By Olivia Isaacs and Dixon Jones.
Knowing what a link is… that’s SEO 101 right? Think again! It’s one of the most fundamental parts of SEO, but when we really asked ourselves the question, we found there were many different ways of counting links. We use both Open Site Explorer and Majestic SEO and decided to see if we can get a common census between the two tools. (Full disclosure: In case you didn’t know, one of the authors of this post is the Marketing director at Majestic.)
Firstly, let’s get all the confusion out the way about the different terms that are used to describe this one key word. "Backlinks", "inbound links" and "in-links" and "links" are different words but they have the same meaning. Namely, they are incoming links to a webpage or the entire website.
Open Site Explorer tends to have used the word “links” – but this can be ambiguous as it might also apply to outbound links from a page. (It doesn’t, in OSE’s case, but user perception is reality.) By contrast, Majestic SEO tends to use the phrase "backlinks", but again the user might reasonably ask whether a backlink is still a backlink after it has been deleted (since Majestic keeps the row of data, but just flags it as a delete link) or if a link from within the site should be included (as Majestic drops these).
Both SEOmoz and Majestic SEO realize that terminology is important and both tools have agreed to work towards a common language moving forward where we can. Certainly – when you use both tools, you will invariably get different answers as to “how many links” there are to a site or page. There are loads of reasons – but even if there weren’t, getting a methodology just to count them is pretty hard. So let’s look at an example (bring on the infographic. Full size here)…
Problem 1: Handling the two (blue) links from Page A to Page B:
Have you ever seen a “Link count” between page A and Page B that is more than one? I haven’t. So how does one account for two links between page A and Page B with different anchor texts? They are obviously not the SAME link, but saying they are TWO links would be simply inviting trouble. Even the lowliest of hackers can create 1000 links on a free blog page all linking to a target page and call them 1000 different links in this case. We are pretty confident that Google only take the first link and anchor text into consideration at this time – if that’s any help to you.
Problem 2: Does an Internal link count as a link?
Look at Page A. It has three inbound links and three outbound links. So is that three links or six? OK – 80% of you will say “3” even though technically the 20% would be right. Let’s take it a step further… what if Page A links to itself? Oh GOD! Here I think we have a difference between OSE and Majestic.
Majestic currently drops any links from internal pages (sites within the same domain), so that last example would not apply… but it is not true to say internal link don’t have link value. It is just a lot of extra data that Majestic doesn’t store as there are tools available for free that let you analyse internal links. OSE, by contrast, did seem to be able to keep a track of internal links last time I checked.
Problem 3: 301 redirects
You will notice that there is a third way a user can get from Page A to Page B with one click – via a 301. So are there two links from Page A to Page B or three? If you say “three” then you are also saying by the same logic that there is a link from page I to page B, which passes through the same redirected 301 page. Now that’s a problem. Is there a link from the 301 page to Page B? Or do would you say that the 301 page should be transparent?
Well I can tell you that Majestic (and I believe also OSE) counts the link from the 301 page as one link. Both sites mark it as a redirect link, but neither one will add in the links from pages H and I in their link counts. Technically – Google does though! At least sometimes. That’s why so many people buy up expired domains and 301 redirect them to other pages. They have a chance of getting link juice as a result. In recent years we have seen a bit of a downgrade in the effectiveness of 301s –but they remain a vital part of the web infrastructure. OSE shows you that the link is a 301, which should give you a good clue. In MJSE the link is flagged as a redirect, but you would need to pick up your “clue” from their ACRank on that link and if you wanted to investigate further, you would need to actually analyse the url that returns a 301.
Problem 4: Other Redirects
If either OSE or MJSE were to try and “act like a search engine” in making your decision that 301s should be effectively invisible, the problems simply multiply. Spam links would appear to increase, for a start, but what do you do with 302 redirects? In theory, Google does NOT treat these as invisible – although some .NET sites use 302 redirects in their menu structures, so goodness knows how Google handles that bad piece of programming. What happens on a multiple redirect? A 301 onto a 302 onto a URL which 301s onto the landing page and – itself – has 100 inbound links? No – it all gets a headache – so a link from a 302 or metaredirect page simply HAS to be a single link, even if there are hundreds of links going into the redirection URL. So how do the different systems actually keep tally?
- Majestic SEO says there are 4 links to page B. One from page A (oops) and one from each of the three orange redirect URLs.
- A Search engine obeying the official line on handling redirects would probably count 5 – but they would be very different links. One from Page A, one from each of pages H and I, one from the MetaRedirect page and one from the 302 redirect page.
How many does Open Site Explorer count? I don’t want to speak for OSE but hopefully they will be able to say.
Note from SEOmoz - At this time, Open Site Explorer would count 2 links from the diagram above, one from Page A and one from the 301 redirect. Although you may see both H and I in an Open Site Explorer link report, redirects do not share or include their totals with the targets of the redirect. We continue to evolve our measurement and reporting as our knowledge search engines continues to expand.
In Summary
Counting backlinks is not a straightforward logic. Whatever logic you choose, Google sometimes counts 301s and sometimes doesn’t. It may be that Google takes the Juice but not the context of a redirect. So if you are building a technology like OSE or MJSE we need to either give you data OR give you opinions.
Thanks for a great topic that is often neglected or glossed over.
Another way to count inbound links to Page B is to look at how PageRank flows (or doesn't). Using this approach and assuming as you do that Google ignores all but the first link from a page, the inbound link count is 3. One from Page A and two through the 301 redirect page from pages H and I. Pages G and C don't count since neither a meta refresh nor a 302 passes PageRank. Since PageRank is a significant part of why we want inbould links, it might make sense to consider it when we are devising methods of counting links.
Nice to see the cross-industry post, and always interesting to dig into just how complicated what sounds like a simple topic can be. We get so many Q&A questions that boil down to why a link is or isn't counting and what makes one link profile stronger than another. It sounds simple enough on the surface, but it's extremely complicated when you dig into a site in the real world.
What a powerful amazing post!I learnt a lot by reading it and havn't been aware of the diversity of counting possibilities.
Perhaps somebody of SEOmoz can explain why they count links from the 302 redirect of C?
Sharp eyes! We actually meant the 301 redirect. I've corrected the post to reflect this. In Open Site Explorer, I believe you can see all links that redirect through a 301, but Linkscape only counts 1 link in its total. Does that make sense? In other words, 301s don't pass the total of their links to its target.
The question is a good one. How many links should you count? Do 301s = canoncical? Are redirected links treated equally? Link counts are relevant when measuring link growth, but do they carry significance beyond that? I'd love to hear everyone thoughts!
In my experience 301s don't pass as much juice as canonical links, but how that may affect an actual 'count' of links I'm not sure. - Jenni
With my past experience 301 give all the benifit regarding link juice or counts, it will never make any bad imapct on to your site
This would make a great exercise for a job interview =)
That's what I thought I mean if you really want to mess with someones head and see how much they really know it could work.
How about a link coming from a Cannonicalized page? will it count as many links or only 1 link? (say there are 100 pages that has the same content and the same rel cannonical, will we get 100 link count or only 1 from the source)
A way to interesting post!
I think this is what makes one tool different from other and this important to have different results to get the maximum data by using multiple softwares.
Imagine big brands like Majestic and SEOmoz sit together (Although competitors but I still think it’s very easy for them to sit together) and set 1 straight forward way of counting links… the software might be started to give more or less the same results but there will be no more charm and what will be the difference and why would I gonna use SEOmoz or Majestic or any other tool if all are going to produce the same result?
I my Opinion this is good that different software are producing different results as this helps us collecting the maximum amount of data by using multiple softwares.
Great post, thanks. More interesting to me than the counted number is to get aware of all the backlinks. So i prefer getting backlinks that my not be counted instead of not seeing them at all.
Ah! And I that I was thinking that a link was just that... a link.
Jokes apart, i really enjoyed this post that made me discovered how something I was not really aware of.
Personally, when I check the link profile of a site I look more than just the raw numbers of links, putting my interests more in the number of the unique domain names linking. In this sense all the explications given about the ambiguity about how to count a 301/302 as a link is extremely interesting.
Thanks for the enlighting post and I wish more cross industry post (now I'm waiting a post by Rand on the MajesticSEO blog :))
Wow. I'd never thought of the complexities of counting or not counting a 302's page, or just exactly what the total number of "inbound links" entails. It's too easy to see the total number of backlinks and gloss over all of the nitty-gritty details of what's invovled in that number. Great to see someone taking a "SEO 101" topic and showing it's far more complicated than it first seems.
Fantastic post!
This. Post. Delivers. Not only that but its really nice to see majestic and SEOmoz playing nice together in the sandbox. Keep these sort of posts coming guys :-)
haha! Post. Deliver... brilliant.
I'll get my coat.
What a fantastic post! Love the graphic too. That just might be posterized in my cube! Thanks!
Excellent post, use a variety of online tools, SEO software as well as data provided by search engines themselves and differences in number of links picked up can be extraordinary!
However, can be fun and interesting as an SEO analysing the data, measuring effect, etc, however from a clients perspective would push them towards looking at rank, visits, enquiries and conversions as a measure of progress as opposed to number of links.
If client wants this info fine, send them a list of links submitted and then 1 - 4 sources of links picked up, however once project is established data should become irrelevant from a clients point of view.
Would say links data from client and competitors sites are most useful and important to be accurate when quoting for projects initially.
I found this line interesting, "That’s why so many people buy up expired domains and 301 redirect them to other pages. They have a chance of getting link juice as a result."
I definitely consider that cheating. Would be nice if Google found a way to discourage it.
Only just got to this post - maybe should have left it until this afternoon when the brain was more awake - however, this is a great demonstration as to why 'linking' is not always a simple task - thanks for this, it is really helpful.
This is an awesome post! I love it how you guys simplify things up! great infographic.
Really an awesome post and beautiful infographic.
301 redirects are really the most annoying part when checking backlinks for a single webpage, or for an entire website.
Wow, brain melt. Seriously though, as steviephil points out it's educating clients about links that's tough. It's hard enough explaining to some that more links doesn't = awesome SEO. So, explaining that Google counts the 10,000 links they bought for £100 from a spammer as just 500 links? Tricky.
Thanks for the insightful post though, the more we understand as SEO professionals, the better equipped we'll be to explain the issues to our clients.
This is awesome post for people like me :)
I am not from this industry, but try to look for some good tips/tricks that help my online business.
This is really useful, kep posting blogs guy!
Thanks for sharing this. It will help to filter our concepts and help us to perform better. Keep it up
https://itechbee.blogspot.com/2011/08/square-now-processing-4-million-in.html
Great post I've learned a lot
Counting backlinks that complicated and tiring job. But what may make should we do to all that we do see any results, or even for me who do not really like counting backlinks simply by looking on google alone analityc
Thank for sharing this. It will help to filter our concepts and help us to perform better. Keep it up
That diagram frightened me. But afetr reading your post now I am feeling relax. Thanks for such an interesting post.
I think I'm even more confused. Haha - Oh the logic behind backlinks. I'll just keep building links and let the SEs sort this out. :)
This post was overdue as I have been wondering the differences between the two for months. Better late than never I suppose. Would be great to see these type of partnership posts more often. They provide a lot of value and answer a lot of questions
I personally feel very comfortable about the links I see in Google Webmaster Tools. Seeing them there makes me feel like they are counted, valid, and worth something. I'm aware of the other links I get (using mainly OSE or YSE), but if they never show up in GWT, I'm always left wondering how much value they really provide me, if any at all.
Sure - seeing the links in Google WMT is helpful - but these are only links for your own, validated sites. The data on the other sites above you in the search engines is very poor indeed on Google, because the link: command bears no relation to the WMT data. Agin, I can't speak for OSE, but Majestic would hope that we have a significant overlap between the WMT data and that Majestic data.
Makes me want to try and redirect all pages in a infinate loop and see what happens!
No Offense to anyone, but doesn't Majestic SEO have a bigger repository and indexed pages, and therefore return more links than OSE?
I think the fact I have had to read this post twice demonstrates the complexities of something which sounds as simple as counting links!
me too, i should have read this BEFORE spending an hour on the bbc news site.
Maybe this has been changed but I've seen OSE count 301s as two links in some cases.
I can confirm this, i have also seen this happen.
"Even the lowliest of hackers can create 1000 links on a free blog page all linking to a target page and call them 1000 different links."
That is exactly why quality is more important than quantity! That kind of spammy link building may look good on a piece of paper, but what kind of value are you actually getting for your site? The search engines know when they are being played, and they don't appreciate it.
Very well constructed post albeit a little inconclusive. Google keep the very best of us guessing.. hence their dominance.
I must state... I believe that all links, regardless of their positioning on a trail, are counted in their own manner to some degree. If a 301 re-direct included seven high-authority links pointing towards the original page/site, compared to a re-direct that had 100 low quality links pointing towards it, then the former option would provide much juice to the page/site that the re-direct points to.
Do we have an up-to-date estimate on how far behind OSE is in recording links?
Keep up the good work..!
I agree with the points made, I also agree that counting backlinks is a hard topic especially on websites which have very complex link profiles with huge amounts of re directs, various old sites and many more factors.
Hence why I try to include multiple variations in my reporting in this area.
But overall great post and good graphic =) If you have been in link building for a long time I am sure this is all second nature.
That's a busy graphic. But I'll get it.
Great post and it is good to understand what you guys think of the whole linking process.
Personally I believe 302's should not be classed because they are temporary redirects (To me), highlighting the original page will be back at some point.
I'd like to know how SEOmoz's logarithmic 10-point scale would work on the provided chart and how that would affect most importantly 301's and 302's ranking/domainjuice when passed.
Just for example: Page C has 20 Domains linking (each with PA of 50 and DA of 40)
What would be passed (Approx) to 302 Page and then on to Page B.
Likewise with the 301 Page
Pages A, H & I each have 20 Domains linking (each with PA of 50 and DA of 40)
What would be the juice passed (approx) and is link via Page H through the permanent 301 to Page B now classed as a reciprocal link?
Cheers,
Joseph Gourvenec
T: @gbnetworks
L: linkedin.com/in/gbnetworks
Great post and infographic. I like posts like these as they reaffirm or clear up the myths and common misconceptions.
One of the biggest issues I've had with 'link counting' is trying to educate the client/boss/etc. about some of the above issues, as well as:
If anyone has any good tips, analogies, case studies, etc. of how to educate those who don't know (or don't want to know) much about link building then I'm all ears.
Hi Steviephil - You're right, explaining this to clients is difficult.
In my reports, I give my clients a list of 'live' links and 'pending' links, so they know what I have built over the month. As well as this, I am constantly explaining to them (ad nauseum) in my email correspondance and during meetings that links take time to become effective, so even if I have built 100 links in a month for them, then not all of these will be 'working' for them at once... if that makes sense?
It is confusing, but at least it's truthful! I have taken on clients who have sent me through reports from previous SEO providers - and, well, the lack of 'proper' reporting is enough to break your heart - lists of links from all sorts of dodgy places, or, worse - no link information at all. Having said that though - I have refined my reporting over the years to include some stuff that other companies have unwittingly given me over the years.
Regarding quality vs. quantity - when I take on a new client I explain to them the differences between these links, giving them an explanation on why I go after quality links. One of the best ways of convincing clients about quality links, is to explain that a link built for the search engines alone (e.g spammy link farn rubbish) will never bring them traffic by itself (as in people clicking the link) but a quality link that has been built for people to use, is likely to bring referral traffic by itself (let alone the benefit given to listings).This is usually a good enough explanation for them to get the gist - and most of them are happy if they get results!
Regarding Follow, and Nofollow - if they ask about it, then they usually have a pretty good idea already - I just ask them to tell me what they think the difference is, and if they are right, I tell them so... if not, then I can build on what they already know, explaining that although 'nofollow' sounds negative, I don't consider it to be, so long as the ratio of nofollow to follow is healthy and natural then it's OK.
Hope this helps! It can be a minefield out there trying to explain SEO to clients, I used to think 'Why did they hire me, if they want to know it all for themselves' but actually, I've found that a healthy relationship with clients is built up through communication - some want more than others!
Yes I think that all links, regardless of their positioning on a trail, they counts thieir own ways but If a 301 re-direct included seven high-authority links pointing towards the original page/site, compared to a re-direct that had 100 low quality links pointing towards it, and would provide more juice to the page/site where redirect points to.
However, I love and a good fan of this and like to count my sites links with this.
[links removed]
Thanks for the positive responses. I think it's the start of a number of discussions on how we "explain" stuff. Who would have thought that counting a link was so difficult? But there you go.
Kip said: I believe that all links, regardless of their positioning on a trail, are counted in their own manner to some degree.
I am sure that the engines are currently counting many links via some redirects, but equally there are many ways that links can exist that the engines either can't or won't count. I can think of many ways this would haooen - from Geotargeting meaning googlebot never sees valid link to robots.txt blocking a valid link and many things in between.
Dixon.
I'm sure he meant all 'accessible' links
:)
Strangely after running multiple tools on a single website today I was actually thinking about this, but at the end of the day does anyone really know which ones the Search Engines are counting? It's not a secret that they like to keep there secrets especially Google, but none the less it is interesting to see I guess you really have to decide what to count and what not to count in links.
Love the chart and explanations, good info here on inbound linking. I'm not a huge fan of 301 redirects but use them as needed in specific situations. This is another reason why integrating intelligent SEO practices in link structure and content management is key at an early stage, this will really set you up for success or failure in the next year.
I've only been working in the SEO field for about 5 months now and just recently started learning/working on linking... the terminology and vast differences between some tools have been driving me crazy. I'm glad to hear that I'm not the only one who doesn't know what's going on with links and that even veteran SEOs don't necessarily agree on the subject. :)
Nice post! Nice Infographic too. I guess at the end of the day there really needs to be a lot more standardization of terms and formulas where and when applicable. I like the statement above, "Both SEOmoz and Majestic SEO realize that terminology is important and both tools have agreed to work towards a common language moving forward where we can."
It would really be great to see a lot more of this standardization of terms and formulas in the Internet Marketing arena.
Cap
Great post. Thanks for starting this conversation. Often I'll look at OSE and see links that are no longer on the page, or must have been only an ad.
thanks.
How much impact do you think buying an expired domain has when 301 it to another page? Is it worth the cost?
It depends on the domain and the cost! Google has known about this as an SEO "trick" for some time and I felt their was a change in their approach to 301'd domains a few years ago - but domains 301'd BEFORE the change in the algo seemed (when I looked bak then) to be treated differently to domains redirected after the change. Now - knowing how hard it is to revisit everything regularly at scale, I can see why this may be. What it does say, though, is that again - it is not always up to Majestic SEO or OSE to tell you if a link (redirect) carries juice. In the end you wll have to decide for yourself.
Great post with a very informative infographic!
Thanks for giving such a valuable information on how to count backlink of a website.But i think we can also check no. of backlink with the help of
link:http//url.
Not really. Compare the quality of that response on Google to the dedicated tools and you will soon stop using that command!
Graphically a good infographic, but a little busy!