Today, I'm very excited to announce that Moz's Spam Score, an R&D project we've worked on for nearly a year, is finally going live. In this post, you can learn more about how we're calculating spam score, what it means, and how you can potentially use it in your SEO work.
Update: Spam Score is now available as part of the Mozscape API. More info here.
How does Spam Score work?
Over the last year, our data science team, led by Dr. Matt Peters, examined a great number of potential factors that predicted that a site might be penalized or banned by Google. We found strong correlations with 17 unique factors we call "spam flags," and turned them into a score.
Almost every subdomain in Mozscape (our web index) now has a Spam Score attached to it, and this score is viewable inside Open Site Explorer (and soon, the MozBar and other tools). The score is simple; it just records the quantity of spam flags the subdomain triggers. Our correlations showed that no particular flag was more likely than others to mean a domain was penalized/banned in Google, but firing many flags had a very strong correlation (you can see the math below).
Spam Score currently operates only on the subdomain level—we don't have it for pages or root domains. It's been my experience and the experience of many other SEOs in the field that a great deal of link spam is tied to the subdomain-level. There are plenty of exceptions—manipulative links can and do live on plenty of high-quality sites—but as we've tested, we found that subdomain-level Spam Score was the best solution we could create at web scale. It does a solid job with the most obvious, nastiest spam, and a decent job highlighting risk in other areas, too.
How to access Spam Score
Right now, you can find Spam Score inside Open Site Explorer, both in the top metrics (just below domain/page authority) and in its own tab labeled "Spam Analysis." Spam Score is only available for Pro subscribers right now, though in the future, we may make the score in the metrics section available to everyone (if you're not a subscriber, you can check it out with a free trial).
The current Spam Analysis page includes a list of subdomains or pages linking to your site. You can toggle the target to look at all links to a given subdomain on your site, given pages, or the entire root domain. You can further toggle source tier to look at the Spam Score for incoming linking pages or subdomains (but in the case of pages, we're still showing the Spam Score for the subdomain on which that page is hosted).
You can click on any Spam Score row and see the details about which flags were triggered. We'll bring you to a page like this:
Back on the original Spam Analysis page, at the very bottom of the rows, you'll find an option to export a disavow file, which is compatible with Google Webmaster Tools. You can choose to filter the file to contain only those sites with a given spam flag count or higher:
Disavow exports usually take less than 3 hours to finish. We can send you an email when it's ready, too.
WARNING: Please do not export this file and simply upload it to Google! You can really, really hurt your site's ranking and there may be no way to recover. Instead, carefully sort through the links therein and make sure you really do want to disavow what's in there. You can easily remove/edit the file to take out links you feel are not spam. When Moz's Cyrus Shepard disavowed every link to his own site, it took more than a year for his rankings to return!
We've actually made the file not-wholly-ready for upload to Google in order to be sure folks aren't too cavalier with this particular step. You'll need to open it up and make some edits (specifically to lines at the top of the file) in order to ready it for Webmaster Tools
In the near future, we hope to have Spam Score in the Mozbar as well, which might look like this:
Sweet, right? :-)
Potential use cases for Spam Analysis
This list probably isn't exhaustive, but these are a few of the ways we've been playing around with the data:
- Checking for spammy links to your own site: Almost every site has at least a few bad links pointing to it, but it's been hard to know how much or how many potentially harmful links you might have until now. Run a quick spam analysis and see if there's enough there to cause concern.
- Evaluating potential links: This is a big one where we think Spam Score can be helpful. It's not going to catch every potentially bad link, and you should certainly still use your brain for evaluation too, but as you're scanning a list of link opportunities or surfing to various sites, having the ability to see if they fire a lot of flags is a great warning sign.
- Link cleanup: Link cleanup projects can be messy, involved, precarious, and massively tedious. Spam Score might not catch everything, but sorting links by it can be hugely helpful in identifying potentially nasty stuff, and filtering out the more probably clean links.
- Disavow Files: Again, because Spam Score won't perfectly catch everything, you will likely need to do some additional work here (especially if the site you're working on has done some link buying on more generally trustworthy domains), but it can save you a heap of time evaluating and listing the worst and most obvious junk.
Over time, we're also excited about using Spam Score to help improve the PA and DA calculations (it's not currently in there), as well as adding it to other tools and data sources. We'd love your feedback and insight about where you'd most want to see Spam Score get involved.
Details about Spam Score's calculation
This section comes courtesy of Moz's head of data science, Dr. Matt Peters, who created the metric and deserves (at least in my humble opinion) a big round of applause. - Rand
Definition of "spam"
Before diving into the details of the individual spam flags and their calculation, it's important to first describe our data gathering process and "spam" definition.
For our purposes, we followed Google's definition of spam and gathered labels for a large number of sites as follows.
- First, we randomly selected a large number of subdomains from the Mozscape index stratified by mozRank.
- Then we crawled the subdomains and threw out any that didn't return a "200 OK" (redirects, errors, etc).
- Finally, we collected the top 10 de-personalized, geo-agnostic Google-US search results using the full subdomain name as the keyword and checked whether any of those results matched the original keyword. If they did not, we called the subdomain "spam," otherwise we called it "ham."
We performed the most recent data collection in November 2014 (after the Penguin 3.0 update) for about 500,000 subdomains.
Relationship between number of flags and spam
The overall Spam Score is currently an aggregate of 17 different "flags." You can think of each flag a potential "warning sign" that signals that a site may be spammy. The overall likelihood of spam increases as a site accumulates more and more flags, so that the total number of flags is a strong predictor of spam. Accordingly, the flags are designed to be used together—no single flag, or even a few flags, is cause for concern (and indeed most sites will trigger at least a few flags).
The following table shows the relationship between the number of flags and percent of sites with those flags that we found Google had penalized or banned:
ABOVE: The overall probability of spam vs. the number of spam flags. Data collected in Nov. 2014 for approximately 500K subdomains. The table also highlights the three overall danger levels: low/green (< 10%) moderate/yellow (10-50%) and high/red (>50%)
The overall spam percent averaged across a large number of sites increases in lock step with the number of flags; however there are outliers in every category. For example, there are a small number of sites with very few flags that are tagged as spam by Google and conversely a small number of sites with many flags that are not spam.
Spam flag details
The individual spam flags capture a wide range of spam signals link profiles, anchor text, on page signals and properties of the domain name. At a high level the process to determine the spam flags for each subdomain is:
- Collect link metrics from Mozscape (mozRank, mozTrust, number of linking domains, etc).
- Collect anchor text metrics from Mozscape (top anchor text phrases sorted by number of links)
- Collect the top five pages by Page Authority on the subdomain from Mozscape
- Crawl the top five pages plus the home page and process to extract on page signals
- Provide the output for Mozscape to include in the next index release cycle
Since the spam flags are incorporated into in the Mozscape index, fresh data is released with each new index. Right now, we crawl and process the spam flags for each subdomains every two - three months although this may change in the future.
Link flags
The following table lists the link and anchor text related flags with the the odds ratio for each flag. For each flag, we can compute two percents: the percent of sites with that flag that are penalized by Google and the percent of sites with that flag that were not penalized. The odds ratio is the ratio of these percents and gives the increase in likelihood that a site is spam if it has the flag. For example, the first row says that a site with this flag is 12.4 times more likely to be spam than one without the flag.
ABOVE: Description and odds ratio of link and anchor text related spam flags. In addition to a description, it lists the odds ratio for each flag which gives the overall increase in spam likelihood if the flag is present).
Working down the table, the flags are:
- Low mozTrust to mozRank ratio: Sites with low mozTrust compared to mozRank are likely to be spam.
- Large site with few links: Large sites with many pages tend to also have many links and large sites without a corresponding large number of links are likely to be spam.
- Site link diversity is low: If a large percentage of links to a site are from a few domains it is likely to be spam.
- Ratio of followed to nofollowed subdomains/domains (two separate flags): Sites with a large number of followed links relative to nofollowed are likely to be spam.
- Small proportion of branded links (anchor text): Organically occurring links tend to contain a disproportionate amount of banded keywords. If a site does not have a lot of branded anchor text, it's a signal the links are not organic.
On-page flags
Similar to the link flags, the following table lists the on page and domain name related flags:
ABOVE: Description and odds ratio of on page and domain name related spam flags. In addition to a description, it lists the odds ratio for each flag which gives the overall increase in spam likelihood if the flag is present).
- Thin content: If a site has a relatively small ratio of content to navigation chrome it's likely to be spam.
- Site mark-up is abnormally small: Non-spam sites tend to invest in rich user experiences with CSS, Javascript and extensive mark-up. Accordingly, a large ratio of text to mark-up is a spam signal.
- Large number of external links: A site with a large number of external links may look spammy.
- Low number of internal links: Real sites tend to link heavily to themselves via internal navigation and a relative lack of internal links is a spam signal.
- Anchor text-heavy page: Sites with a lot of anchor text are more likely to be spam then those with more content and less links.
- External links in navigation: Spam sites may hide external links in the sidebar or footer.
- No contact info: Real sites prominently display their social and other contact information.
- Low number of pages found: A site with only one or a few pages is more likely to be spam than one with many pages.
- TLD correlated with spam domains: Certain TLDs are more spammy than others (e.g. pw).
- Domain name length: A long subdomain name like "bycheapviagra.freeshipping.onlinepharmacy.com" may indicate keyword stuffing.
- Domain name contains numerals: domain names with numerals may be automatically generated and therefore spam.
If you'd like some more details on the technical aspects of the spam score, check out the video of Matt's 2012 MozCon talk about Algorithmic Spam Detection or the slides (many of the details have evolved, but the overall ideas are the same):
We'd love your feedback
As with all metrics, Spam Score won't be perfect. We'd love to hear your feedback and ideas for improving the score as well as what you'd like to see from it's in-product application in the future. Feel free to leave comments on this post, or to email Matt (matt at moz dot com) and me (rand at moz dot com) privately with any suggestions.
Good luck cleaning up and preventing link spam!
Not a Pro Subscriber? No problem!
I'd love to see a disavow tool that allows you to test the effects of disavowing on PA and DA within the context of Mozscape. In cases where the site considering disavow is in the red it could easily be a net positive depending on how much Spam Score detracts from Authority scores.
Hearty cheers to Dr. Matt Peters and the rest of the Moz team.
PA/DA won't be affected by disavow because we don't know what folks submit to Google Webmaster Tools. Mozscape/OSE also doesn't take spam into account - we're like Google prior to 2012 in that we just discount spam links (rather than having them actively penalize/hurt scores).
That said, longer term, we'll be looking into using Spam Score to modify the PA/DA algorithms and may be able to provide something like what you're describing.
Hey Rand, really stoked to see what we can do with spam score.
Quick question - You mentioned that Mozscape doesn't take spam into account; Do you mean that DA/PA won't effectively scale with spam? As in prior to the release of spam score, rogerbot would attempt to recognize if a link is spam and if so, not count those spam links in the DA/PA calculations?
If so, I think it would be an interesting feature to be able to toggle "raw" DA/PA without any spam filtering to the calculations, or more "pure" versions of DA/PA with spam discounted. I've been acting under the impression that those metrics typically do scale with spam.
Hi Nicholas,
Rand was referring to the fact that Mozscape doesn't currently use the spam score or flags to calculate PA or DA. We do not attempt to discount spam links when calculating these metrics.
Since sites with high spam scores are often low quality and may have suppressed rankings as a result, we hope that including the spam score as a component of PA/DA in the future will improve them.
Thanks for the response Matthew. I think the addition of spam score is a huge step in the right direction towards adding some qualitative checks to a quantitative measure.
It's actually fairly common that I hear clients (and even SEO vendors) say things totally off base about Moz's metrics.
"Please only target sites with DA 30+, this means that the site has lots of good links and isn't penalized, sites under DA 30 are most likely penalized."
"Oh, our DA/PA and such will be improving soon, we've made many onsite changes, our title tags are more relevant to the content, etc."
@Rand, yup, my comment was purely addressing future scope. This is a great launch though and another welcome tool. Cheers!
"PA/DA won't be affected by disavow because we don't know what folks submit to Google Webmaster Tools."
...But what if you did, for a small portion of the Web...?
Hypothetically, if you kindly asked a number of webmasters (maybe Moz members?) to pass on the full lists of disavowed URLs on behalf of their clients and their own sites (if applicable), could Moz factor that data into the PA/DA algo? With enough submissions (100s or even 1,000s of websites' disavow data), could you even make a judgement call across the board for those site where you didn't have access to the disavow data? And in fact, would it being useful info to plug into your new Spam Score tool, to help make it more accurate? Or would it be too hard to do/tell?
Given the size of the web and the number of disavows Google gets vs. the small sample we could acquire, I think it just isn't feasible unfortunately. Over time, we have lots of ideas and ways to improve PA & DA that will likely be more meaningful, though.
And yeah - it would be hard to use disavow files to improve the Spam Score system. We really want to keep it based on things that correlate with what Google's penalized/banned, and plenty of links that one might disavow don't necessarily come from penalized/banned sites.
"PA/DA won't be affected by disavow because we don't know what folks submit to Google Webmaster Tools."
But what if you did... All ya gotta do is ask... imagine how many you would get. You could remove those links from the calculation and you should be able to just re-run the site and see the estimated change.
Sounds good and logical
Is the "PA/DA won't be affected by disavow (...)" statement really correct?
I believe Moz uses real ranking data to make sure PA/DA reflects reality. It is well known that Penguin needs a refresh to take disavow data into account (and it seems that Panda needs a refresh too according to recent information released in Google Hangouts). These have not happened in 6 month. Panda and Penguin detect webspam. They have a strong influence on rankings. Therefore, they must have some influence on the fabric of PA/DA.
Therefore, if someone has improved their website's content and worked on their backlink profile, it won't be visible in the rankings until the refresh happens. Since the PA/DA model depends on existing rankings, does it mean their PA/DA is irrelevant until Google refreshes and Moz takes the new rankings into account in its model to compute PA/DA?
Yeah - even though the PA/DA model can re-compute with changes to Google's SERPs, there's no way we can ever know which links a site may have disavowed, or which ones Google's actually counting/not counting.
Things Google does in the SERPs can impact PA/DA, but nothing that happens privately between Google and a site owner (re-consideration requests, disavows, actions exclusively inside Webmaster Tools, etc) can be seen by us.
Hi all. I explained my concerns in an email conversation with Rand, but thought I'd write a few here as well. It needs to be really clear that this tool is not meant to replace a manual link audit. I would highly suggest removing the "create a disavow and send it to Google" link.
I can see the tool being a useful adjunct when you are manually auditing links, but to suggest that you can automate the process at all, is not good.
One of the big problems is that the Mozscape index is not big enough to go to the depths of the web where many of the scuzzy links exist. As such, so many unnatural links are going to be missed. Some of the sites that I tested that had hundreds or thousands of unnatural links had less than 10 showing on the spam tool. Many business owners are going to check Moz and say, "Woohoo. I only have 2 unnatural links" when really the problem is immensely bigger. Others are going to look at links that are labelled as spam score 5-7 and say, "Hmmm....Moz says these could be problematic". I can see people disavowing some good links as a result.
I would highly suggest having a BIG button somewhere or even a popup that tells people that they have to click to say that they understand that this tool alone is not meant to be used in isolation, but that it can possibly help in your link auditing efforts. Even with all of the disclaimers, there are people who are going to mess up their sites by blindly submitting the links suggested by Moz.
Also, you've got a link at the bottom that points to a Search Engine Watch article on using the disavow tool. It's got some decent general information. But why not link to some of the great detailed tutorials that are on Moz? I'm biased because I wrote a couple of them, but there are quite a few that can really help people when trying to figure out if their links are unnatural.
Another concern that I have is that I often see people who are madly disavowing links when their problem has nothing to do with links. It might be good to mention that as well. It seems that many small business owners think that they only way they can recover Google rankings is to disavow.
I wanted to like this tool, but I am really concerned that it could do more harm than good. Perhaps I have misunderstood its purpose. If used as an adjunct to a manual link audit, it could be helpful. But to me, it came across as an all in one solution to link problems. I think that other people are going to assume this as well.
Thanks for sharing your thoughts Marie - both here and with me over email. I agree completely that this tool cannot take the place of a manual review. It should only be used to augment the other types of data/metrics you're using to evaluate things.
We did not intend it to be a complete solution to all link auditing or spam recovery issues, and were very careful in our marketing to simply present it as is - a data source that includes 17 flags we found correlating with sites Google had penalized/banned.
Also appreciate the feedback on Mozscape's index size issues. A few others brought it up and I totally agree. While we have been getting much bigger (from ~100B - ~300B in the last year), we need to keep growing. We've got plans to do that and a sizable team working on it every day - it's a huge challenge because of our metrics (which require processing vs. other indices that use primarily raw counts), but we will get there.
That's awesome that the Mozscape index is growing Rand. Keep it up.
I am hoping that my criticisms are not taken the wrong way as I don't want to lose my good relationship with Moz. But I am just so concerned that people are not going to understand that this tool is not to be used on its own to create a disavow.
I read the blog post quickly, checked out the tool, saw the "create a disavow and send to Google" button and my first assumption was that you were trying to completely automate the process. I'm sure I am not the only one who made this assumption.
After spending the entire morning reviewing the tool and re-reading the blog post I see now that that is not the intent. But I'm still very concerned.
Understood - we've got a bunch of warning (when you try to generate the disavow file, in the blog post above, in the disavow file itself), but we will try to do more to make the risks clear.
And certainly no offense taken or relationship hurt! We LOVE constructive feedback, and we don't mind getting criticized either.
Awesome. To be fair, I downloaded a disavow but am still waiting to receive it, so I haven't had a chance to see the disclaimers there.
If I can help at all improving the tool I'm happy to contribute.
This is wonderful!
I am probably missing something obvious, but I can't seem to find the detailed breakdown of the score of the site I am actually entering in OSE? (I can see how to get that for the links, but not for the URL the links are linking to.)
Great suggestion! The URL's Spam Score in the Show Metrics section should be click-able for a more detailed view. It will be in our next update. Promise.
My work-around for this was to go to a website that we link to and look at our site in their results. A bit unwieldy, but it got the job done.
That's our bad! As Chris notes, we'll definitely get that added. In the meantime, you can hack this by using: https://moz.com/researchtools/ose/spam-analysis/flags?site=moz.com&target=subdomain&source=subdomain&page=1&sort=spam_score&subdomain=www.yourdomain.com just replace "www.yourdomain.com" at the end of the string with whatever site you'd like to see
If your report Rand comes back as '-- /17 No spam data exists for this subdomain'?
Is it treated as being 'Zero'? as below. I think I was just expecting it to say 0/17.
The Pope's site gets a report like that.
We're looking into this right now. We may be accidentally showing -- when we should show 0 (the latter meaning no flags were triggered).
I think the Pope's site should be 1/17 or higher because he doesn't have contact info. :-)
Haha. That made me laugh EGOL.
Understood! We did get -- /17 and our parent org. got 1/17 (no connection to the Vatican I may add).
Which made me think, is this right. - Thanks for clarification.
Right on! I was wondering this same thing! 1/17 for my site! Woot!
If i remove a spam refer, how many time did Moz to update and fix my spam score?
Thanks for this complet information!
We update Spam Score as we crawl sites, so within an index cycle or two you should see it change.
I have a site with many domains, each is a different language version of the same, more or less.I am wondering if google is penalizing us because the language navigation bar is on every page in every site, so one can link between all the other sites. Should I remove this, how should I instead link between language sites if so?
There's something off about our rankings, we're doing all we can to follow googles advice on site design, yet we're so low in search results.
Google is penalizing web sites, I think you should choose one of the domains .com and then add variations for each language. for example .com/gb for united kingdom, .com/no for Norway...
Graphically, I recommend that you have on your home page an area with the flags of different countries for the user to select their language.
Do not forget in your web, put in your logo a link that redirects to your web, this is a thing that helps us to improve positions
Regards
Awesome! Looks great - does this mean you'll start indexing more borderline links so we can see more of a site's complete link profile? I love Moz tools in general but avoid OSE because I get 3 million times more links with Ah & Maj. If I had the same type of link database with Moz that I get with those, I would roll everything into one and it would blow them away. I'm looking forward to that day!
We should be pretty close to Ahrefs' size. Right now they say 319 Billion unique URLs on their homepage and we've got 323 Billion in our latest index update. We'll be working on growing the index over time too (although the next 1-2 indices may be a bit smaller as we're re-jiggering some stuff). I'm hoping to break 400 Billion in the months ahead.
Hi Rand,
Moz mentions 367 million root domains but ahrefs lists backlinks and URLs only. I wonder what the difference is in the number of indexed root domains. I think that's more important than the actual link count, because one root domain can have millions of links.
Huh. Yeah, that's odd. I would have expected them to show the number of unique root domains, too.
Rand, I want to leave a really open comment to help you see what I see every day "in the trenches."
I'm stunned that the Ahrefs & Moz numbers would remotely resemble each other on the top level. I use both tools every day all week and the numbers are never *close* to each other. Here are a few examples:
For my own site, Moz shows 2 links from 2 domains. Ahrefs shows the links that were built when my site got hacked - thousands of links. It also shows a couple hundred real links from about 40 domains.
I can't find ONE example where Moz has within 50% of Ahrefs data. Usually it's 10 to 1.
As I use backlink data for disavows, I want to find everything. I would rather have too much data than not enough. Majestic historic links blows everything away by insane numbers as well. So for a "complete" link history, I can't use OSE. Make sense?
(Mattcutts.com - Moz shows 218/2850. Ahrefs 20k/500k. I mean ...)
I'm sorry if this note comes off a bit rough on Moz.
Not at all! We've been working to improve index size, and we know it's something folks really want from us. I also have the same perception you do - that Ahrefs & Majestic always report a lot more link counts than we do. I wonder if some of that is us de-duping and them not? But that can't account for all of it. Totally agree we just need to keep growing our index.
Great new addition, and has been mentioned above, if OSE if OSE had more effectively detecting links like other tools like Ahrefs such be a more effective module. After having tested a little, flags, well understood, seem pretty accurate.
Congrats!
Ooo a new toy... ;-)
Quick UX suggestion when you look at the Spam Flags for a site: in addition to greying out the non-relevant flags, remove the tick next to them (or change it to a cross). Leaving the tick next to all of them is a little distracting to the eye, but only showing the tick next to the relevant flags will help to make them stand out even more.
Hey Steve, thanks for the great suggestion! I'll get this prioritized and hopefully get this in the next round of updates.
FYI - Ben's the designer on Spam Score :-) Thanks for jumping in man!
Hi
This tool is really bad for all websites because it can scare away all customers of this website by making a calculation error.
It is impossible to give a afirmation that google has made it harmful and dangerous website. If this is a paid directory, such www.annuairefrancais.fr, all customers will see a 2 spam rating will cancel their registration.
Off, this directory will soon celebrate its 20th anniversary, it consists of 1000 subdomains Categories and No spam. 6 million visitors consult the official website each year and thousands of companies enroll each year.
It is irresponsible to make tools like Nostradamus predictions, the best engineers and scientists we never managed to decode the google algorithm, and this is just scorespam tool created to make money.
It is completely irresponsible to make such a dangerous tool, without having certainty. It's like saying that because of probability calculated, your neighbor has 2/17 chance of being pedophile and that you show this result to everyone.
And finally, if you want to know the source of this calculation, we must paid ! This is blackmail?
No sorry, is not serious...
This is an excellent new feature. I was just looking this morning for a link analysis tool that would automatically calculate spam risk for a given site's backlinks. I didn't want to pay for a new tool, so I'm glad to see that OSE is adding this to their toolbox. Thanks!
I am trying to better understand the use of a text to HTML code ratio as a spam flag, especially in light of the answers to this Moz Q&A: https://moz.com/community/q/text-to-html-ratio I've seen it utilized in that manner in at least one Microsoft patent, but without much in the way of explanation:
Using content analysis to detect spam web pages
[0030] Some classes of spam web pages can be detected by analyzing the content of the page and looking for "unusual" properties, such as (1) the page contains unusually many words, (2) the page contains unusually many words within a <TITLE> HTML tag, (3) the ratio of HTML markup to visible text is low, (4) the page contains an unusually large number of very long or very short words, (5) the page contains repetitive content, (6) the page contains unusually few common words ("stop words"), or (7) the page contains a larger-than-expected number of popular n-grams (sequences of n words) for example. These metrics or filters can be input into a classifier for deciding whether or not a page is spam or determining the likelihood or probability that the page is spam, by comparing the outputs of one or more of the metrics, alone or in combination, to one or more thresholds, for example.
I suspect that this problem is one that might be common in Some WordPress Blogs, and I'm not sure if adding more HTML markup to the site is likely to improve the quality of the site. I could write shorter blog posts.
I'm trying to understand why this is considered to be a potential indication of spam by the Moz team. Was it a metric that was tested against a body of sites that has been penalized and tended to correlate higher than other signals, and selected because of that? Thanks.
Sorry if I wasn't clear - NONE of the flags above are necessarily signs that something's wrong/broken with your site and none of them are intended to evoke action. These are merely features we found that, when taken together, correspond well to sites Google has penalized/banned. For example, not showing an email or a social account on your pages may be a totally fine thing, but it was a feature that correlated well with sites Google had penalized, so we made it one of the flags.
In general, I wouldn't worry too much about a flag being triggered or about fixing them. The list of flags is merely meant to show which ones we found, not to say "this is wrong/broken/work on it" and we try to avoid that kind of language.
Good evening!
I would like to make an answer, because I'm not sure about this. Google and moz are two different things, but in this case moz is one of the few webs where we can analyze this parameter. For example in Google search console we don't know nothing about the spam score or similar parameters. So the spam score in moz has any influence in the positions in google?
Thank you
Hi there
We are using the MOZbar, but the last days disappeared the SPAM score, and is (was) a great feature.
What happend ?
I think the problem is already solved, it works for me and I can see the spam score perfectly
Should be a momentary error
sounds very useful Rand, I can see complete spam signals form inbound links domains, but what about my own domain?
You can also know the position of your domain for the spam score. But if you want to know why your spam score has that score, you can only know it with the payment moz tool
Do you know some free tools for know the percentage of do-follow and no-follow links?
Tank you so much Mr. Fishkin
Hi again Solaire! I have the same question. Anybody can help us?
Glad to know that my Spam Score is 0 /17. Thanks for the new service. A good indication before storm :).
Gonna check couple of more sites Im curious about. Thanks Rand..
I have checked it for few websites. For one website namely "askpctechies" it is not showing any value.
"SPAM SCORE: ---"
What could be the reason?
For your kind information the site is new and it got live in 2nd week of Dec 2014.
There's a few possible reasons, but my guess would be that we haven't seen many links to the site yet (not enough to do the sample crawl of pages and to have data for the other flags).
Thanks a lot Rand. But, it seems that link is not the actual cause because one more website "go4hosting.in" is having enough links but the result is same as above. Which other reasons could be there?
We may have been blocked from crawling the site. The site could have been down when we attempted our crawl. It could have been an error on our end (anything from the crawl itself to a data issue to hardware). Or, it might be an unknown unknown. Regardless - thanks for sharing the example and we'll certainly look into it!
I have this result for one of my website: https://prntscr.com/6oh4vl
What is means?
Interesting article and really helpful
From Psychological perspective, this will motivate non-spammers to continue their work.
It would be more fun if it was available to those who aren't subscribe, at least one check per month, so it would be an accessible tool for every spammer and hopefully demotivate them :)
Great job, who knows what your next invention will be :D
Good suggestion, I'll pass it along to our product team.
Nikola - the score itself is available in the free version! The deeper analysis and list of all scores is part of paid-only, but you can see Spam Score for any given subdomain free.
Where did the Moz "Community" button go?
Hey Patrick!
If you're logged in, you can click on your profile picture on the top right-hand side of the page. From there, you can click on the "My Q&A" link. Clicking on the "Learn & Connect" link at the top will give you an option to go to the Community page as well.
The spam metrics I'm seeing for some sites (coming from the perspective of a white hat SEO) almost scare me. I've got one particular client that has very good organic rankings. They'd hired another SEO company years ago who did the spammiest of link building. The linking sites serve zero purpose whatsoever. It's so blatantly obvious it's for SEO and nothing else. These sites are getting spam scores of 3-5 on average. The client has never been penalized. The vast majority of their links are CRAP. No usage of keyword-rich anchor text. That's the only 'saving grace' I guess. I'm amazed they haven't been penalized and continue to rank well. As an SEO, if I'm looking at this, I'm almost encouraged to mimic this type of strategy. It's produced solid organic rankings, no penalties, and Moz is basically telling me that there's a 10% chance of spam.
Steve - be careful! Just because you see those links doesn't mean Google's counting them. The site may have disavowed them long ago, and there's just no way to be sure if they're what's causing the rankings.
Rand - The site/the client has not disavowed the links. Are you suggesting that Google has disavowed the links/isn't putting any weight on these links? There's very little else that they have right now that would drive their authority/rankings. If I just look at the site's legitimate links, there's no way that they would rank as well as they are for relatively competitive terms. This is about more than just this particular client. I'm just very surprised at the low spam scores for clearly spammy looking sites. I trust that you guys know your stuff, so if you're telling me that a site is a 3/17, despite what it may look like, I'm trusting you that it's legit. I'm surprised, but the conclusion that I'm coming to in this particular case is that very questionable-looking links are actually working.
Steve - I completely follow where you are coming from. I've seen several sites with only linkfarms as backlinks and they are still ranking good. But even though it is easy to come to the conclusion that it's the way to do it, there may be many other factors playing in to why they are performing well. Links are important, yes, but not at all the only thing that matters. Maybe they have great engagement (low bounce rate, high CTR, etc.)? Maybe they are very active on social media? I don't know.
I just wouldn't be tempted to go and do some fishy stuff linkbuilding-wise since it might be something else completely.
Thanks Moz finally we have great tool for analyst spam backlinks for our website.
Working with it now - I think it will be helpful - thanks moz team :-)
Is there a known issue with the 'no contact information' flag? I've always had a contact page with form and email details plus every page (a) has a link to the contact page and (b) all social media links but I have a MOZ spam flag. I've waited a few updates to see if it self corrects but it doesn't.
The problem that I see with this tool is that some of the 17 "flags" are rather meaningless (in my opinion)....
"You don't have contact info across your site."
"Your domain is MothersFantasticSpaghettiSauce.com"
"Your site has 3 pages."
"Your domain is Route66Guide.com"
"Small site mark up."
These are either determined by the topic your site or your way of doing business. They have nothing to do with spam or low quality. The five of them combined should never be a problem. If it is a problem then Google is really messing up.
Several other flags could exist on a site for no bad reason at all.
However, if you have a bunch of cookie cutter pages... BAM! The Panda is likely to get you.
So, I can easily imagine a site with 7 or 8 flags being A-OK, but a site with one flag taking a heavy hit.
Yeah - I think this is something we're going to have to keep trying to communicate. The flags we identified are purely those that CORRELATED with sites we saw Google penalize/ban. They aren't causal, they don't represent our opinions of what spam is/might be, and they are not meant to suggest any action on the part of a site owner. They're simply the features that, in our large dataset, were useful in predicting the percentage chance that a site would be penalized/banned.
In a way, you're absolutely correct - many of those flags are meaningless beyond their ability to predict the percentage correlation. I'm trying to think of good, new ways we can communicate that in the product and in our content about it - you're certainly not the only one to have that perception!
p.s. 69.4% of sites w/ 7 flags and 43.2% of sites w/ 8 flags were, in fact, totally fine.
In other Moz products, the user is given a prioritized list of things to "look at for improvement". The same could be done here.
Flags such as "small number of pages" could be given a small flag as
"opportunities for improvement". These folks might receive a link to an article that would explain that "Increasing the number of pages on your site and the number of keywords that they target can increase the traffic pulling power of the site. Also, adding more pages of customer support about how to use your product, how to select the product, etc, will increase the engagement of your visitors and their satisfaction with your brand if you do a good job.
Flags for "thin content" and "bad links" could be given big flags as a "high probability of causing problems". These diagnoses would be accompanied by links to articles about Panda and Penguin so the visitor can understand them fully.
I think that the tool has potential as a multiphasic website analysis that flags potential problems, explains them, and then links to content resources and other tools on Moz to help the user learn more and improve the website.
I think that it would be a great way to help webmasters and an even better way to introduce webmasters to the great content library on moz.
I think there is some confusion about the different ways that the spam tool can be used. One use is to evaluate whether a link that is pointing to you is likely to be spam. What EGOL is talking about though is using the tool to determine whether there are factors on your own site that could be improved.
I think that most people are confusing the two.
When I first saw the tool, I definitely thought that Moz was attempting to automate the process of link auditing. I think that Moz needs to find a way to make it much more clear that this is not the case. But I see now that what they are saying is that in all of the data that they have collected, looking at spam sites, there are certain things that are often correlated with being "spam". For example, a site that has a number in the url, no contact info, and a low DA was determined mathematically to have a higher chance of being spam than one that is not. But, this does not mean that ALL sites that have numbers in the url, no contact info and a low DA are spam sites.
Another way to look at this would be to say, if I manually looked at 100 sites that contain a number in the url, no contact info and a low DA, I might determine that 80 of them are spam. (I'm making that number up.) So, if my sample size is big enough I can make an assumption that 80% of the sites that have those criteria are likely to be spam.
But, it's important to know that these are guesses that Moz made. No one outside of Google knows what criteria Google uses in their link based algorithms. The Moz spam tool is a neat idea that looks at a number of factors and determines mathematically how likely a link is likely to be spam. In many cases it's going to be correct, but not in all.
I am guessing though that Moz did not fully intend the tool to be looked at on your OWN site to say, "Oh no....I have a number in my url, I must be spam!" I think that there are likely a few things that we can take from looking at the information the spam tool puts together on our site and apply common sense to them to see if perhaps there are ways that we can improve the site that we hadn't thought of before. But I don't think Moz intended us to evaluate our own site to determine if it was spammy or not. (I could be wrong here though.)
I'd be interested to know how Moz determined what spam was as this can be subjective. What I consider spam, someone else may not. Many of the sites that are coming up with high spam scores are not ones that we necessarily need to disavow in my opinion. They're crappy sites that just link out to every site on the web. The disavow tool was meant to ask Google to not count links that we made for SEO purposes. There's no harm in disavowing these crappy sites...probably....but still, they're not ones that Google is expecting you to disavow.
And one final note...I was auditing links yesterday and I came across a site where I was having a hard time deciding on whether or not it was spam. I realized that having the Moz spam score in the Moz bar could be helpful to me. I wouldn't base my decision solely on what Moz says, but if I was trying to decide yes/no on a disavow and I saw that Moz had the site highly correlated with spam then this could be helpful to me.
I wish I could give this comment 5 thumbs up! Very well said Marie.
We're working on better ways to communicate what you've said here in the product (and I filmed a WB Friday yesterday to hopefully help with this too). And yes - we are planning to get this into the Mozbar very soon!
Re: How we determined if something was spam - you can see the methodology explained in detail in the post above (in the section from Matt). Basically, say the site was "www.randsmustache.com" - we'd search for "randsmustache" and "randsmustache.com" in Google and if the site "www.randsmustache.com" didn't rank in the top 10 results, we'd presume the site was penalized or banned. There were likely a small number of false negatives in this case, but given the very large sample size, the statistical impact would be negligible. And yes - that definitely means that Google probably gave lighter penalties to plenty of sites we did check (i.e. they rank for their own name, but might not rank for other stuff), but we wanted to be confident that we were only running our analyses against stuff that we had extremely high confidence was penalized.
@EGOL - In other tools, we do give a list of things one should look into/attempt to fix, but that's not what the spam flags are meant to suggest. As I tried to say in my earlier comment - firing a flag doesn't mean you should necessarily "fix" something, or even that something is wrong. The flags here are simply correlations - not causal, not negative issues (necessarily), and not recommendations of stuff to fix. Marie's comment does a great job explaining/re-iterating this, too.
Thank you, Rand. Marie's comment is very good.
That may be true but if sites are showing a 5-yellow for spaminess it is highly likely lots of people are going to want to reduce this scary looking feedback about their site.
It may well be changing to pass by adding twitter buttons and making whatever tweaks to get rid of several more flags is what is likely to happen.
My guess is a spaminess rating that wasn't just x/17 but a factor of how many of 17 tripped plus an understanding of how important that was (I would imagine including which interactions of spam flag were more critical...).
I would be surprised if there isn't a big difference in a certain 3 flags being tripped versus 3 other flags being tripped (plus say 4 other random flags). That is to say, even with your limited ability to know what Google is directly reacting to versus correlations you can observe. I would imagine this could big improved into a 100 point (or whatever) system that gave a much more valuable spam site insight than just treating each flag as equally important (and ignoring especially deadly interactions between flags - which flags when they are tripped together cause the likely spam hit to be seen in google results.
@curiouscat - actually, when we looked into the relative penalizations percentages for various unique flags, the deltas were so small that we decided against using them individually, rather than as a count. It's not the case that one flag is much more risky than another - it really was, at least in our data, the counts that lined up best with penalization.
Hi, I join this blog just to have a rant about, how unfair this spam scoring is. My website reached 6 flags and I don't have many links to check up on.
* low moz trust and rank score (of course you won't trust the websites not using your services)
*large site with few links (well it's new! and I am not spaming for links)
*small proportion of branded links
Stop scaring people with your statistics, you almost gave me heart attack!
Now when I have it of my chest, have a nice day
Exactly you are right.
For eCommerce sites, there are few automatic flags though e.g. Thin content, large number of pages from start etc.
I am not sure if there is an easier solution to it though
Coooool... checking it out now - thanks moz :-)
I'm concerned about how "no contact info" is determined. For instance my site, Tour de Force 360VR has phone and email in the footer of every page. Street address is on the About page. Similarly my other site, Scott Witte Photography, has all contact info on the HTML page and in the Contact section of the Flash interface. Yet both these sites are cited as having "No contact info." Been this way for some time so it isn't a case of "Maybe we will catch it on the next crawl." I find similar false negatives on a number of sites I've checked.
How exactly is "no contact info" determined? What exactly must be done to prevent it?
sounds like a very useful tool! Will have to play with it and see what to make of it
Quite helpful post!!
Hats off to Moz - Awesome tool, very metric centric but also very practical in the sense that I can now use this tool and apply within the campaigns of all our clients starting today. Personally I think this is what the SEO world really needed to bring a little bit of stability into an extremely complex problem. I don't think many have the balls to even think of doing something like this, yet the guts to actually go ahead and research it and then have the integrity to release it to the wider community so we can all use it.
can we just try and get links with a highish spam score to be made no follow, would that solve the issue ?
I wouldn't feel super-confident in that for every case, but certainly if you can, Google seems to ignore nofollows from spam sites.
In one of my attempts, Google has showed nofollow links in their penalty rejection notice!!!
That's interesting, Google does tend to have a fairly discretionary view when it comes to no index and no follows which is really frustrating. It makes sense for them to ignore no follows from spam sites as they are still spammy! Really good tool Rand, especially for seeing/checking on signals at a glance. As you said, it is not perfect but with some constant tweaking from Dr. Pete it will only get better. Tools should be used in conjunction with manual testing instead of trying to make them do everything for you.
Yes, indeed Simon!
Thank you very much for this good article , and how I could fix ranked 4th flag
thanks again
Good one Rand.
I would have liked to have an option to sort the links in ascending or descending order based on the spam score from the OSE 'Inbound Links' tab itself with a quick drop-down or an overlay window of spam flags against each link. And a link at the bottom says 'click here for more detailed view' which then gets me to the 'Spam Analysis' tab.
Also, in the 'Spam Analysis' tab, it will be much good if we can have details as to whether the link (from a sub-domain) is followed or no-followed.
Love it! Curious...I didn't see anything about checking for many links from the same C-block. With many of the clients I work with--who've been naughty and they know it--I'll find they've bought into simple link farms, where the sites share the same C block (often, even the same IP address). Do you have the ability to spot/flag this in there somewhere, or if not, do you plan on it? Extending one step beyond that, detecting many links from domains with some of the same ownership info would be helpful (of course, Google being a registrar probably has access to some of the domain info that you can't get your hands on).
Hi Michael,
Great question - while we include flags related to diversity of links across domains we don't include any flags for diversity across IPs or C-blocks due to technical reasons (this is currently difficult to compute). We do hope to include some flags in the future using Whois data to capture the multiple domain ownership case (but again, there are some technical hurdles to making this happen).
It looks good at first glance. I noticed one thing that had me puzzled. I owned a domain name before (www.biondocommunications.com) and about six months ago or more pointed it to my current url (www.biondocreative.com). The old site shows a medium spam score. It has been a 301 redirect for more than 6 months, so I am not sure if the Open Site Explorer will eventually detect a 301 and eliminate the domain from its index or follow the 301 and perform the spam analysis on the new site if the link exists there.
thanks much,
Anthony
Good question -- we skip any 301 when we crawl the site to assign spam score. The score you see now for the 301 is from an old crawl that will eventually be refreshed (and flushed from the index).
Thanks for all for giving such valuable information. I am also facing some of those problems and now i have a solution's because of you people. Thanks you friends.
Thanks Rand and Moz team.
I've been working off your spam score pages to build a disavow file. We are a very large website with thousands of spam links, and I was on 1050 domains checked, when it appears that Moz did some sort of update last night (6/9/15)... new spam scores have been assigned, new sites added, and my list is now totally off. Is there a way to go back and access the data prior to the update?
Hi Sam - unfortunately, each index replaces the last and we don't show historic spam scores via OSE. However, the good news is that the latest data is almost certainly of more use/value than the old index (which was a couple months out of date as of yesterday). The new scores and the new domains you see should be a better representation of reality. Sorry about the timing!
One more thing - we do have the ability to expose previous index data via the API. If you email [email protected], he can give you access to that. It will be purely via API though, not included in Open Site Explorer. Hope that helps!
Yes, that helps. Thank you!
Always doing case studies the latest one I would like to share quickly and see people's comments...
Moz Spam Score:
Base Line Site: Wikipedia has a score of 4 out of 17.
Test Site: I set up a website and spammed the heck out of it with the worst links (most people will avoid these types of links for Tier 1 or 2). Most of the links had no relevance, placed on hacked sites, on adult/pharmacy websites and high OBL over 1000+. So what was my final score 2 out of 17. Yep it went and beat Wikipedia's score?
Conclusion: Moz Spam Score Tool does not work and gives the wrong answers clearly.
Hi MozTester - I think you might be thinking of the score the wrong way. The "worst links" that you pointed to the target site are the ones that likely have high spam scores. Pointing a lot of bad links to a site that doesn't trip any of the other filters isn't going to increase the number of flags fired. If you look at the list of flags, you'll see that most of them aren't about incoming links at all.
Also note that the scores are about correlation, not causation, and that given the web's size, saying 95% of sites with this flag count have been penalized/banned still means there's hundreds of thousands or millions of sites in the 5% that haven't been.
Hope that helps!
Nice post, i just to blogging today but never this is very good
Hello Guys,
I would like to integrate spam score on my website, how can i do this?
You can integrate Spam Score into your website by purchasing a subscription to our Moz API.
https://moz.com/products/api/pricing
Hi, I am the webmaster of https://www.uhren-shop.ch. I checkt the homepage my moz and was shocked to see, that my private homepages hubergenealogy.com, hubergenealogy.ch, patrikphilipphuber.ch, patrikhuber.com and patrikphilipphuber.com are listed as 10 in the spam score. The sites have all the same content (relink) and very old (made around 2005) and never updated since. I have certainly not linked to spam sites nor purchased any links, the external links (also to uhren-shop.ch) are no follow links.
Since I do not want to negatively affect my business website, I would like to ask if anyone has a tip on what to change. Since I do not want to negatively affect my business website, I would like to ask if anyone has a tip, I can change.
BR from Switzerland
Patrik
Hi Patrik - the links likely won't hurt you, but you might use Google's Search Console (https://www.google.com/webmasters/tools/home) to disavow those links just to be sure that Google knows they're not ones you intentionally created. https://moz.com/blog/guide-to-googles-disavow-tool from Marie Haynes can also be helpful if you want more info about the disavow process/tool.
Thank you very much for support!
Awesome. A grea tool.
My wesite has some backlink spam core is 4. I am not sure how bad it is. That is reason why i search to this page.
Thanks Rand Fishkin!
Ho Nguyen from Luoi Cong Trinh
Hi, our site got a spam flag of "No Contact Info" and yet we have our email address displayed in one of our footer widgets. We also have contact numbers displayed there and in our footer page. Is Google discounting the contact info in the footer?
Hi Ronald - it's possible that our crawler didn't see it when we last indexed the pages, or that we visited pages on your site that didn't have it, or that we simply had a crawl or parsing error of some kind. If your pages show in Google's index with the correct contact info, there's nothing to worry about.
Thanks for this great information and tools that helps us to improve the quality of our site.
A great new feature - a good way to quickly examine our backlink profile and potential new link opportunities for spam.
Any recommendations if one of the spam links is a forwarding domain? I had to forward a client's domain over for a number of reasons. There are some bad links pointing to that domain, which are likely read into this.
You can disavow links that point to a forwarding domain (just make sure to register that domain with GWMT).
That's really nice work you guys have done. It will help to identify the spam links. Moreover, webmaster will be more aware of the situation of their sites.
It'll be interesting to see how well this correlates in real life. I agree with some of the flags - numerals in the domain is a big one. Not so sure about others, eg branded links. It very much depends on the type of site. Ecommerce sites should have a lot of branded links, info sites very few because people linking to info sites tend to link to specific pages with anchors that pertain to that page. They couldn't care less about the brand.
I understand where you're coming from - but whether we agree with the flags or not was irrelevant to us in this process. We didn't want to use our opinions, but rather what the data showed (and whether a particular feature was correlated with sites Google had penalized/banned). So while I agree there's plenty of times these flags shouldn't indicate spam and are not necessarily a problem for a site, I'm confident in the math showing the counts that line up with the percentages of penalized/banned sites.
I am not sure at the moment about this metric. I will monitor it
My blog has 8/17 score SPAM ! I do not know why to fix all !
Hi Rand I am Vikas Singh Gusain, can you explain : how to share content /links in best possible way without detected as spam. Startups need to do these thinks ( social sharing ,link building ) .
Spam Score is a great idea, there should be negative marking too...
Thanks for the helpful tool to detect the probability of a link being Spam. I had a basic query as part of toxic link removal for a website. Even after disavowing, is it also important to send mails to individual webmasters to remove the link from their website? I have heard from many websites that website owners even resort to pay off those webmasters to remove the link. Some websites are so poorly managed maybe using robots etc. and the webmasters hardly cares for a request to remove a link from a mammoth directory. Is it still the responsibility of the website owner to do whatever it takes to get the link removed? Should Google not consider the Disavow list as a reference to stop counting any link juice from disavowed links? Please share your views. Thanks!
What the hell? 48% of site is penalized because have numbers in his URL? Where is the logic?
The way it works is a little more subtle, but not too complex. Basically, the 17 flags (one of which is "numbers in domain name") correlate highly with sites Google's penalized/banned. That's not to say that these flags are the reason why Google penalizes them - they're merely strong indicators we found of sites that were penalized. The more flags, the more likely to be penalized. E.g. our data showed that, on average, ~52% of sites with 8 flags were penalized.
Spam is definately something to avoid for any PPC or seo consultant. Great tool and metric Rand. Matt Cutts and Google is always cracking down on it. I would fix any spam found. I woldn't want more than 25 percent at most on or off any sites and set a goal of no more then 10%
Awesome new tool mate. I would love to see it working while having the group by subdomain and social media thingie. Would be much more tidy :)
Gotcha! Yes - I think we can get on the list of stuff to add
@Rand Fishkan, Sir as mentioned earlier by you issue has been resolved or not ? if it has been resolved , then pl guide me what to do for my website i.e. https://www.servostabilizer.org.in OSE is telling my spam score 4/17 with following problems :
Large Site with Few Links
Site Link Diversity is Low
Small Proportion of Branded Links
My website aalpha.net shows low number of pages found, is this referring to actual pages crawled by MOZ? whats the threshold to calculate low numbers?
I want to check my website spam level. how can in do that ???
This is really useful. Great to see the Moz tools gradually improving
Now if only we could improve the uptime of OSE :)
We had some network issues last week, but since Friday afternoon, those should be resolved (we had gated this launch on whether Saturday and Sunday were error-free, and that appeared to be the case). We'll keep working hard to get back to 99%+ uptime.
For some of my blog its showing --- only . What the reason ?
You'll see that when we haven't yet completed the sub-crawl of pages on a site to determine the Spam Score. Usually happens if the site with the -- is very new or we didn't discover any links to it yet.
can you let me know How to check spam score in bulk?
Hey Vikas,
You can get Spam Score in bulk via our Mozscape API. https://moz.com/products/api/pricing
Well, I was going to check this tool and it seems it has been removed from Moz.com (the URL https://moz.com/researchtools/ose/spam-analysis is currently redirecting to the OSE homepage).
If you have disavowed spammy backlinks, and those were causing a spam score, will the spam score go away after a while or will it stay forever anyway, since the links are still there but are no-followed?
Is there an appeals process to ask for a manual review to change a spam score?
Hi,
My score is a whopping 7 and I do not have a spammy site in my opinion, and I am not a spammer, this I know for sure. But on the front page I have a 3,000-word article with a large comparison chart that contains info about each product, an internal link to each review, and an Amazon affiliate link to each product. I don't think my content is thin, each individual review is over 1,000 words. Two-thirds of the posts on the site are non-affiliate.
I did not bother to no-follow the Amazon links as Matt Cutts said you don't have to. I disavowed all spammy backlinks (none put there by me; after that, my DA went down).
I have up to 50 external and 50 internal. In SEMRush I get a "too many links" flag and it says 100 is too many. But Moz says only external links are a problem. Is too many links the issue, regardless of their affiliate status?
My chart shows in the Google Answer box for some good queries, so I am reluctant to alter it at all, but I will if it will lower the spam score.
Nothing in the spam-flags list mentions affiliate links. Would no-following the affiliate links help, or do I need to delete most of them? Is the 3,000-word article too long?
The only thing I see in the Spam Analysis is that there are 8 sites with a spam score (from 1 to 4) linking to my site. Is that enough to cause me to get a 7? They are not spammy sites.
The other thing that mystifies me is this: "Spam Score currently operates only on the subdomain level—we don't have it for pages or root domains." The chart with affiliate links is a page, not a post. Does that mean you are not even looking at that page with the chart?
I have a TLD. As far as "subdomains," I am not sure what this means. When I built the site, it was on (mymainsitename.com/thissite) in my hosting platform, since they make you do it that way, but it is its own site, not a subdomain as far as I know. I have zero understanding of this issue, can you explain?
Sorry about the long post but I am very frustrated with this spam score thing. I don't want others to be scared to link to me or get linked to from me, and I certainly do not want to be penalized by Google. I was assuming it was the Amazon links but that paragraph about the subdomains has thrown this theory off.
Thanks very much.
How accurate is the Spam Analysis?
My site is showing up a spam score of 4, but when I looked at the issues flagged up, I found that 2 out of the 4 flags are not accurate.
This is what's coming up for my homepage:
A subset of pages within this subdomain has a large number of external links.
There's a high ratio of anchor text compared to content text.
There's a large number of external links within sidebars and footers.
Crawl only gets a valid response to a small number of pages.
Issues 1-3 are not accurate. I don't have any external links on my homepage at all, let alone in the navigation.
I'm wondering if this is a known bug with the tool?
If not, then how do I actually find the links it's referring to?
We have seen a few inaccuracies of late and have an engineer looking into a fix - hopefully in the next index. Thanks Jay and sorry about that.
Thank you Rand. Any idea when the index will be? I've seen this spam score since last year and it's been baffling me since. I guess I can now ignore the score without worrying too much.
Hi Jay - the next one is scheduled for March 30 (you can always see the latest ones here: https://moz.com/products/api/updates). That said, until we have the fix in place (which isn't quite yet), it won't be reflected in the index update, so it's likely that won't be this index - possibly the next one or one after depending how long it takes our engineers to figure out what's wrong and patch it up.
Hi Rand,
Will you also fix the "no contact info" bug?
Yes - we should be updating all the data once we get this collection issue solved.
Awesome! I'm just relieved. Spent ages trying to work out was wrong and where the links were being seen. I was starting to worry that my site had been hacked and injected with hidden links. I can at least stop worrying and wasting time trying to fix this issue. :-)
Maybe I missed it in the article, but when you were analyzing the data, what defined a penalized site? Does it have to have a manual action? If not, what did you use to make sure these sites were penalized when determining the ratio of sites penalized to each flag?
To find penalized sites we ran a Google search with the subdomain name as the keyword, e.g. "subdomain.domain.com" and looked for ones that didn't have a result in the top 10 (except for an optional "www")
I'll try that you suggest and see in 3rd page competitor site, but when try with "keyword.domain.com" with his url google show his URL in 1st position. What I'm doing wrong Matthew?
Hello Rand,
I just have one question about "Large number of external links": does that refer to the fact that the website has too many links towards external websites? I am asking this because someone on Moz's chat told me that this actually refers to links from a page to other internal pages as well as to external pages.
This is a doubt I have because analyzing a client's website, it has 4 spam flags and one of them is "Large number of external links". Each website page has lots of links towards internal pages but not that many towards external sites. Actually the only links it has towards external sites are links (few links) towards the company's YouTube channel. That's why I was surprised seeing this flag for this website. Hence my doubt. I just wanted to make sure with you that what I was told on the chat is correct.
Thank you,
Frederic
Hi Frédéric - yes. Large number of external links refers to links on the site/pages that point out to other sites. Spam sites tend to have hundreds or thousands of external-pointing links, while most quality sites have only a few dozen or maybe a hundred across a small sample of pages (which is what spam score looks at).
Thank you Rand for the answer, it's all clear. I am very surprised though because my client has only a few links pointing to YouTube, nothing else. It's the website's new version. The old version had lots of links pointing to external websites but it's not the case anymore. Cheers and great job on the tools, Moz team.
Great,Thanks , Just saw the spam score, Metric , it will be great help for those who are looking for clean and better domain names.
Risk of buying penalized domains will now decrease, I just purchased a domain name with 7/16 spam score. I hope to improve it.
Regards
Pulkit
This is really good tool but its not working well right? because i analysed https://www.mucizefikir.com web site and i cant see my updated links. All links are old.
One silly question:
Why a site has different spam scores depending on whether you use www. or not?, i.e.:
www.walksofspain.com --> spam score 0
walksofspain.com --> spam score 7
Thanks
Hey there! Tawny from Moz's Help Team here - not to worry, this isn't a silly question at all! Spam Score is determined by an aggregate of 17 different flags we set up to identify traits that correlate with measured Google penalization, it depends heavily on the backlink profile to a site. The higher the number of flags on a link, the higher the chance that it's spammy.
Since it's entirely possible that some sites will link to the www subdomain of the site, others will link directly to the root domain, walksofspain.com, which could mean that the www subdomain and the main root domain have different backlink profiles, and thus, different Spam Scores.
The flags represent a wide variety of potential signals ranging from content concerns to low authority metrics. Since this is just based on correlation with penalization, rather than causation, the solution isn't necessarily to disavow links from sites with higher Spam Scores. Instead, I'd recommend using it as a guide to help you prioritize your disavowing efforts. Be sure to check out a site's content and its relevance in linking back to you before disavowing. :-)
I hope this helps! Let us know at [email protected] if you have any other questions or if there's anything that needs clarifying!
Hello Rand,
This post is amazing and incredible for me cause I didn't understand about my sites spam position but now according this post I am clear about it and i see that my blog spam score is 6 OMG. Now please describe it or give me a way which help me for reducing this score.
This post is amazing and incredible for me cause I didn't understand about my sites spam position but now according to this post I am clear about it and I see that my blog spam score is 6 OMG. Now please describe it or give me a way which helps me for reducing this score.
Hi, My Website spam score is 5. How Can i Disallow Links ??
Results, Jobs and Education are Such Keywords, i Make Backlinks on.. If you Could help me Get Better Ranking and increasing organic Traffic Startegies ?
That's a great question to ask in our Q&A forum!
I can't understand the "Small proportion of branded links" flag. For example I have "Astronomy" anchor text, which link to a page with Astronomy Games. Because it is a category I have it on almost all pages. If I put more words in the anchor text, it become too many text. For the sake of the design, I do not want to do it. Or should I? I too should have known, that numbers in the domain name is spam, before I get it, now its too late.
Great addition to your suite of software. Why is it that subdomains are the primary focus? A domain which has a spammy subdomain would suffer, right?
Is there a correlation with your website spam score and email spam score?
Glad you like the feature! We chose subdomains since we wanted to capture the case of large sites like "blogspot.com" that have many millions of subdomains, many of which are spam. We haven't looked at email spam, so unfortunately can't answer you second question.
Allright that makes sense. Most of the free blogs out there operate on a subdomain level, so it is easy to setup and use as a spamming site.
But if e.g. https://freshwebexplorer.moz.com is signaled as potential high SPAM, how does this affect the main domain, moz.com in terms of its own score?
A subdomain's spam score won't have an impact on the root domain (as we don't calculate flags on root domains), or on other subdomains hosted on that root domain (we'll calculate those separately).
BTW - one weird thing to think about is that moz.com is technically both a subdomain AND a root domain. We think about it as *.moz.com - the root domain, which has metrics related to every subdomain on the pay-level domain name, and moz.com, the subdomain, with all of the pages that exist behind moz.com/*. The subdomain has the spam score count of flags, the root domain doesn't.
All right, so each subdomain is calculated seperately from the root domain in terms of your spam score, good to know.
I am not quite sure about your thoughts on moz.com both being the domain/subdomain, but are you referring to pages of the root domain and an actual subdomain such as freshwebexplorer.moz.com?
actually i wasn't knowing about this, but your post was clear enough. thanks for the information
Hey,
Thanks for posting up this blog. When I am checking my website on Opensiteexplorer, I am getting a Spam score = 0. What does that mean?
It means that none of the flags we identified were found on that site.
Hi Rand ! This article really help me to measure penalization risk and got lot more knowledge from it.
Thanks for sharing buddy
Hi, Great News for us that now we can find out spam score too with the help of OSE tool. But I have just checked this site kojaxx.com in your tool, it is showing 2 spam score but this site has been penalized by Google. What you say Mr. Fishkin.
Can we make back link on this site?
That's a great example - 2% of sites with only 2 flags were found to be penalized in our data set, so if you reviewed a bucket of 100 sites with 2 flags, you'd expect to find 2 that Google had penalized. Kojaxx fits in that bucket.
In general, I'd recommend against getting links from penalized sites, especially if those links themselves appear to be manipulative or against Google's guidelines.
It asks for credit card info. Can't we just try it without filling card info. ?
Wow, I am really excited about this new feature. Thanks Rand and Moz team.
Tools looking at spam are helpful - but just out the KPI tool at ExtraDigital to view your KPI score.
1st April 2015.
cool!!
That is a great addon to OSE, I am already liking it. But somehow it's showing some of the organic links as potential spam, its in early stages let see how it goes on.
Hello Rand,
Could you tell us if the fact to put blog comments externals links in nofollow will reduce the spam score on your tool ?
I'm speaking about my own blog where comments url are still in dofollow.
By the way, what do you think about the way Google will take this change in term of scoring or ranking (giong from dofollow to nofollow in blog comments)
Nice Feature Added in Moz Tool as Spam Score of Website & Awesome Explanation about new feature of the tool...It will help a lot to find out the site spam score.....Thanks all the moz staffs make such a good tool for SEO Industry... Thanks A Lot Again....
This is a fantastic new addition to Open Site Explorer - definitely helpful in providing more confidence when reviewing the backlinks of competitors to understand where to and where not to target. It's also thrown up some interesting results in terms of sites that are valid and not spam but which are coming back as spam (I think this may be because they're not implementing best practices in some areas).
Hats off to you guys and Dr. Matt Peters, this is great high-level info to see.
Do you foresee the 17 factors being added to over time?
Great tool, thanks!
I really like this addition Rand. I noticed it just today as we're double checking backlink profiles of our sites, and then this blog post came up in google news feed. Only limitation I see is open site explorer only captures a fraction of the backlinks for a client's domain. Can we upload the full list into moz and then have it work through all the links? Some of our sites have tens of thousands of links, it would be really helpful if we could upload it and say pickup a spreadsheet the next morning. I'd also suggest listing the # of links that domain has coming into it on the Spam Analysis, that would provide a more holistic picture when glancing at the spam report and all the links listed. Thanks again for continually innovating at Moz!
Nice. Is there any way we can upload/import a list of URLs to then check their "spamminess"? (and in the meantime, their DA, PA,...)
Not right no, but there should be. We've talked about adding a "bulk data checker" so you can quickly get metrics for a number of sites/pages all at once. I think that would be a great feature.
It would be indeed. I think Netpeak Checker does this (i.e. giving lots of data for a list of URLs), or use to. But it would be great to have that straight into moz
That would be incredible, looking forward to that possible update!
In the meantime, if you have a paid Moz API subscription, you can do what you are asking.
https://moz.com/community/q/spam-score-and-social-contact-data-are-now-in-the-moz-api
We thought we could use one, so we had a team build one. You can check it out here: https://blog.cbtnuggets.com/2015/06/introducing-bulk-backlink-spam-score-checker/. Feel free to use!
This is excellent! Very excited to add this to my monthly link hygiene checklist!
I think a great tool!!
Hopefully soon be for all users, so we can see first-hand operation.
I like to see as more and more tools that make us much work is created.
Happy to have discovered this blog, where every day you learn something new.
Congratulations to Doc Pete on another project coming to fruition as well as others who put in work. As a guy interested in public relations, I'm wondering if this could produce 'hurt feelings' or 'witch hunts' regarding linking patterns or particular domains. A post by Eppie (about three years ago) comes to mind, where he was introducing a tool with good intentions for SEOs. There was a huge debate that day, and some people had strong opinions about an organization with the high reputation of Moz being part of a tool that "labeled" sites/intentions. Of course, I understand this was made for "good," but, again, I'm interested to hear any thoughts. Thanks.
Can never have too much data and intelligence! Glad you are bringing this to market.
Clear rules from google should be great to understand what is spam for them
Just to be clear - these ARE NOT Google's guidelines or rules for spam. These flags merely correlate to sites we've found Google penalize/ban. Don't worry about avoiding, for example, a number in your domain name if you and your customers like it.
I can see this being very useful for disavowing spam links. Although, at the moment, all my links pass the spam test :)
Just make sure to manually review! Plenty of links that have low flags could be spammy and plenty that have high flags could be OK. Like any other metric, it should be helpful to sort and to add to your consideration set, but it can't replace good old-fashioned SEO smarts :-)
That's EXCELLENT!! It's indeed a great feature and I must say it will make our job very very easy.
I just tried it, the details section is working great but I cannot be able to see the "Spam Score" of my main domain. What could be the reason?
Thanks,
May be the case with Moz crawler, it might not have crawled your page yet.
Great news and a good tool.It's good that you agregen, as would be the DA or PA domain to examine whether these links disallow
Really helpful information about the links to our website, which help us to clean up all spam links and help us to monitor and play safe with Google.
Hello, our site does not have a spam score? could you tell me why?
I've been waiting for a tool like this! Nice work Rand and Moz team. I've already started leveraging this tool for link profile audits. Of course you need to do an eye test on the links that are flagged but this tool has been pretty accurate so far.
Moz tools gradually charming.
Wow - this is brilliant! I have wanted to have some insight into this for quite some time. Thanks Rand - you guys keep making my life easier!
It´s an espectacular tool!!! Congratulatios!
That's really great to have an eye on the links and to know the quality of the links.
Really helpful tool for me. First I wish thank moz team for it. Many tools provided my moz help us to reduce our work time. Spam score checker is amazing one. Hope it will provide significant spam signals affecting our project websites in future. Before we need to record all back links and analyse each and every pointing links separately now it much more easily keep our website projects away from spam back links. Once again thank you Rand to publishing the post with better introduction to the new tool. Thanks to Matt also.
Yes I'm also agree with you. This tools is very helpful for me. I'm facing a problem with my domain spam score. I have cheek my domain if I cheek by
www.answersmode.com - spam score = 0 but if I cheek without "www" it shows me
answersmode.com - spam score = 9
How it possible? Will you please tell me?
Thanks
[link removed by editor]
I like to appreciate MOZ's effort !!!
Thank you Dr. Matt Peters for effective big data analysis
& Thank you Rand especially for serving it here!!!
I am not sure about the spam factors used. I agree they are common to spammy looking sites but I simply would not have them on my sites or as clients. Out of the three big tools around I prefer ahref with their new "Near Duplicate" feature that will save me having to shell out for a separate feature for Panda related duplicate content networks.
Great update Rand, It will help allot
Excuse my English.
Excellent work, great tool and interesting contributions in the comments.
They will work hard to make the guys at Google to improve the signs by which to detect spammers.
Your excellent work is also: Hi spammers, these are the 17 factors that you have to hide ;)
I understand that while working signal identification and study of the correlation they have considered other signals its relevance you do not have considered important.
Other signs may include:
Perhaps the Analytics tracking code is the least reliable data because not all webmasters incorporate it but I'm sure that 100% of the spammers do not incorporate.
Congratulations. Great Job
That's a great set of additional ideas - I believe we have considered a number of them, but I'll make sure we look again in the next round of analysis.
Stunning Tool By MOZ! Thanks rand for sharing the spam radar so that we can check what links are Good & Bad in our site. this tool have one more merit that we can easily disavow the spammy links from our site.
Guys, please add Export to PDF feature to the Spam Flags for subdomain page.
Thank you.
PS Great tool BTW!
Hey Slava! Thank you for the suggestion. Can I ask what you'd be using the PDF for? That would definitely help us to understand your use case(s) a little better. Thanks!
PS - I'm the UX designer for OSE in case you're wondering who this random Ben guy is ;)
Woow !! Good functionality to Open Site Explorer'm going to try and ask the already reports of spam links. Thanks for all your help Rand
Hi Mozzers,
It's a great feature in OSE. One thing I could not understand is that your focus is on sub-domains, does www and other sub-somains are equally precise or there is a differece?
Moreover, I would request your team to add features to analyze content quality on websites, it would also help a lot in creating quality content. It would also demotivate spammers who focus on quantity instead of quality.
Hi there! One of the link flags is the ratio of follow to no-follow links. What is considered a "normal range"? Thanks for adding this metric by the way. I've been using it the last few days for sites in my niche and it seems to be pretty accurate.
Oh. Just one thing to add. I do find the graphic rather counter intuitive. It looks like the signal bars for a cell phone which should be better the more bars but this graphic works in reverse. My first time viewing it I thought 1/17 and 1 bar was bad. Just my two cents.
I can't recall exactly what the ratio was - in a way it's not particularly important because these flags aren't indicating problems or things to fix, just features correlated with higher or lower percentages of being penalized/banned. Correlation is NOT causation!
Re: the graphic - I hear you; we talked through a few options and this one tested pretty well with folks. We'll keep listening to feedback, though, so thank you for letting us know.
Really good article. Very useful and simple. Does anybody know if the website is real time analyzed or has the tool a DB?
Another UX suggestion for the Spam Flags Page : Ponder dividing the 17 items into 2 SECTIONS:
One for all the readily Actionable On-Page elements that could be fixed quick (footer/sidebar links, thin content, anchor text, missing contact info, etc.)
The other section for External / Link profile Off-Page issues that are far more complex to address.
INTERESTING TOOL spam score, but you need a subscriptionit is also very intense work to check one by one the potentially dangerous links
@Rand I was surprized to see that one of my blog has been linked from spammy site through anchor text link. Accordingly, it shows spam score 5. So, what should be done from my side to prevent those activity.
Absolutely nothing! Again - the spam flags aren't indications that something's necessarily wrong with your site (or any site where they're triggered). The flags themselves and the counts are merely meant to show the signals that correlated (not causal - just correlation) with sites we saw Google penalize/ban. I'd worry about only fixing the items on your site you think are actually problematic for search engines and/or users.
Hi Randy,
I have checked it for my website Brickseo.in
but didn't get any score. This site is about 1.5 yr old.
Is there anything wrong ?
Be really good if I could upload my current disavow file to you guys so that you can ignore the links I currently tell Google to ignore - I guess this would be a huge strain on resources but would make the tool even better.
But love the tool and glad its something that is part of Moz and not something else I have to pay for.
Nice tool. Will you add also general overview of "good" - "border" - "trouble" amount of links for tested site?
Nah - that was just me messing around in Photoshop. Those aren't the most accurate terms to use, and we'd prefer folks think about the flag counts in relation to the percentage of sites penalized (e.g. 8 flags = ~56% penalized).
I meant to have on top of the page general information about how many links were found as spam or almost spam (or even amount of 17 flags, 16 flags.. whatever). If i have many webs, it gives me fast idea about which one is how good or bad on it comparing to others (as i will see XY links as spam, YZ links as almost spam).
Now it is not possible to see, as there are only 50 links on one page at once. As well, many times i am not interested in links as such, but just want to check link portfolio of domain, so i will see 45 spam links, 23 medium spam and i have fast idea what was site history and approach like. Plus having this info in api, would be reasonable for agencies. By the way, there will be API for this?
Ladislav, Spam Score and Social & Contact data are now in the Moz API! Details here: https://mz.cm/1HlE0qI
Excellent tool, looking forward to using it to review my client's sites and have a better evaluation of links.
How can fix following issue: