How You Should Start Choosing Anchor Text Starting NOW

How You Should Start Choosing Anchor Text. Starting NOW

Hands down, the most important skill you can have when it comes to offsite SEO is anchor text selection.  Having the correct anchor text distribution can make you (or break you) in the SERPs.  If you do it right, you’ll find yourself at the top of page 1, and you’ll do it with less links than your competitors.  If you do it wrong, say hello to Penguin penalty.

Not surprisingly, the most common question I get from customers is: What anchor text should I use?   While I’d love to be able to answer this for everyone, but there simply isn’t a 5 minute answer.  Selecting the right anchor text takes analysis and practice, and this is a skill that you want to craft for yourself, rather than rely on others for.  But let me just say, once you get this down, it’s like graduating to the next level of SEO.

The first step in anchor text selection is to figure out where you’re headed by determining what is the target anchor text distribution for your niche.  Here’s how I do it, broken down step by step:

Analyze the Competition

Many SEO’s often talk about needing a “natural” anchor text distribution when they describe it, it typically looks something like this:

  • 50% Brand/URL anchors
    • For example, a branded anchor for would be “Golf Genius”. Branded anchors only apply to non-EMD/PMD sites.
    • URL anchors are variations of the naked URL:,,, etc
  • 25% topic anchors
    • Example: “sports equipment” would be a topic anchor for the keyword “best titanium driver”
  • 10% keyword and longtail anchors
    • Examples: best titanium driver, what is the top titanium driver, buy titanium driver online
  • 15% miscellaneous anchors
    • click here, read more, go to website, etc.

While many people stick to this idea that you need this perfect, “natural” ratio of various types of anchors, this simply isn’t the case.

Protip: Whenever you do a search for a specific keyword, Google shows you exactly what they like to see in the SERPs.  Do a search for your target keyword and toss the results into a backlink checker like Majestic or Ahrefs.  Here, and only here, is where you’ll find the anchor text distribution Google is looking for.

Often this looks nothing like the cookie-cutter anchor text distribution that is supposedly “natural”.

Case in point – Keyword: “sell my car online”

sell your car

Your anchor text plan often isn’t what you initially expect.  I’ve been in niches where I had to consecutively hit the site with target anchors for a month straight until I was able to break into page 1.  Sounds like SEO in 2005, but sometimes that’s what the niche is asking for.

Find the Average Anchor Text Distribution of the Top 5 Rankers

To figure out the target distribution of anchors for my particular niche, I rely on Majestic and Excel.

Step 1) Download the Majestic Anchor Text data for the site in position #1


Step 2) Categorize all of the anchors by type

I use the “Referring Domains” column so I don’t double-count anchors sent from the same domain, possibly from a site-wide link.


Step 3) Find the anchor text distribution for this URL by creating a pie chart.

cars dot com

Step 4) Repeat for Sites in Positions #2-5

Step 5) Your Niche-specific Target Anchor Text Distribution = The Average of the Top 5 Sites

final anchors

Next Step: Do It Better than the Competition

Now that you know the anchor text ratio that Google is looking for, it’s time to begin your link building campaign whose anchors will mimic this distribution.  The only difference is that you’re going to be doing it much better than the competition.

In the next article, I’ll go over 6 crafty anchor text techniques that will allow you to get much more value from your links and push you past your competitors’ amateur SEO efforts.

Like what you read?

Don’t miss out on next month’s post.

Stay Updated!

Get these articles sent straight to your inbox


  • Hi Matt!

    Wow, this post was really an eye opener for me… I’m guilty of following the “one size fix all” anchor text method. So thank you very much for opening my eyes to a new approach.

    Now, if you don’t mind me asking, I have two questions:

    1) I can totally see myself doing this approach for 1 or 2 keywords… probably is gonna take me 20-30 minutes to gather all the data (the anchor texts for the top 5 sites) for each keyword… but here’s the thing… what about if my site is targeting 40, 60 or even 100 keywords. That can really take A LOT of time if done manually. Is there a way to automate this process when working with several keywords?

    2) Let’s say I found a winner keyword. I gather all the anchor text data from my top 5 competitors and it turned out that they ALL have 50-60 BRAND anchor texts links. Now… since I’m gonna primary gonna be using PBN links… how can I match that 50-60 brand anchor text ratio?? (in order to mimic their anchor text pattern). Because you know… in our community we love the idea of outranking a site with 2,000 low quality backlinks, by only using 5,10,20 PBN links. But… I love your approach… how to make it work in this case?? (with mostly PBN links).

    Thank you very much Matt. Your blog is my #1 SEO resource nowadays and I’m so looking forward to learn more of your own personal approach to SEO.

    • Hi Jorge,

      Thanks for your kind comments. To answer your questions:

      1) Unfortunately, I haven’t found a way to automate this in any way other than having my VA do it. That being said, this is a great task for a VA and should be a research tool available to you at any time. So it’s well worth the money.

      2) Press releases and inner-page single payment PBN posts on high quality networks.

    • Depends on the niche and if your competition is other SEOs. If that’s the case, think outside the box. Find the few that screw up and forget to block ahrefs and use them. There’s always some.

  • Do you ever find yourself dialing back your exact match anchor a little after this analysis to account for sites that might eventually get hit with a penguin penalty?

    What happens if the Top 10 results are being dominated by sites that will be hit with a penalty and it’s only a matter of time?

    It’s okay to say I’m over thinking it. 😉

    • Hey Rob, great questions… really got me thinking.

      • 1) The only times I’ve been in a situation where I’m pushing a penguin penality is when I’m taking over a new client site. In which case, I’d definitely dial back and start making the anchors match the niche-specific ratios better. It takes a bit for the results to kick in, but with certainty, they eventually will.
      • 2) Really good question. Normally I’ll ignore results that have a unruly anchor text ratio that just doesn’t make sense -or- hidden links (like Lance pointed out).
  • Hello, Matt. Great blog. I’ve learned more reading your small handful of articles than I have reading hundreds of articles on other blogs.

    I have a question, if you don’t mind. When you use Majestic to gather up all of your competitors’ anchor text, do you use the fresh or historic database?


  • Hi Matt,

    Really enjoy your blog. I’ve decided to try out your average anchor text distribution method, but it seems my competitor’s anchor text distribution %’s are all over the place. For now, I’ll just ignore the data & go on with the typical “natural” anchor text profile. Is this method only suitable for keywords with higher competition (non-local)?

    • Hey Ivan, this method works for all niches, but you’ll definitely see more reliable numbers with a larger data set (high number of links).

  • Matt,

    How about when you are going after a niche that is full of authority sites in the first two pages and they all only have a handful of links each that lets them ranks.

    This is for a terms that is a physical product we sell.

    Would you just start out by modelling what you mentioned in ‘Analyze the Competition’ in the blog post?

    • If page 1 is jammed full of authority sites, then its likely to stay that way. Give it a shot using agressive anchors and a better optimized onpage, but be ready to call it quits if you’re not making gains.

  • Hey Matt, it’s amazing how sometimes the best content can be found on small corners around the internet, your blog is truly one of the best.
    Anyway, I was wondering if nofollow links are good for pillowing?
    Let’s say I want all my dofollow links to be targeted anchors, and only pillow with nofollow links, would it work?

    • I’m referring to all regular website links: PBNs, Comments, white hat links, etc. Social signals, bookmarks, etc are not included.

  • Nice post. I have actually had a site hit with a spam penalty before for the very reason of poor choice of anchor text. Still haven’t managed to fix it but I put it down to a learning experience!

  • Great post, thank you. If an anchor has e.g. 10 RD’s out of a total of for the site of 50, would that equal a 20% slice of the pie? i.e. Multiple RDs means you’re counting that anchor multiple times towards the total?


  • Hi Matt,

    Quick question regarding EMD, you said that branded anchors should not be used for emd’s (which is understandable do to over optimizing) and if you’re competitors are not using emd’s how do you know which ratios to use?

    Should it be mostly naked URL’s?

    Also, second questions I wanted to ask you is, if it is naked urls, do you think it’s possible to rank #1 using naked URL’s without any EMA or LSI (assuming you were going after the main keyword only)? Since the keyword is in the domain name and title of the website, in theory you shouldn’t need EMA since google already knows what your site is about, correct me if im wrong?

    • 1) If the competition doesn’t have any EMDs for comparison, use the branded domains to come up with your ratio, but make it significantly more conservative.
      2) Yes, I suppose its possible, but I’m not sure why someone would want to try to rank without a EMA, considering its likely the most powerful ranking signal there is.

  • Hi Matt, another great article!
    Got a quick question for you..
    After purchasing a branded press release or even building citations, not all links will show up in majestic. Is it best to assume all links have been counted when trying to match the average anchor text ratio of competitors? Or should I base my results on the same source (majestic etc) I pulled the original data from?

    • For competition, you should be checking Majestic, Ahrefs, and Moz and pulling all their data together. While simultaneuously making sure they’re not hiding any links. For yourself, count anything that is indexed by Google.

  • Hey Matt, I am freshing up my site and I had your article about the anchor text ratio in mind.

    Since it is my first site I messed up the anchor text ratio quite a bit. But somehow I managed to rank for some Kw´s on page 2 without a single link pointing to them, so I decided to put my site back on track.

    Since I need to clean up the target anchor ratio a lot, I contacted the domain owners and asked them to delete my links. Some of them did, some of them didn´t. As a result I need to clean it up more and I was wondering if I should disavow these links.

    Is that a good option or shall this leave the impression to big G that my site is not good enough for page 1.


  • Hi Matt!
    So say I’ve already done all this analysis. I already have my average number of anchors to use (say 50) and the the type of anchors (say 50% URL, 30% Brand, 20% Target). Three questions:
    1) What link velocity should I use? I mean… Should I build these links (in my example, 50) all within the first month? or should I spread them? How do you determine that.
    2) And, say after having built all these 50 links… What after that? It is recommended that I just wait and see the rankings increasing, withoug doing anything else from my part?
    3) Say 3 months have passed after I built the last of those 50 links (imitating the correct anchor text distribution for my keyword) and my page still isn’t on Page 1. Do you just continue waiting, or at what point you start building more links?
    Thank you!

    • 1) Depends on the age of the site and the amount of signals you have going to it. If you have some age and signals, then feel free it hit it with 4+ per week.
      2) Start linking to the silo pages that are linking to this page. In fact, you should be doing that all along.
      3) Continue to hit those silo pages.

  • Hey Matt,

    Do you actually build out type of links out too? I’m not sure what these are called but the N/A ones in the example you showed.

    Thank you!

  • Its an amazing article Matt! I couldn’t have come across this article at a better time as I am just about to start linking out. Well I have a complicated question for you. Suppose I want to rank for 5 (related) keywords for one article. Now for each keyword, I will get a different anchor text ratio according to the above guide. And for all those 5 keywords I will get different anchor text ratios and they might contradict each other. What to do then? Look forward to your suggestions. Thank You!!! I already have several pages of your blog bookmarked!

  • Hi Matt!!
    Couple of questions if you don’t mind my friend:
    1) Is there a special reason why you chose to analyze the Top5 and not just the Top3?
    2) So we of course want to use PBN links for the “Target” anchor texts, but can we use Web2.0 sites for ALL THE REST of the anchor texts? (Topic, URL, Brand, Misc, N/A)
    Say my average number of backlinks for my keyword (a product name) is 35. Only 5 of those links are “Target” type (so I’ll be using 5 PBN links)… Can I build 30 Web2.0 sites and use those for the rest of the backlink profile??
    I’m asking because I remember you once said that backlink diversity is a myth… So now I’m wondering if you can imitate the average backlink profile for your keyword with PBNs + Web2.0s only (allowing you to have 100% control of the entire backlink profile).
    Thank you!

    • Hi Jorge,

      1) The top 3 usually get a special leniency when it comes to their optimization, simply because they get traffic. Averaging the top 5 dilutes their advantage.

      2) I actually pillow with PBNs, citations, and sometimes press releases. I never use Web 2.0 these days. Diminishing returns on what they provide.

      • Thank you so much for clearing this up Matt (as well as answering all my other questions I left on different articles yesterday hehe)
        Now I have a better idea of what to use for these other types of links. So glad I asked this before I started building those 100 web2.0 for pillowing I was thinking of using 🙂 Cheers pal!

  • Hi Matt,

    Great stuff. Just one question. On your pie chart, what type of links are you classifying as “NA” and “Misc”? How are those two types different from each other? Thanks!

    • Misc would your typical “read more”, “click here”, and “website” type anchors. NA’s would be the anchors that simply dont fit anywhere else: blank anchors from empty alt tags, author names from blog comments, etc.

  • Hi Matt,

    how can I perform this deep analisys if all of my competitors are ranking with their internal pages and not with their main root domains, please?


    • Anchor text optimization is performed on a page-to-page basis. You don’t look at the domain. You look at the page. This is the same when it comes to your own site.

      • Hi Matt, thanks for reply. I did it and, with my bigger surprise, the top 15 ranking pages have no backlink pointing to them…

        My website is 19th since Dec.2015. I have a 2000 words related article, H1, H2, LSI keywords and theme words used, backlinks from wikipedia related post, from 8 Web 2.0 blogs including google sites, wordpress and tumblr, links from the main related online magazine…but I am still always on 2nd page.

        I really don’t know how to rank this keyword on first page…also considering that ALL my competitors have no backlinks and also very thin content (max 300 word post).

        Any suggestion will be highly appreciated.

        PS: To clarify, I am talking about an italian serp on

  • “…I’ll go over 6 crafty anchor text techniques that will allow you to get much more value from your links and push you past your competitors’ amateur SEO efforts.”

    I like your style 😉

  • Hi matt thanks for providing this excellent piece of info, I am doing this now for the first time and following this closely, however have 1 or 2 questions relating.
    After downloading the stats from majestic I feel that not all the anchors are provided (is that common?) after looking at the websites themselves there as some that are not added. Also Is there an easier way to distinguish a keyword anchor to a topic anchor as I get a little unable to make my mind up easy when looking as I suppose a topic anchor can be a keyword also? Thanks again

    • You definitely need to use all of the backlink crawlers. Ignore all sites that have hidden backlinks, or discover them by more clever means.

      A keyword anchor contains at least one word of your target keyword phrase. Topics dont.

  • Hi Matt!

    I was wondering… Does the order on which the anchor-texts are built actually matter??

    Say I’m targetting a keyword which I believe is going to need 15 links. After doing my anchor text analysis, I found that:

    5 need to be Target
    5 need to be URL
    5 need to be Misc

    Is it the same thing if I:

    A) Build the Target first and then the URL and Misc

    B) Build the URL and Misc first and then the Target ones

    c) Build all 15 totally random

    My thoughts are that maybe, as long as all the 15 anchor texts are built at the end, it doesn’t matter the order… But… I thought of asking the expert first before making any final conclusion 😀

    Thank you!

    • Order matters.

      Just think, Google is the company with the most PhD’s on the planet. Their spam team’s primary goal is to detect unnatural patterns and if these patterns trigger a threshold of improbability you’ll pay the consequences either algorithmically or manually.

      Is there any reason why you have to build the anchors in an unnatural sequence? Probably not, so do whats natural.

  • Thanks for this awesome post. I was looking for something like this and found at the right time.
    One question, when I put my site in Majestic, I was freaked hell lot of anchor text spam.

    Any idea, how to remove this anchor text spam?

  • Hi Matt!

    Thanks for the great post. Just a quick question – How would you recommend to use keywords for EMDs in post URLs?

    Say my EMD is and primary key word is “best dog toys”

    So what would you recommend for articles like “best dog toys for XXXX” and “best XXXX dog toys”


  • I like the idea there is only a problem if all competitors are hiding their backlinks.

    How can you know the right anchor text ratio if all websites in the TOP 5 are using hidden PBN links ?

    I have checked many backlink checkers and searched for links manually on Google already but it’s quite difficult to find PBN sites if they are hidden correctly and not using brand/url anchor texts.

    Any advice or tips ?

    Kind regards,


    • Hey Alex, if you have the means to uncover their PBNs then take them into the analysis. If not, ignore these sites and move on to the next.

  • This takes quite a while to do and it seems like I get the same results you did in the South African automotive niche. See here

    1) Is it really worth doing this per niche/keyword? Have you seen major variances?
    2) Does a 5% swing make a big difference? E.g. Let’s say the average target % is 30% and you have a target % of around 35%? Seems like a lot of work and curious to know if it makes a big difference?

    • 1) Absolutely. Numerous times I’ve had to increase/decrease target anchor density to levels that made no sense at all. The niche dictates what should be on page 1, not common sense.

      2) Not really.

  • Just revisiting this article again after seeing it in a related Facebook Group. Such a great post.

    I’m going through pages now that I’d like to move up from the bottom of page 1 to the top. I’m going through my competitor’s anchor text ratio and had a question:

    If an anchor text contains the brand AND the target, should that be considered a brand anchor or a target anchor?

    For Example:

    buy books at Barnes & Noble

    If the keyword phrase I’m trying to rank for is “buy books” should I consider this anchor text a Target anchor text or a Brand, since it has the text “Barnes & Noble” in it?

  • Great stuff Matt. Couple ? if you have time.

    When checking anchors for a URL, are you looking at the Path or URL results in Majestic? Or does it matter?

    Thanks bud,


      • Hi Matt,

        When you are doing this for affiliate sites, you analyze only other affiliate sites, right? Excluding authority sites, e-commerce sites etc?

        And what if there are only 3 affiliate sites on 1st page and rest are e-commerce? Would you pick rest 2 aff sites from page 2?

  • Hi Matt,

    1) I was wondering what kind of a links you would exclude from your anchor text ratio? I saw that you exclude social signals and bookmarks but what about all these auto generated backlinks like website lists, random analysis sites etc (all those useless links)? Would you exclude them as well?

    It makes sense to exclude them, but would love to hear your opinion!

    2) Also what’s the best way to find out if competitors are hiding their backlinks? Of course it’s easy if they have only few RDs to the root domain but they are ranking on top. But what if they hide only part of their PBN? How do you conclude that this guy is hiding his PBN?

    • 1) I’d count any website links that are indexed. Including press releases, directories, citations, etc.
      2) There’s always ways to reverse engineer an entire PBN. Maybe I’ll get into it in a future blog post.

  • Hi, Matt. Thanks for the post. I am curious about target anchor, Ex: my main keyword is: best vacuum cleaner, so “my vacuum cleaner” anchor is target anchor or topic anchor?

  • I don’t quite understand how this works for multiple keywords.

    Let’s say I analyse a page that’s ranking top for keywords ‘sofas’ and ‘coffee tables’

    I look up the anchor texts and fill in the pie chart for keyword ‘sofas’

    There are 100 backlinks.

    There are 50 Brand/URL anchors.

    there are 15 misc anchors.

    There are 11 keyword anchors referring to sofas.

    There are 9 keyword anchors referring to coffee tables.

    There are 10 topic anchors referring to sofas

    Therea are 5 topic anchors referring to coffee tables.

    Would I fill the anchor ratio chart in as:


    50 Brand/URL
    15 Misc
    11 Keyword
    10 Topic
    14 NA

    and Coffee Tables

    50 Brand/URL
    15 Misc
    9 Keyword
    5 Topic
    21 NA

    Or combine them and just have

    50 Brand/URL
    15 Misc
    20 Keyword
    15 Topic

    For both keywords.

    Hope that makes sense. Thanks for the guide!