and Authority aspects of the algorithm. Sure inbounds can help with Relevancy and Authority too but the inbounds stop helping you for Popularity. Thus link building campaign becomes more about finding Relevancy and Authority than finding Popularity (i.e. number of links no longer matter to improve SEO the Relevance and Authority of those links matter).
In a perfect world a site would gain thousands of inbound links from relevant and authoritive sites but . . . we know that's not always possible. When it's not possible the number of inbounds will only get you so far.
Great post Rand . . . sorry I don't frequent SEOmoz more often. I'm slammed with trying to fix Tribune's network of media sites. ;-)
Rand,
I have seen one website. In that website there are no content but they placed lot of links with their products anchor text in Top bar, in the middle and in the footer. And in all links they have added lots of Keywords in Title Tags of Links. I also checked the back links of that website... But it has no more back links
This website comes in the ranking last 3 months... So I think may be soon that website will get penalty (If Google catch it).
Hmmm, not sure about this. One of the most popular sites on the web: Wikipedia is all about internal optimised linked text, do they trigger some 'Google filter'? No. IMO it adds usability if it's on context also it makes sense to use it. SEO benefits hmm, probably works to some degree also.
Great Article and a great heads-up - I do have to agree with Jill Whalen - non-competitive and, perhaps, longer-tail should be thoroughly investigated as well.
As for the Wikipedia, their use of Internal Text Link-Optimized or ITLO as we call it around here, is completely reasonable given:
They provide the link as part and parcel of the large context of knowledge exchange and the linking structure of the internet itself. IMO though, there are plenty of entries on Wikipedia that could use some human authorization from Google if Google is in fact rewarding all their ITLOs equally.
@ Socialguy: but surely that's what you would do anyway? I wouldn't make a link appear out of context or that had no relevance to what document it pointed at. Also what about sites that have many hundreds of links in the footer a la insiderpages?
Oh yea, hot topic... It used to carry a lot of weight in the past --- but as of late, my experience has shown the same things you have outlined.
There really is no need to spend time optimizing hundreds of internal anchor text links. I just do a dozen at most per site and then move on to more important things.
Best,
Mike
I am not sure how navigational links that are clearly useful to a user (Google is all about the user experience) could be excessive? I have a feeling this may be in direct correlation to the number of indexed pages or number of external links as a % of one of the above?
Keyword based internal links are the key ingredient to quick user friendly navigation of large websites. Often these keyword anchor links are the only way to rapidly find what you’re looking for, something is being missed here.
If you are seeing a penalty and you think it is 100% quantity of keyword links penalty why not add a no follow to a few links every couple of days and see if the “penalty” is lifted, this would also clue you in to the quantity of acceptable links for that particular website.
Interesting research - I need to start having a look at some of our anchor text following a site redesign so this was well-timed.
I wonder if your experience re: targeted anchor text to the homepage comes down to usability. A homepage is the only thing that every website will have in common and it's easy for Google to distinguish it from all other pages. On top of that, users should be given safety nets against getting lost in navigation which is offered far better by a nice clear "home" link than by "New York Apartments" or "Antique Rifles". Just a theory - absolutely no evidence to support it :)
Erika
Rand, internal linking has been a big part of my strategy for some time. Your SEO and ny apartments examples are IMO, indicating less effectiveness because the degree to which the page is already affected by IBLs and internal contextual links for those terms ie a lot of these links also come from many sources with more authority. Whereas the lesser terms or terms not strongly associated to your site by IBLs and contextual internal links have far less competition for the new term you are trying to position higher.
I actually got hit with this penalty (internal link anchor text overload) back in 2004/05. We were experimenting with our on-line retail site and discovered back then that even with extremely competitive terms we could rank well with manipulation of this technique.
Then all of a sudden we were penalized. People were runnigng in a panic and i recommended that we take down the "SEO'd" internal anchor text and replace it with the old standard. The rankings came back for our "stanbdard" terms within 30 days.
Then we started manipulating external anchor text severely, which again it worked like a charm, then one day.... You get the picture.
The overall moral of the story is when you find somethign that works well do not overuse it.
Right on the money. I've been experimenting with the overuse of 'optimized' anchor text in internal links as well and have concluded the same thing. I was dropped between 5-20 spots that were immediately restored when I removed the excessive anchor text.
internal body links VS internal footer links
What is everyone’s opinion on internal body links and internal footer links? If you have a fairly new site with many back end pages that need to be crawled on a regular basis. Is it A. - better to focus on natural looking internal links in the body of your pages or B. – your normal SEO internal footer links? I always try to combine my use of these internal links but always seem to put more focus into natural looking internal body links.
Surely the internal body links are crawled first and seem natural, as oppose to the internal footer links. Furthermore, by using internal body links, you have a degree of flexibility with your keywords and anchor text. With internal footer links they always need to be organised and have some structure which to me makes them seem unnatural and fake. Also they are in the footer and wouldn’t this have less authority than internal in text links with the body copy?
Do you think you are more likely to get penalised for using internal footer links or internal body links?
Sorry about that ramble, any feedback or opinions would be great.
Cheers
Good post
whenever something is over done for search thn it will not reap benefits.Overdo for the reader.For SEO follow basics to reach reader.
Complimentary reading -
https://www.webmasterworld.com/google/3756885.htm
Rand,
Keep up the good experiments.
The internal linking to the home page with targeted anchor text worked for a while, I believe it was 2004 when my site was hit with a very significant penalty for it.
Since then, I only utilize page specific anchor text topic specifc linking to the home page to a small extent and it seems to have a minor assist. I have never tested the point of diminishing returns with this technique though.
I like to build up an internal pages page rank and then link off of it in the context to a related page, which seems to offer some benefit as well.
very good!
来自中国的SEOer,像你们学习了！
I have just changed our internal links from default.asp to https://www.proseo.dk and is now wating to see the effect.
If it has a big effect I will be very surprised, we have been steady in the bottom off page one or top off page two with our primary keyword.
So if that all I have to do to move a couple off places up I will let you know.
Great site you have here...
Peter
Hi peter I dont know if there is major effect from linking to the default page itself or to the root domain name. I will be glad to know what the effects are.. thanks
1. Internal Anchor Text Has Very Little Impact on the Homepage
Did the link to the homepage with keywords appear below that with "HOME"?If so, would be this applied for?
https://www.seomoz.org/blog/results-of-google-experimentation-only-the-first-anchor-text-counts
nice post, ive recently updated the links so that there is a 70/30 split around internal links using two different keywords, the only problem is that if you leave out one link or have a single broken link this can give your ranking quite a knock down the SERPS.
ive run some tests with linkscape and some sites which were getting 3-5% benefit are now getting 0.01% link juice. The site or links have not changed, and their ranking in SERPs but im guessing if they change the structure of the site, the new site will not rank quite as well.
I am surprised that anchor text to the homepage has no real effect.
I haven't been doing this too much, but now I know that it doesn't really help anyways. Thanks!
-Brenelz
Interesting.
Rand, did you test a variety of keywords pointing back to the home page, in terms of their competitiveness?
Obviously the one-word "SEO" is super-highly competitive and not going to be necessarily indicative of what might happen if you were to use a much less competitive phrase in the links back to the home page.
great post Rand and thanks for sharing.
From my experience, we always try to have some specific keywords in those links, although it has not been proven that it did help. but I guess having those KW in the content and in the links could only benefit one on another (is this English?).
Good post. I didn't know that internal linking can lead to penalties at all ... wouldn't it be more fair to just ignore those links and not pass any linkjuice?
Great post. i was wondering about this aswell for the past few sites I've set up, and a couple I optimised recently. Great to see that you have tested it.
I have experimented a little with anchor text also and found that putting too much seemed to penalize me. I have found the balance I believe.
Thanks.
Thanks Rand - this is very interesting I've always been interested to know how much effect the sites neighbourhood has on rankings. If you manage several sites on the same server or bank of servers (but the domains are registered with different people/registrars as is often the case) how much weight is carried by using optimised anchor text and linking between them. I've never analytically tested this but it does appear to benefit ranking and it is something I'm hopefully going to test out this year at some point. If I find anything interesting I'll add it to YOUmoz.
That's true, I've just had that experience on my client site...Never do that again
I always enjoy reading test results - thanks Rand for sharing. Just tweeted saying as much.
Interestingly, I think there is a broader match here that 'over optimisation' is always one to watch out for...the more natural everything seems the more likely thresholds are not passed. It's why I'm still keen to use a wide variety of synonyms, stop words, etc...
Great post Rand, cheers!
Thanks for the thought provoking blog post Rand! I alway enjoy the stimulation whether I agree or not. In this case I agree for the most part, but I can not verify nor deny that Google is penalizing for this.
As with everything, we use sitewide footer links with moderation.
We use to see huge gains in the SERPs from sitewide, internal footer links with anchor text. I've never noticed a penalty necissarily, but it does seem that the jump we use to get from doing this has been virtually unidentifiable for at least the past 12 months or so.
Slightly off topic, but interesting none the less, we noticed a major increase in positioning for a competitive phrase after an extremely content related linking partner removed sitewide footer links pointing to our web property in exchange for five in-content links on higher value pages. None of the new links were on the home page of the linking partner's website (PR 7), but were on a couple PR5 and PR4 internal pages. We noticed a 12 position increase that has been sustained for nearly two months and counting.
The main area in which I highly recommend using good anchor text on internal links is when the page is not included in some part of your navigation.
It's more appropriate (IMO) to include generic terms like "company" or "services" as a link in navigation, but when you need a deeper page to get indexed outside of your navigation, I find it most important to link on good anchor text without losing the flow from the context of the page.
I've just set-up a test for this myself. I did an internal link on one of my cllient's major target search phrases where a current page ranks #7 (local listing is #1). I will check back in a week to see if results have changed. PM me if you want follow-up.
this is definitely very useful. I wouldnt say and confirm that I get a hell of a lot of good results from using optimised anchor text on internal links, but i have definitely seen one of the main websites (300 pages) I manage get higher rankings within a specific timeframe for targeted keyphrases I have been optimising for in the internal links. I am going to try and monitor this more closely from now on... all in all, a very intersting post and comments, thanks, guys
"Excessive Internal Anchor Text Linking / Manipulation Can Trip An Automated Penalty on Google" Please define "excessive" Rand. :) Thank you.
I can't Marty! I can only say that some degree of excessive got these two sites penalized. Finding where the line is drawn would require willingness to do a lot of burning in your testing and I don't have the bandwidth or resources :)
A couple of comments on the comments to this post.
How do you know a site is being penalized for excessive keyword stuff anchor text unless you test without it?
Is it really a penalty, or do you hit a value limit with idential links, and when you use a variety of anchor text you get the value limit again and again from each anchor text variation?
A few years back this would work, but only for a short time. Perhaps once Googlebot had found a large number of links and the penalty threshold had been crossed did rankings tank.
I'd expect to see far more of this from Google as they try to weed out over-optimisation (and perhaps even optimisation).
Might be that to rank well you've got to get very clever and copy some aspects of worst practice...
And then you've to question whether you're in it for the SERPs glory or for the conversions.
This is great information to know, thanks for doing the experiment!
P.S. Ever since the new launch of the website all the blog entries the font is really hard to read. In IE 7 it is fine but for some reason in forefox 3 the font is very hard for me to read. Anyone know what is wrong?
IMHO, there are many aspects what can change the value of any on-site links.
Regarding penalties,
I think just in special cases can get penalties, anchor amount normally will not a sign of manipalation.
But here is one idea for a possible reasons:
Same anchor texts in repetaed blocks in mass on the pages, targeting authority names (very popular brand names which have associated own domain with the phrase) in the text.
How do you know if the penalty from Google is because too many internal links. Do you really know why? anyone that tells you if you do wrong because too many internal links?. google does that tell you about it or is it just your speculation? I increase the number of internal links with 2 of the reason. The first is to share the popularity among the articles. and the second is to increase my pageview Thank you