Just how good are backlink tools such as Open Site Explorer, Majestic SEO and ahrefs at identifying links to the pages of your site? That’s the question we set out to answer in this study, and for many of you, the answer may come as a shock!
(Want a quick summary of the results? See our 3-minute video and 12-slide deck at the end of this post!)
Bottom line is that they all fail to report on many links, even from very prominent sites. However, this study is not meant to indict them. They remain incredibly valuable, and a core part of any SEO toolset.
In this post, I will present the results, as well as provide you with my thoughts on how this impacts your SEO strategy. At the end of the post you’ll find a video where we summarize the findings.
[Tweet “No backlink tool finds all links, but they are still useful. See study at”]
How We Structured the Test
We picked a group of 20 sites that share all of these characteristics:
- Strong tendency to link out to other web pages in their content
- Implemented links are NOT NoFollow
- Reasonable to a high degree of prominence
The 20 sites we tested were:
We then reviewed the outbound links in three articles per month on each of these sites, starting in June of 2014, and ending with November of 2015. For each of those outbound links (the “Tested Links”), we noted the page that the link pointed to (the “Receiving Page”). We took the URL of Receiving Page, and plugged it into each of the three tools to see which tools identified the Tested Links.
Across the 20 sites, we ended up identifying 2681 outbound links, so this gave us a decent amount of data as to what was happening with each tool.
The Results
Let’s dive right in with a snapshot of the results:
In our test, Majestic discovered the most links, netting just over 60 percent of the Tested Links being found. OSE came in at 55.3 percent and ahrefs came in at 50.5 percent.
[Tweet “Study finds no major backlink tool finds more than 60% of links.”]
Important: Considering the size and the scope of the web, this test covers a very small sample size. In addition, this test is deliberately biased toward high authority sites, so this test says nothing about the effectiveness of the tools on lower authority sites. For that reason, I’d argue that all three tools are pretty competitive with each other.
We also looked at the overall performance of the three tools over time:
You can see that no tool identified more than 74 percent of the links in any given month, even on these very prominent media sites. Moz returned the most links in 12 of the 17 months tested, with a strong tendency to find more of the older links on these sites. Majestic returned the most links in five of the cases. While ahrefs didn’t return the most links in any given month, it was competitive throughout.
Also note that the Moz data was pulled via their API. Moz only returns links via their API that have already had their metrics calculated, such as Domain Authority and Page Authority. As a result, these numbers don’t include their “Just-Discovered Links”, and this probably causes October 2014 to come in lower than it otherwise would have.
We also looked at the aggregate coverage of all three tools together:
In aggregate, this brought overall coverage up to just over 80 percent of all the links, a significant improvement over any one individual tool.
[Tweet “The three major backlink tools together find about 80% of links to a site. See study”]
One last thing that I looked at was to see if the data would change much if we looked at only the sites that were Moz Domain Authority 90 or higher:
This did indeed cause a slightly higher success rate for all three tools.
Summary Conclusions
The major conclusions are as follows:
- All three tools were competitive in finding links across the tested sites.
- Even on very high DA sites, no tool found all the links.
- Using all three tools together provided the greatest coverage.
- There was evidence that the tools bias toward coverage on higher authority sites.
More conclusions and takeaways after the video and slide deck below.
Click here to get an infographic summarizing the findings of this study.
Here is a video summarizing the findings of this study:
And here’s a slide deck summary of the study findings:
If you are serious about backlink analysis, it makes sense to obtain access to all three tools. Their coverage has significant overlap, but they also all find links that the other tools don’t. That’s why at Perficient Digital we built our own in-house tool to aggregate the data from these three, plus add data of our own, to give our clients the best possible link analysis.
It is also likely that coverage of links drops in all three tools across sites of lower authority. We saw some evidence of that, and it does seem to make sense that they would spend more time on discovering the highest value links. It’s also my (speculative) opinion that the value of using all three tools together goes up even more as you get to lower authority sites.
Need better SEO analysis of your site? We can help!
Thanks for the research! I’ve often wondered about which tool to use, but usually default to Majestic and Moz. Sounds like incorporating all 3 is the best, most complete way to go. I also didn’t realize Moz was so competitive with Majestic, so I’ll probably use OSE more now since I’m a loyal Mozzer and like the interface better.
I’m curious, if these tools seem to only find a portion of the links, are we even sure Google runs near 100%? Concentrating on high DA sites makes sense as its going to be the strongest ranking signals and provide the highest contribution to understanding search results.
Hi David – there are definitely portions of the web that Google does not crawl, so they wouldn’t see links on those pages, but I’d bet that they get 100% of the visible links on the high authority sites, and for most of the web.
What most people don’t realize is that software companies, such as a Moz, ahrefs, or Majestic, don’t have the same infrastructure to crawl the whole web that Google does. Each of the tools does a great job, but with far more limited overall resources.
Thanks Eric, makes sense.
Hey Eric,
Do you have plans to expand beyond OSE, Majestic, Ahrefs & include openlinkprofiler.org, Google Search Console link data, & SEMrush backlink data (or others)?
We should probably do that. I’ll see when I can get something like that queued up!
Hi Eric, you know I have great respect for you. But these findings are not surprising, and I have been telling my clients these exact same things for over a decade now.
There are many other topics worthy of discussion when it comes to third-party backlink analysis tools. Just to touch on one that you mentioned above, these third-party tools do tend to focus on higher-quality sites, whereas Google crawls whatever it feels will help them improve the results for their end users. The little discussed issue here is that these tools have completely different goals and uses than Google. As an example, when Google comes across 70 million pages in a blog network in Russia, it’s in their best interest to crawl them to learn whatever they can from it. However, a third-party backlink tool is not going to crawl 70 million pages that ultimately don’t serve any value or surface any link targets for their paid users. You can see the conflict here. In Google’s case their algorithm actually improves by crawling the crap, because it learns from it. The third-party linking tools however, have a different goal; they need to please paid subscribers by surfacing link targets. This is just one of the issues at play here, and I have spent hundreds of hours explaining to clients and others just how different do use tools work from Google, and just how careful one must be in how they utilize that data . Thank you for the data and analysis.
I’ve been happy with using Cognitive SEO lately and the portion of backlinks it returns when compared to Majestic directly.
Do you have such a good tool that we can buy?
Thanks for the detailed and insightful comment Eric. I agree that the findings are not surprising. I’ve likewise been trying to educate people on this for a long time, and it’s one reason we ran the study, so I could point to actual data on the matter.
In addition to your points, Google owns the largest computing infrastructure on the planet, whereas the 3 companies we looked at are small software companies. They don’t have the capability to crawl the entire web, even if they wanted to.
Where this becomes an issue for people is when they are trying to use the tools to recover from penalties and/or Penguin. At that point, they aren’t just looking for new links to pursue for their site, they’re looking for links that they should get removed or disavow.
PS – the respect is mutual!
Needless to say, but we knew this but only needed a study which could provide some concrete data over this. However, Rand endorses Majestic when it comes to backlinks. Perhaps this is a great opportunity for people working on tools to take the ship ahead and give some better tools.
Nice post, nice study and nice discussion.
Yes, Google is far too big and far too comprehensive than the three put together. Google’s goal too is different.
Then, backlinks are not the ultimate piece of armor for SEO purposes, though it is one of the strongest.
We have always suggested that users gather all the sources they can and we would strongly suggest all of the following are gathered: –
Search Console
Majestic (Fresh and Historic)
Ahrefs
OSE
Previous link building reports
Anywhere else you can get some…
Even if you do that and manage to de-dupe it all (without a tool, tools help there) you will still only get a percentage of the total and you also cant know how much of the total link map Google has for that domain…
Its amazing what fluctuations you see in the links over time also…
Morning Eric.
Can you confirm that the two big dips in the OSE data are due to MOZ’s recent crawl troubles?
Hi Randy – the October dip is due to Moz not yet having run their metrics (DA, PA, etc.) so those numbers will come up with their new update as of yesterday. As for the earlier dip, I don’t have confirmation, but I think that’s probably the case.
Fingers crossed!
I’ve been using OSE for 3+ years and it commonly misses 100’s of links that Majestic and Ahrefs catch… maybe I’m an outlier. :
Hi there from Slovenia.
Thanks for nice and really educational post.
We, small seo freelancers and homeworkers would also love to read some similar study perhaps about backlinks impact.
For example: I’ve noticed from open link profiler one backlink to my sites marked as 100%. When that happened, I thought I had it all. The entire sample what a perfect and most powerfull backlink would have to be.
Unfortanately my study of it is in Slovenian language, so I can’t share it here, but maybe I can share some info I’ve learned out of it.
My 100% open link profiler backlink was:
– no follow
– no topics related
– no anchor
– a small blog comment on site with far more than 15 OBL
So it learned me one thing:
tools are great, but trust in your own experience is greater.
Regards, Matija from SLO 🙂
Great insights, for site owners/seo’s on a low budget I reccomend exporting majestic, search console list and import into spyglass for deduping.
Hi Eric! Something doesn’t add up for me (maybe it’s due to the 10.5% Belgian beer in enjoying), but could you please share the full data set?
I’m extremely curious why links from such prominent sites are missed (or are being excluded) by these tools. Are entire articles along with all their links being missing by the tools (which ones?) or are some links being missed and others not from the same article? What’s the correlation between links missed and the metrics of the receiving domains (e.g. total number of total links, number of high ranking keywords, link detox score etc.)
I’m almost certain there are links not being crawled on purpose, and/or links excluded from analysis due to the reasons Eric Ward mentioned (I.e., these tools don’t have the bandwidth to crawl everything they come across so they have to pick and choose what to crawl based on what brings the most value to their current and future customers).
It’d be fascinating to dive into the data set and try to figure out what’s going on and figure out the crawling logic of these tools.
Thanks and great work running this test over so many months!
Nice piece of research Perficient Digital what could be interesting if you did this same study with different groups of sites 20 x High DA Sites 20 x Mid DA Sites and 20 x low DA sites. The 20 sites in the original test group are all high DA sites so I presume players like MOZ crawl these at a higher rate, fresh sites MOZ seems to lack big time on them because they delay their crawling.
Honestly day be day, many of people blocking those SEO bots, because they just consume webmaster bandwidth. I have also blocked those bots into my website, and non of them collect accurate link data to my website. Yes they collected some of links, because other sites don’t block them.
But I have seen, MJ12Bot(Majestic) crawl our website more than Googlebot, so once any webmaster look out server log, they surely ban it from robots.txt or .htaccess.
Good one Arjun, this definitely was a factor in the test! To find out to what degree, it’d be easy to validate by fetching each of the domain’s robots.txt files and comparing it to the excluded results of the tools.
Arjun, do remember though that if you block the MJ bot with your robots.txt, Majestic will still have ALL of the inbound link data for your site (and will also be available to the public including your competitors). If your goal is to hide your link profile you will either have to ensure everyone who is linking to you are blocking the bots themselves, OR, more realistically, do some redirect work through a different domain (that you don’t publicize) and block all bots there on the server level so they can’t follow the redirect. That and few other things. It’s not worth the effort for most cases, (better to build a link profile that is so strong it can’t easily be copied) but I’ve been paid to consult on this more than a few times for clients in very competitive verticals.
Hi Eric!
Just curious – what was the breakdown of the types of pages making up the “Receiving pages?” In other words, were most of them home pages vs. deep pages, and do you know how deep, on average, the non-home pages were? Do you think that would influence the outcome…it seems that deeper pages are less likely to get crawled by the tools, especially on lower-authority sites.
I suppose at the end of the day it’s a wash since all 3 tools are dealing with the same pages, but it could suggest some other interesting take-aways:
– Which tool is better at picking up links to dep pages?
– Filtering for only home pages as the “Receiving page”, how good are the tools then?
– What are results for each “level” – i.e., 1-click away from home page, 2-clicks, etc.
Thanks for sharing the study!
Cognitive is amazing. Blows my mind. Anyone working with linking related KPI’s needs to take a look at it.
Hi Eric,
I am using ahrefs and opensiteexplorer to find out my blog’s backlinks, I thought I am getting all the backlinks information, after reading this post I came to know that they are capable of discovering 60% of backlinks, thanks for sharing this study with us, see you soon with another article.
Very true. They all seem to miss some backlinks. I also like to use SEO Spy Glass
Hi Greg – great question. Actually, the pages receiving the links do not need to be crawled, so that should not affect the outcome of whether or not the tools see the links. Once they crawl the page that is linking out to other pages (the “Sending Page”), they should already have the information they need to report on it.
So I think the question is more about how they handle links on Sending Pages, based on their depth. Does that make sense?
The results here are surprising to me, mainly because the tools all performed a lot BETTER than I’d have expected them to. I was especially expecting to see a much vaster trough of difference between Majestic, Ahrefs (both of which I hold in high esteem) and OSE (which I do not).
In almost every piece of analysis I have produced, OSE only found between 1 and 10% of the links which were uncovered through Majestic or Ahrefs. True – Majestic and Ahrefs links were thinned out once I ran crawlers to check whether the links were still live, but there was still a massive gulf between OSE and the other tools.
Most of the stuff I was doing was at the domain (root domain) level. I was also adding in links found via SEO SpyGlass (which is a bit naff, but cheap!) and Google Webmaster Tools / Search Console.
As far as I am aware (from personal experience) this test paints OSE in a very positive light.
As noted in the article, we did focus on very high end sites, with the lowest one having a Domain Authority of 72. I do believe that if we had gotten down to sites with a much lower DA that the performance of all 3 sites would have gotten much worse. My experience also suggests that OSE would have returned fewer links than the others if we had done that.
Backlinking tools are fine for analysis purposes and testing our build linking strategy, but I think that what it really matters are the backlinks that Google has recollected (and shown it the Google Search Console), as they are what it’s used for the search results. In any case, comparing the Search Console’s backlink set with the results of other backlinking tools may arise interesting insights about the way our website is indexed.
Didn’t we all know this already?
You may have, but there are many people who didn’t
Hey Chris,
Speaking on behalf of Ahrefs, I can say that our numbers in this study were heavily affected by the fact, that we’re blocked by 2 of these 20 sites: via robots at http://www.newsmax.com/robots.txt and Techcrunch blocks us at server level: http://screencast.com/t/XrJ7KoyH
So not crawling 2 sites out of 20 makes us -10% behind from the start
Great point Tim , and thanks for letting everyone know about that!
Hi Eric, I hope you could do a study to include Cognitive SEO, SEM Rush, Raven and SEO Powersuite.
It’s interesting to see how these three competitive tools stack up against one another over a period of time. It’s clear that majestic performed the best according to the results but Moz and ahrefs definitely still held their own. Before reading this article I used a combination of Moz and ahrefs but now I am inclined to try out Majestic and add it to the mix. Thanks for the research!
I started using Majestic as a result of reading this article / watching the YouTube video. I really like Majestic but am now using multiple tools. Awesome article Eric Enge.
I think that the best tool for backlink is OpenProfileTool, you can consult links about competience and other websites.
I have used lots of online tools to check the backlink profile but i feel there are four tools which are quite helpful and show the detailed information:
1. Opensiteexplorer
2. SEMRUSH
3. Ahrefs
4. Google Search Console aka Webmaster Tool
I have been using these tools from last 2 years and all of these tools are quite useful to check the backlink profile.
It’s interesting to see how these three competitive tools stack up against one another over a period of time. It’s clear that majestic performed the best according to the results but Moz and ahrefs definitely still held their own. Before reading this article I used a combination of Moz and ahrefs but now I am inclined to try out Majestic and add it to the mix. Thanks for the research!I started using Majestic as a result of reading this article / watching the YouTube video. I really like Majestic but am now using multiple tools. Awesome article Eric Enge.
There is a lot of information on this subject, including steps that you can take today, to improve your link profile and boost your rankings along the way.
In the past, I’ve provided a lot of advice on building quality links, where to find the best links and tools that can provide assistance.
But, there’s something else that you need to do: examine your link profile.
If your website is brand new, this may not be a big deal. And, here’s why: you don’t yet have any links pointing to your website.
Conversely, if your website’s been around for a few months or longer, there is a good chance that you have some links pointing to it. Some may be good, some may be bad and some may not move the needle in either direction.
It’s important to understand your link profile, as this will give you a clear idea of whether or not you’re on the right track.
1. Ahrefs
An absolute favorite among online marketing professionals, Ahrefs gives you all of the tools that you need to track your backlinks and keywords. Furthermore, you can get an inside look at what your competition is doing.
Linkody
From the moment you visit the Linkody homepage, it’s clear what the link checker tool has in mind for its users:
“Never check backlinks manually again, Linkody automates the whole process.”
cognitiveSEO
cognitiveSEO provides a variety of internet marketing tools, all of which can help push your business forward.
Kerboo
With so many Kerboo tools to use, you don’t want to rush through the process of learning more about this company.
LinkResearchTools
The name of this tool says it all, but you need to learn more about the features, to determine if it’s right for your website and approach.
Thanks for sharing that great infirmation
Hi, Eric Enge
I completely agree with you that backlinking tools are not as comprehensive as people may think and keep doing this great job.