The ‘over optimization’ of anchor text has been coming up a lot recently in conversations that I have been having and has been the subject of a few recent blog posts. For the sake of this post and to quell any arguments stemming from the phrase ‘over optimization’, I am, for this post, defining the term as: building too many links with targeted anchor text such that they a) no longer provides value or b) actually takes away value – basically you’ve built too many targeted links and you’re not seeing your rankings increase.
When I have talked to people about this recently, I have suggested that a 7:3 ratio of non-targeted: targeted anchor text would be a good frame of reference for emulating a ‘normal’ link profile. I got curious about this though and decided to do some research. I looked at product and category pages on ten different websites - these websites are all large national and international brands/ecommerce sites that are well linked to. Between the ten sites I analyzed the anchor text associated with 28 category pages and 31 product pages.
For each of these pages, I downloaded the anchor text report from OSE and looked at whether the ‘Number of Linking Root Domains Containing Anchor Text’ were optimized or not. Specifically, I looked at if the anchor text had the following attributes:
- Brand Name - The brand name is the anchor text, e.g. Giant Bicycles
- Branded - The brand was included in the anchor text (these did not include anchors that were the brand name), e.g. this cool company, Razoo
- Keyword Branded - The anchor text combined a keyword (exact/phrase/partial match) and the Brand Name, e.g. a ton of motorcycle helmets at Dennis Kirk
- Exact Match - The anchor text exactly matches a keyword, e.g. Downhill Bike
- Phrase - The anchor text contained a keyword phrase or partial match anchor text, e.g. the monthly mountain bike magazine I get
- Random - The anchor text does not contain the brand or any keywords (or partial keywords), e.g. this website
- URL - There were no actual anchors, it was just a URL, e.g. https://www.malibuboats.com
I chose to only look at the number of domains linking with these anchors as site wides can disproportionately skew these ratios pretty quickly.
Finally, I looked at the sources of the links and did not include pages if it looked like they had links manually built.
Ok, now on to the interesting part.
Targeted vs. Non Targeted Anchors
Across category pages and product pages, I found that 34.6% of links were targeted (targeted anchor text collectively refers to exact match anchors and phrase while non targeted anchor text is everything else).
Here is a breakdown of this distribution:
Here is a simpler breakdown, consolidating brand related anchors:
Category Pages
If we take a more in-depth look at category pages, we find some variance from the collective distribution above. The data shows that only 25% of links to category pages are targeted - people are less likely to link with good keywords to your category pages.
Looking at the ‘other’ anchor text distribution, the number of links for branded and URL anchors increase 5% and 7%, respectively. Most of the gain in the branded links were keyword branded links.
Product Page
The product pages show a higher proportion of targeted anchor text, making the targeted and non-targeted distribution roughly equal.
Looking at the distribution of anchors for product pages, we find that there are more links with exact match and phrase match links to product pages than to category pages. Exact match links jumped up about 7% and phrase match jumped up about 4%.
So What Does This Mean
For a lot of people, this means you should probably decrease the amount of targeted links you are building and add in some varying anchor text. It is important to keep in mind that this research, while it was time consuming, is by no means exhaustive, so you shouldn’t take this as fact. That said, I think it gives a pretty good rough estimate of what normal might look like. I like to be a little more conservative so, especially with the product pages, I will probably keep trying to stick to the 7:3 ratio I first mentioned.
If you have been doing a lot of SEO work and are still having trouble ranking, this is a factor that you might want to look at as you may need to start building different anchors to balance out your profile.
The sites sampled spread several industries. Your industry will probably look a little different, as such, you should do your own research and determine what 'normal' looks like for your industry. To do this, just pull the anchor text report from OSE for sites from the SERPs. If all the pages ranking have a lot of SEO'd links, look at those sites and try to find non optimized pages and use those to help establish your baseline.
To help you do your own analysis, I made a Google Doc that will help you calculate these percentages.
Indeed Awesome post Geoff specially for those folks acting like Running horses in race. In Schools we were tought that over performing any task leads to unwanted and unfaithful and fake results. Here in the SEO world , it seems much applicable. Many times , people just keep on running behind the anchor texts and without making any estimate they keep on trying to get links for the same anchor texts again and again(and because of this Google would Think--"Why does this Page always get lots of love just for this single Keyword?"). That's How,indeed extra wastage of time and efforts due to over optimization. Thank God, i haven't done this kind of effort instead I am following Rand's Key Tip i.e. Always try to do Branding of your company with multiple variety keywords and always think about the long tail keywords. Hence I always keep on thinking about long tail keywords and branding while trying to aquire links. Thanks Geoff. Thumb Up to you!
Ajay - You make a good point, of all the posts, this is the only comment that is constructive.
Don't forget the importance or synonyms. They are pretty valuable and somewhat overlooked.
The part in your analysis that "almost" completely negates the results is this:
"Finally, I looked at the sources of the links and did not include pages if it looked like they had links manually built."
I wonder if the results would be different if you included the links that "looked" like they were manually built. Other than that, excellent method for analyzing link profiles. It's very very very important to consider all of the links though, not just the ones you "think" are working.
Geoff,
thanks for your post and the good definition of different keyword categories.
As you mentioned your results are nowhere as representative as they might seem, especially given that you only sample 10 web sites.
What is missing for me is also the information in which keyword space/niche, language and country you were conducting your tests.
We have been doing similar analysis for thousands of keywords (and their top 10 rankings) every day with your Competitive Landscape Analyzer (CLA) as part of the Link Research Tools in the last year and I can tell you - whatever you generalize here in terms on numbers is just wrong.
Brand bs. money keyword rules differ not only by topic but also for different keywords in the same space.
FYI - for keyword categories we usually only differentiate between branded keywords, money keywords (exact/phrase) and compound keywords (your "branded" and "KW branded")
I know it's not simple to present results based on research with very few data, but the best to do is point out to which samples it applies.
I did give some examples in my presentation at A4U on brand keywords last year also that you can read/listen to here https://www.cemper.com/seo-knowhow/brand-links-twitter-tweets-facebook-likes-power-your-seo
Apart from this factual problem I do like the presentation and definition of different keyword categories alot.
If you would like to give our Competitive Landscape Analyzer a try, just let me know and I'd be happy to get you a review account and share more details/background with you!
best regards, Christoph
I think exact match is quite overdone, but I also think what you have defined as Phrase and Keyword Branded are both excellent and far less spammy alternatives.
I like that you pointed out that people need to sample their own industries. If you're building tons of exact match links and noone else in your market is doing that, you're going to stick out like a sore thumb, when you could just as easily get away with less aggressive techniques and be just as successful. Using the right tool for the right job is key.
Also, I can't think of any situations where having more exact match links than exact brand links looks natural. If anyone can think of an example, please share it.
I tend to err more on the side of caution here as well. If anything, Google will get more strict on this topic and the ratio will be less..
nice post although I believe to my experiance anchor text match is all about the "source". If you get exatch match from trusted sources then you are good to get as many as you want but if not then mix it up.
Yeah I have suspected this as well. I would love to see a study on the anchor text distribution of posts that have gone viral - which means they usually have hundreds of links from good sources with anchor text that is all the same as the title of whatever the thing is, but they also probably have resulted in many non-targeted links like the ones discussed in this post.
Hey Geoff, I always love a data driven post!
We always deconstruct the serp - as there are often very specific (and sometimes unique) characteristics that may influence the top ranked search results. I agree, generally with your sentiment - as I've seen sites rankings improve as their anchor text distribution has changed from "competitive" to "branded" or "brand + mixed" anchor terms.
I also think "brand" anchor text is a safe signal - so it makes sense that, if you're not sure, always go with branded and see where that takes you.
Richard
I agree with the research, it is amazing the number of sites who do well who push an index of mainly brand related terms.
But yes diversification of your link profile and making it nautral is key if you only build links to say 5 urls and only use 3 different terms over a period of time you are heading for trouble if it is a large site, yet if it is a very small site it may work well for you.
You really need to sit down and map every thing out and play smart with link building it isnt something you delve into quickly ;)
I haven't seen very many sites compete with a brand only link approach. I did hear Greg Boser say (about a year ago) that if he had the choice to take an exact match or a brand link from the homepage of a PR 9-10 site he would go for the brand link. But, looking at the spread from above, 25-46% of links relate to keyword and not brand on category/product pages.
In my expirience on sites with local intent, exact and phrase links are still the major driving factor. The bigger the site gets, attention to brand links becomes more important but remembering that without exact/broad keyword matches...nothing happens.
Exactly, well put James :-)
There's that saying, 'Variety is the spice of life'; it holds true when it comes to Search. Having a varied (yet still relevant and value-adding) content strategy with the same for seeding and distribution, will likely lead onto a varied link profile (in theory).
I think mapping everything out is key to developing everything out.
Thanks for presenting the research. It is definitely something to watch out for. I think a good rule of thumb in SEO is to have a good variety of everything (when it makes sense).
I totally agree, it's like much in life, 'Everything in moderation'.
Of course though, everything ethical, honest and truthful especially when it comes to Search :-)
I'll keep the 7:3 in mind as a rule of thumb. Thanks for sharing, it keeps the comunity awake!
I would tend to agree with you that exact anchor text optimization is old fashioned now. With time search engines have become smarter and they can easily figure out such an inorganic campaigns. Recently, Cyrus Shepherd published a story similar to this one on seomoz. For reference and more info on the topic see this https://www.seomoz.org/blog/beyond-exact-match-anchor-text-to-next-generation-link-signals-whiteboard-friday
Next generator of Search Engines is trying to become more human. They are catching social signals, link signals etc and trying to react like normal humans so anything that strikes them whether links coming from same anchor texts they are going to find it out and who knows they secretly penalize it!
Geoff, could you please restrict people from editing the spreadseet to force them to make a copy of it. I am pretty sure it's been edited and is now quite hard to use!
Thanks!
Sorry about that, should be good now
If you target the long-tail - as any SEO worth their salt should - then your anchor text will be highly varied by default.
Great point. I combine my target keywords with city names in my area. Makes for a large variety of anchor text.
Actually, I doubt it would in many cases. If targeting long tail, you're also having long tail landing pages which are designed to work for 1-3 phrases most likely. And you'd use variations of those phrases for your built links.
Currently, I think it's well worth dropping non-targeted anchor texts into links for those pages.
These valuations are done on the URL level first, then on the domain level - meaning that if you have only an isolated problem, that page will vanish from rankings. Too many pages like that, and you'll get pecked on the homepage / domain in general.
It drives me nuts when people don't take the time to design charts beyond the Excel 2007 defaults.
thankx Geoff, saved me some time to work out my own ideas on this topic.
Really liked your keyword category definitions. And your google doc is very appreciated!
When I look at the old SEO work which is based on "exact" links anchors and compare it with effects of divide anchors for lots of combinations I see that the dividing anchors make much better work for positions.
My exact phrase versus other is 8:1, I rank pretty well but I haven't gon up in months so I will try cutting out on the exact phrase a bit. Hopefully that increases my rankings even more.
Would you treat your homepage like the Category page?
What if you have already overdone it and seen a drop in SERP. What is recommended to fix the problem?
I think variety is a must as far as keywords are concerned but relevancy should not and cannot be ignored either.Well done research but 10 websites samples might be a bit less. But still you did provide some new ideas and food for thought here. Good work Geoff and thanks for the Google spreadsheet on the Anchor text distribution.
When I read through this I start laughing (not about the author but google) and it absolutely clearly shows how useless and manipulative this whole google link stuff is. Considering this formulas here you never will achieve anything unless it is heavy manipulated, so what is it about? Unfortunately google shifted within the last 12 month to a increasing unnatural manipulating business philosophy this is not good at all. The second heavy manipulations is the weight to authority, the result is that since several month now I see increasingly pages with almost no content or totally unrelated to the theme plus only a few links and they are within the top 5, what do they have in common? they are from the big ones, like expedia, lonely planet, huffington post, nytimes etc. it is a shame how google discriminated webpages from smaller companies who have the real good and innovative content, elsewhere they call this cronyism. Before the new team took over this was a somehow honest business, it is not anymore.
hello,
can any one answer thit question
does it work if i made 88 % of non exact anchors in one page with nofollow tag
and 12 % exact anchors in 12 page dofollow?????????????
Hello Geoff, this is very informative post about the anchor text distributions.
I have one query, How to balancing out the over optimized anchor text?
I tray to make it 70% dof and 30% nof
The same anchore for many pages and the result is very good in the top of google
As long as you're not actively hyperlinking your main keyword, you really should be all right in terms of over optimization to be honest.
Hi Geoff! I totally believe that changing the anchor text is very important and will continue to become even more important as the social sharing of links is loaded for rankings. Thank you very much for sharing this and I really do appreciate this kind of topic.
Thanks alot, i advertise on other sites just by one word for a year and i havent any rank for that keyword, many thanks for your article
Wow. What a great article!. You can't imagine how much it helped me!
I finished analizing the qualitiy of my link profile, based on your study and see that, not only I have a Targeted 71.16% vs Non Targeted 28.84% distribution, but the Exact Match anchor text represents the 66.51% of the targeted links, having a Variance from Normal of 48.46%.
The question is: should I reduce the targeted anchor text percentage by removing some targeted links or better add new links for non targeted anchor texts?
Thank you and again I love this study and the spreadsheet!
Thanks Geoff (and the team collecting the data of course).
Excellent stuff as per usual.
Google love natural linking that's why too much anchor text usage is not recommended. That's why there are ways to diversify the use of anchor text in linkbuilding.
Read more here
https://searchenginewatch.com/article/2171174/Google-Penguin-Update-5-Types-of-Link-Issues-Harming-Some-Affected-Websites
Good info. I this is definitly a good guide. I dont think you can go far wrong if you go by this.
Are you planning on carrying out the same tests on home pages?
Also, what about static sites?
I think this is a great article and appreciate the fact were discussing the breakdown of anchor text of links naturally for a website. Now I bring up the next question, what should the link type distrubution be or look like rather in some of these large conglomorates to get a whole picture. For example what % are blog links, forum links, QnA, Social bookmarking, press release, social media. etc? Any insight into the link type breakdown distribution?
Thanks,
Thanks for taking the time to make these charts and share your findings on anchor text distribution. Good info and thanks for the Google Doc, very useful reference.
Nice post! At the very least, this is a great example of how to break down and analyze your anchor text to make sure you're not in the 9:1 range. Also, OSE doesn't count image alt text as anchor text, while some other software does. Although not as strong as text links, I wonder if it would be valuable to include them into such an analysis.
It would be good to include alt text associated with images as they do impact the link profile. I'm not sure if they show up in the anchor text report - I'd assume so as they show up in the inbound links report.
Thanks for the great post Geoff, it really opened my eyes to some new exiting ideas. I think that if you can get quality links and get your keywords in there, you won’t have any problems with ranking higher. I like the 7:3 ratio idea as well, 30% of your anchor should be targeted, sounds about right and I shall stick to that and see what happens. Thanks again, very cool post.
Geoff,
+1 for the Natural Link Building! I agree that varying the anchor text is very important and will continue to become even more important as the social sharing of links is weighted for rankings.
Interested to know how artificially keyword rich anchor text might damage the chance of referral traffic.
Even if you do have the chance to suggest anchor text when link building, I don't think you can/should decide what your anchor text without thinking about the people reading the page their goals, and the context of your link.
Should your link anchor text contain a compelling call to action rather than a keyword rich phrase?
I don't there are absolute rules for the anchor text you should use as every site is different and has different strengths/weaknesses - I'm sure there's value in using CTA's over keywords in some circumstances (especially if there is a lot of opportunity for referal traffic as you pointed out) but it really depends on the situation.
Thanks for the research - and I may well do exactly what you have said in terms of link text analysis. I think it comes down to the age-old 'natural' character of your link profile - a page that only has (or mostly has) exact match just doesn't seem natural (except maybe in terms of brand names) - people left to their own devices will link however they feel like which will always include a spread - retaining this natural profile is less likely to get you in trouble in terms of ranking and potential penalisation.
I will look at the ratios for some clients (and our own business in this part of the UK) and see what ratios apply, which at least can give an indication of 'natural'.
Make a lot of sense to go towards a more natural look of what anchor text will look like. Not everyone (or anyone ) naturally is going to give you an anchor text with your keyword. Thanks for posting and making my Monday a lot more interesting=)
Hi Geoff,
Conducting the test on 10 sites is not statstically significant, as the margin of error would be massive hence the data inaccurate.
Overall I really liked the methodology you took to pull your findings.
Regards,
Vahe
I think you have made a very good point that your link profile needs to look natural (as with many other aspects of SEO looking natural is key), and should fit in with the pattern that other sites within your industry follow. I also think that in a lot of situations the actual URL can contain the exact brand match, so having a larger number of these links can look natural, as many website giving links are more likely to cite the site url in a link as a source. I have often found this to be the case anyway.
We only use 'partial match' and branded KW anchor text, it seems to work very well, don't think we've built an 'exact match' link for over a year! This is of course a general opinion, you'll find that different industries require different backlink strategies, check what anchor text your top ranking competitors are using!
Very well done. Thank you!I've been thinking about the issue of over optimization anchor text for a while now. So this makes total sense.
It belleive its going to help a ton! Thank you again!
Jeff
Thank you for the google doc sheet! Its nice to see a clear analysis approach like this even if the sample of websites is not always relevant for everyone.
This actually has just want I was looking for. Seomoz and the great blogs by users has taken me so far in the past few months. Now I read everything and anything. SEO puts a smile on my face daily. Thanks for everyones hard effort to make this such a great informative community. Special thanks to the users and great seomoz starff and writters whom have made this such a succesful learning tool allowing them to quickly advance in seo knowledge and best practices!!!!!
Great post, I agree with the variety and natural linking seem to work best at the moment. Love the spreadsheet, thanks for the simple charts and the share.
Geoff - solid post. Love the simple charts that are easy to digest. It's a balance we deal with daily and seeing some solid research is always good to have in the back of the mind.
If you overdone it, you'll definitely get a -50 penatation. There is no doubt of it.
Great topic, post, and info/links, Geoff- thanks. This is useful to show a lot of SMB owners who are 'cukoo' for particular terms. From a readers perspective, overuse of anchor text may cause 'blindness,' making readers gloss over linked terms assuming they're all leading to 'expected' in-site sources (service pages, home pages, etc) rather than be an usher to other helpful information (articles, infographs, videos, etc). In most cases, varying anchor texts help better intrigue readers and enriches the browser experience. (first graph box - change "distirbution" to "distribution" unless the pun was intended haha - former English teacher in me!) Again, great post and thanks, Geoff!
This makes total sense. SEs will definitely like to see a natural backlink profile.
For those of you who still use BuiildMyRank, I'm sure you are using exact match targeted keyword anchor text. I guess it's time to start using more "click here" and "for more information" links.
Jeoff, the result of your research has confirmed the importance of anchor text in backlinks. The only “dark hole” (please, do not kill me :) ) is, as you call it, “random anchor text”. It is a good idea to fill this “random anchor text” by semantically relevant words (not by keywords). Why is it a good idea? Well, we all know Google’s patent about Semantic Closeness as a part of search algorithm. At least, it can be filled by synonyms as benjaminkarl pointed this out (see above comments). Thank you.
Hey Geoff, completely agree on the anchor text ratio but what about relevancy of content surrounding the anchor text to the source URL? Does it have ny bearing on the rankings?
I have run several experiments on a number of blogs where the content surrounding the anchor text is completely unrelated to my source URL and I still manage to get top rankings. For eg a health article linking to a SEO article with a combination of exact and random anchor text.
Would love to hear your thoughts.