Why does site with more toxic links rank better?

[onlinesource]

I know nobody here works for Google and only Google can make sense of Google, but I am scratching my head trying to figure this out.

So, I have a shopping cart which uses multiple websites or store fronts such as a .com site that I want to rank in Google.com, a co.uk that I want to rank in Google.ca, a .ca site for Google Canada, etc. So all sites share the same shopping cart, which means they share the same products, same categories, same pages, etc. I do use hreflang tag to point Googlebot to each store when they land on a particular url, but this is what is odd.

So I've run toxic link reports on both my .com and .co.uk sites. The co.uk site consistently ranks very well *knock on wood* in Google.co.uk for top keywords but the .com site is a big mess, as our SERPs bounce between page 5 to 7 on a daily basis. So in the latest toxic reports, the co.uk site has way less links but actually a higher percentage of toxic links. I mean, 19% of the links on the co.uk site are toxic but on .com, it's just 2% after all the disavows i've done. Also .com has 83% of it's links (542 in total) marked as healthy/GOOD whereas co.uk is 46% and just 22 in total. Then I thought, "well, maybe co.uk has LESS links but the good links it has are better!?!" So, the average power trust levels of the 1 compared to 4 for .com. I don't get it.

Is there any correlation between the # of bad links and healthy links or is the problem likely something else? I don't understand this because I was told, lower your toxic links, raise your healthy and you'll get rewarded so that being the case, shouldn't the .com site dominate?

[Robert Charlton]

The co.uk site consistently ranks very well *knock on wood* in Google.co.uk for top keywords but the .com site is a big mess

Top of my head answer to a question that's probably impossible to answer without being Google... but what immediately comes to mind is that a co.uk site on Google.co.uk is probably facing significantly less competition than a .com site on Google.com would be facing. The .co.uk is a ccTLD, which means that you're limiting competition to a country (in this case, the UK), whereas .com is competing all English language sites worldwide.

Additionally, not all links are equally important, so it's unlikely that any measure based simply on a count of good links vs bad links is going to be very meaningful.

[Andy Langton]

As Robert says, you certainly can't rely on numbers.

You should regard the data from link tools as a time-saver, rather than a solution. You need to interpret it for it to mean anything as far as ranking goes. With more aggressive tools, I've seen a lot of links listed that are very unlikely to cause ranking problems, e.g.

- Scrapers and "stats" websites (could be thousands of these)
- Non-canonical versions of a link (e.g. www.example.com/link, www.example.com/link?parameter, example.com/link reported as three links, rather than one). For some sites this could potentially be hundreds or even thousands of "links"

You could also try (for instance) putting in sites that perform really well, and you'll find that unless they have very few links overall they will be listed as having risky links.

Put another way - should you trust the tool or what you're seeing in search results?

[martinibuster]

There is no such thing as a toxic link. What you were told about reducing "toxic" links was a myth. The fact is that successful websites have always attracted strange and off topic links. They're what I call Barnacle Links. It's entirely normal and the search engines know to discount them.

The only links anyone should be disavowing are crap links you are responsible for creating yourself.

[goodroi]

@mb I respectfully disagree with you. Toxic links definitely exist and you even say there are "crap links" that should be disavowed. I think there are very varied definitions of what is or is not a toxic link but I think we can agree there are certain links that do more harm than good.

Personally I think that the automated toxic link detectors are not good and I would not use most of them. They miss bad links aka toxic links and they can also accidentally flag good links as toxic links. I would run a link report using Majestic, Ahrefs & OSE. Then I would cross reference the data from these different sources and manually review the suspected links.

I also think in general people are too concerned (some are even paranoid) about bad links aka toxic links aka negative links and are not concerned enough with developing quality links. It doesn't matter if you remove the poison or are paranoid about poison if you don't add positive ranking factors.

[martinibuster]

I also think in general people are too concerned (some are even paranoid) about bad links aka toxic links aka negative links and are not concerned enough with developing quality links.

I agree with you Goodroi, 100%. You started out disagreeing with me but ended up agreeing with my post. I confess I left out quite a bit from my earlier post, but I guess it needs to be there. So here it is.

1. There is no such thing as a toxic link.
That is not terminology associated with information retrieval research at universities or at any search engine. Search engineers do not use that phrase (I've only ever seen it used in the context of answering a question in which an SEO used the phrase). The phrase and concept of "toxic links" is not associated with any published or patented search engine algorithms. The phrase and concept of Toxic Links has no relevance to how search algorithms were created. The phrase "toxic links" has nothing to do with how search algorithms rank sites. The phrase and the concept it describes is literally not in the literature, not in the science of information retrieval.

So where did this thing "toxic links" come from? Toxic links is a phrase invented by the SEO industry.

2. What you were told about reducing "toxic" links was a myth.
Removing links is easier than building links. The SEO industry realized his and has shamelessly capitalized on the easy money, making millions of dollars pushing the idea that your links are harming you. The phrase "toxic links" was invented by the SEO industry for the benefit of the SEO industry.

When Penguin first came out several years ago there was a real need for some publishers who had experienced a Penguin penalty to remove low quality links that they themselves were responsible for. That was the first wave of Penguin penalties. After that the SEO industry invented the phrase Toxic Links to make easy money off the FUD. The Toxic Link terminology, was an invention of the SEO industry, intended to create a fear-based demand for a service you may not need.

3. The fact is that successful websites have always attracted strange and off topic links.
They're what I call Barnacle Links. All successful sites attract them. I'm fairly certain that those "barnacle sites" link out because they believe that linking to a quality site will help their site avoid being caught as a spam site. Attracting barnacles is entirely normal. Information retrieval researchers are aware of these kinds of links and make allowances for them. The search engines are aware of them and discount them.

4. The only links anyone should be disavowing are crap links you are responsible for creating yourself.
As GoodRoi points out, you would be best served by building, not tearing down. To this I would add that the best way to do that is by cultivating the circumstances that result in people sharing, which includes social sharing, word of mouth and links.

I wrote an article in SearchEngineJournal.com where I pointed out that researchers discovered that people tended to share things that had given them a positive experience. My original insight that I shared in that article (and I will share here) is that people do not share links, they share experiences. In other words, one of the secrets of building a successful site is cultivating positive experiences. Think of all the reasons why a blog post or an infographic go viral. Those aren't links people are sharing on the web, those are experiences that are being shared. This is part of what I call User Experience Marketing. It's a focus on building more sales.

[travelin cat]

We were hired by an agency to try and find out why their client was appearing on page 3 of the serps. The client was a medium size well known brand for their industry. As our first step, we found hundreds of "questionable" links and disavowed them in Search Console. Within 1 week the client was on the first page. We did nothing else for them at that time.

Coincidence? Perhaps. But the timing suggests otherwise.

[martinibuster]

Within 1 week...

One single week, seven days, is an extraordinarily fast time.

We're friends and we've literally broken bread together on many occasions and I can vouch that you are telling the truth. But... are you certain it was not a coincidence?

[travelin cat]

Really cannot say with certainty as we were subcontractors on this case. The agency that hired us said they did nothing else and neither did we.

[travelin cat]

I should add that this was in July 2014.

[onlinesource]

Top of my head answer to a question that's probably impossible to answer without being Google... but what immediately comes to mind is that a co.uk site on Google.co.uk is probably facing significantly less competition than a .com site on Google.com would be facing. The .co.uk is a ccTLD, which means that you're limiting competition to a country (in this case, the UK), whereas .com is competing all English language sites worldwide.

I agree that it isn't fair to compare the rankings at Google.com to Google.co.uk since yes, they are different search engines with different competitions. After reading what you wrote, I thought I should address this from different angle.

So last night, I asked a competitor of mine who operates a similar site to my .com store to run the same detox report on his site. Fortunately for me, we're close so he was happy to help me out and give me his spreadsheets from Google Webmaster Tools and everything I needed. So he did do a disavow himself about a year and a half ago after the Penguin updates hit him but he has been ranking again on the first page for top SERPs for almost six months now with Google.com. Again, his site targets the same keywords as mine, so I think it's a fair comparison. Also his domain is roughly the same age as mine, give it take a year or two, if that matters but still we're looking at domains that are 6+ years old.

Again just to recap, so my .com site according to the most recent detox report (I actually did another one after my post) is now showing:

0.9% toxic links
16% average threat links
83% are good links

63% in total are reported as disavowed.

So together we ran a report on his site and again his site is ranked #3 in Google for the #1 keyword that Google places me on page 5/7 for. This is what his report shows.

55% of his links are toxic
34% are average
just 11% are good

In total he's disavowed 2%

The ONLY difference that gives me pause is the detox percentage. How important is this? Otherwise, 83% healthy and 2% toxic is much better and those numbers should give me the edge over sites like his, correct? Again, I'm not saying I deserve to be #1 in Google but on page 5 to 7? I'm really baffled by this.

The other thing to consider is LRT Power*Trust for our sites.

My domain has a LRT Power of 3 and LRT Trust of 2
His domain has a LRT Power of 3 and Trust of 1

So again, same LRT Power and a slightly better Trust goes to me. I don't 100% understand the whole LRT thing but I assume it's a lot like domain authority? If LRT reports hold true vale, I should be doing better than I am with the numbers I have.

At the end of the day, I know that only Google understands it's own formula but just wondering if others have experienced this same headaches and found the problem goes much beyond linking? I'm just running out of things to rule out because if the co.uk site ranks, than I can't assume Google hates my content. And if another site out ranks me with worse toxic percentages, then where do I go from here? Thanks for any feedback.

[Andy Langton]

0.9% toxic links
16% average threat links
83% are good links

63% in total are reported as disavowed.

So, even by your tool's evaluation, you have disavowed good links?

The ONLY difference that gives me pause is the detox percentage. How important is this?

It's not important at all - it's possibly a clue as to something you might look at. The main thing it demonstrates for certain is that the tool is not matching the reality of how Google ranks pages. You cannot rely on automated tools at all. I have more SEO tool subscriptions than I know what to do with, but I would never rely on their conclusions - I use them for data, not analysis.

[Shai]

I cant stand all those automated link cleaning programs. I would certainly not trust them in deciding what a good/bad link is. We carry out link clean-ups for penalty recoveries (manual and algorithmic) and ALWAYS amalgamate link reports from the normal 5 sources (OSE, Ahrefs, Majestic and Google SC), remove duplicates, and MANUALLY go through each and every link and check using our own guidelines.

[onlinesource]

@Andy "So, even by your tool's evaluation, you have disavowed good links?"

How do you know that I have disavowed good links? I see that this latest report shows that 63% of my .com links have been disavowed, but does that percentage go towards the TOXIC, AVERAGE AND LOW THREAT list or a separate list of links?

I don't understand how any of the detox links would be good links since I ask the program to show me TOXIC links and then disavowed those TOXIC links. Unless a link was once considered toxic, got disavowed by me and then turned good.

I guess I'm confused. So, is this saying that 63% of ALL MY LINKS were disavowed? If that is the case, do I need more healthy non-disavowed links to lower that disavow percentage? Also is there a good disavowed percentage?

[Andy Langton]

How do you know that I have disavowed good links?

I just assumed this based on your wording, "63% in total are reported as disavowed". If this is direct output from the tool, then you would need to check the documentation to verify what it means. It would be a big concern if you're unsure what you've disavowed or have not. I have used the tool you mention, but it's not something I use routinely, or currently have access to, so I can't comment on these particular definitions.

is there a good disavowed percentage?

0% is a good disavowed percentage. If you've created dodgy links or got someone to do this for you, then whatever percentage of your total links that represents is likely the ideal percentage of disavow.

To be blunt, it sounds like you are relying on a particular tool to:

- Match how Google evaluates links (no tool does this)
- Manage processes like disavow (this is a bad idea)

It's very unlikely that this is going to work. To reiterate, a "toxic link" does not relate to Google's algorithm in any way. It is a "best guess" by an automated tool at what links Google might object to. If any of those guesses are wrong, the data is wrong. If the process of bulk disavow is wrong, it doesn't matter whether the data is right or wrong.

Your competitor has (apparently) completely disobeyed this tool and only disavowed 2% of links (toxic or otherwise). And yet they rank better than you. What does this tell you?

Finally, percentages are not necessarily helpful. Does your competitor have 10,000 links and you have 1,000 - do the percentages matter? Probably not.

[onlinesource]

Your competitor has (apparently) completely disobeyed this tool and only disavowed 2% of links (toxic or otherwise). And yet they rank better than you. What does this tell you?

I get what you'r saying. Your saying that my competitor only detoxing 2% of his links proves that it's best to detox less links because obviously he only detoxed 2% and Google still loves him. True, you could make that argument. You also make the argument that he had less toxic links that needed detoxing. :) The fact is, if I have 1,000 toxic links and need to disavow all 1,000 links, that's still 1,000 detoxed links and a much higher percentage than his. At this point I don't know what to do.

This is my story. An employee who I put in charge of SEO years ago, hired a company to build links for us years ago. That company used many blackhat SEO methods and eventually Penguin punished us for it. For years we rode a wave and ranked great, so I never question that was going on but obviously looking back now we were slowly building our demise. I've been trying to rebuild this website from the ground up and the only way I know how is to remove the bad links and add good ones. Even it my data shows that 1000 links are toxic, they all need to go because they aren't helping me rank. Sure I need to then build healthy links in place of the bad links but the percentage will never be 0%. That's impossible.

Apparently, I can resubmit my list of toxic links to Google, removing certain ones from the list and then un-disavowing them. I heard the process can take weeks but I can low that percentage total of detoxed links assuming they were not bad after all. The issue is most are bad and 99% are coming from poorly managed business directories where nobody answer them email; and disavowing was my only choice. If i could ask them to remove it, sure that would make more sense but that isn't the reality.

[Walt Hartwell]

I probably have a different viewpoint than most, but if 99% of your links are coming from poorly managed business directories, I wouldn't disavow those links.
They surely will have very little value individually, and are most likely not considered of importance by search engines. But it is still 99%.

Disavow where there are links to your site in proximity to links for high risk sectors(pr0n, poker, pills)
Disavow where sites linking to you are unbearably bad and are using your preferred key phrases.

Other than that, let the search engines figure it out.

[tangor]

That company used many blackhat SEO methods and eventually Penguin punished us for it.

And google has a LONG MEMORY and NEVER FORGIVES, or FORGETS.

My link strategy is limited to what I link out to. I do not ask for or even seek links from others. PERIOD. IF a site links to me, thanks. If a bad site links to me, thanks also (but I won't disavow it either). I keep my link strategy limited to what I can control and don't worry about all that other krap out there.

Automated tools? Automated fools. (Not meant as an insult to anyone! Just don't believe in them!) Broken link checkers, however, are valuable (and once again, from MY SIDE not the OUTSIDE).

I'll just close with a comment that once black hat is a stain that cannot be removed. Might diminish it a little (hence page 5/7) but never again page 1. This is empirical knowledge from my own work trying to help previous bad seo (black hat and other questionable tactics) try to recover from the last to g animals which start with "p" (as in punished). NONE have ever regained their former glory, though several managed to make it back to page 2 over a VERY long period of time.

I agree with the general comments there is no such thing as a "toxic" link UNLESS you did it yourself.

[onlinesource]

NONE have ever regained their former glory, though several managed to make it back to page 2 over a VERY long period of time.

NONE seems like a strong word, but there may be some truth to that. I read an interesting article years ago where Matt Cutt said, sometimes it's better to start over.

Obviously my friend's site was punished too. Fact is, he had to disavow and he did disavow after falling off of Google. He recovered, so a recovery CAN happen. I've also known others that had to disavow bad links and recovered from it. I hate to say NEVER but there may be some truth in the fact that sometimes it's worth taking a match to the site and walking away.

I'm starting to think maybe that's a smart idea. I have a few domains that I own that operate for internal company use but the domains they use have good age on them, so maybe they should be reapplied as my actual company's website. This is something that I might look into. Not sure if I want to invest in all of the international tld's for each store but maybe I would? It's lot to consider and executing a plan is not going to be easy but nothing worth having is. :)

Probably will speak to some advisers and consider this if they think it's a good idea. I wonder if I should just 302 redirect the old site to my new site (not 301 because i don't want full link credit) but also don't want to loose that traffic.

[tangor]

If you are going to start over, start over. No redirects, no attempt to "save" links... (Meanwhile, 302 "moved temporarily" is not a response that will build trust into a redirect!)

In my comments above the sites did recover, but never saw page 1 again. I will qualify that my NEVER and NONE are limited to those sixteen clients I assisted. None of them ever (and to this day) returned to page 1 and, in fact, a few have virtually disappeared (page 100+).

Old bad acts are not forgotten. Nothing personal, of course, all machine code based on trillions of urls matched, cross-matched and matched again. No human involved. And, if you think about it, no human to plead your case you changed your ways and walk the straight and narrow.

Self-inflicted and never forgotten by others.

If you want a new start, start new. Squeaky Clean. Go from there.

I'd like to report that one of the sixteen did that ... but none were brave enough to give up any bit of "link juice" and "pr" to try again, from a new domain, from a different registrar, with a different business address/profile. But I know two who did that very thing and one failed and the other prevailed. If talking percentages then, to my knowledge, that would be 50-50.

[onlinesource]

I'll just say that I think the whole thing is a load of $#!& because what stop somebody from interesting hundreds in back links for a competitor and destroying their respect with Google? I think the whole thing is absolutely religious but I don't run Google and the argue is one for another day.

I am thinking about taking a match to this current domain and starting over! So my initial thought was to go the 302 route and not loose the business from the .com site but also stop any association with the bad links to my old domain. Clearly that is not the way to go.

The problem for me is that the .com site ranks in Yahoo for top keywords and the co.uk ranks in Google.co.uk and the .ca site ranks in Google.ca and the .in site ranks in Google.in. It's just the .com site at Google.com that can't move up! Maybe since these domains pull traffic from valuable search engines, I should leave everything the way it is and get what I can out of it.

I guess at that point, I should invest in a NEW domain, NEW site, NEW everything. All new content, new products, etc. No connection between this new site and the old site.

My last question is, if I have healthy links to my old site where I can request that the domain url to be changed from my old site to my new site, should I? In other words, should I try to steal whatever healthy link juice I can from my old domain?

[tangor]

In other words, should I try to steal whatever healthy link juice I can from my old domain?

How healthy is it? How can you determine if the move was beneficial if you drag all the old to the new? Redirects by definition are: OLD has moved to NEW so KEEP all the other....

If you don't have a pistol or know where your foot is located this might be something to do. Starting Over generally means Fish Or Cut Bait.

[Walt Hartwell]

I'll just close with a comment that once black hat is a stain that cannot be removed. Might diminish it a little (hence page 5/7) but never again page 1. This is empirical knowledge from my own work trying to help previous bad seo (black hat and other questionable tactics) try to recover from the last to g animals which start with "p" (as in punished). NONE have ever regained their former glory,

With all respect, I'd consider that anecdotal.

Many people on this site consider shades of gray to be "oh so bad black hat". If you stop thinking about hats and start thinking about users and what those users prefer, it's much easier to determine what information you should present. If users click through to "You wouldn't believe what...." links, I have no problem providing those links.
It's the internet, it's not church/temple/mosque/what ever works.

[Nutterum]

Google is well aware that sooner rather than later a website will attract crap links and they usually just ignore them when determining the PageRank and other factors to rank your site. Penguin comes when you went in and built 20,000 PBN, Indian e-commerce, and .ru links to game the system.

Now there are websites who rank well even with those. That is because regardless of what many people believe, Google re-evaluates the "shards" your website is in (read it as topical niche) several times per year. So, you will see a very crappy site funneling traffic for a few months, only to disappear there after. Because black-hatters know this, they abuse the system by making churn-and-burn websites, so long as those bring in more money than the time and effort to build+maintain+opportunity cost(like actually getting a job as an SEO..).

It is how the Google world works. Get used to it.

[FranticFish]

...there is no such thing as a "toxic" link UNLESS you did it yourself

That can't be right. It's not who is responsible for a link (you, an employee, a contractor, a friend, a competitor, or someone unconnected) or what the intent is - good, bad, indifferent - it's what sort of link it is.

[onlinesource]

That can't be right. It's not who is responsible for a link (you, an employee, a contractor, a friend, a competitor, or someone unconnected) or what the intent is - good, bad, indifferent - it's what sort of link it is.

Thank you for saying this! It's funny how anybody can start up a blog and have an opinion on the internet, but that doesn't mean that not all of the information coming from blogs are 100% accurate and verified! Yet we somehow trust linking as a whole. I can think of many industries that would pay $29.95 for 3,000 toxic back links if the thought those bad links would destroy their competition.

[onlinesource]

@ Tangor

How healthy is it? How can you determine if the move was beneficial if you drag all the old to the new? Redirects by definition are: OLD has moved to NEW so KEEP all the other....

The links I would want to change from my old domain to my new domain, are all healthy. At least according to that linkdetox.com site where I've scanned all my sites. These links that detox site considers to solid links or GOOD/LOW RISK links. Most are coming from healthy blog articles were the domain has good age, good domain authority, etc. In them, the blog author mentions our site in a link. I can easily write them, ask them to swap out the old domain in the link to my new site which they should do. Obviously there are only a handful of links I want to switch out because a lot of the old links are toxic and disavowed and ones I have no desire/want for.

[Nutterum]

I never ever used those detox websites. For me they are scams more than actual help. Unless the SEO audit I make for a client involves in to siphoning through 100k links I will not resort to a link quality aggregation service. What is more, I usually have a devilish smile on my face when I see a small local business (think roofing) with 10k backlinks. Seriously do you think Google is that dumb as to not deprecate 90% of those links pointing to your site?

@onlinesource - don't make your job harder. The hours you waste getting paranoid over each and every link will be better spent on actually providing insights to authority content creation that can reach some good long tail keywords and on ways to promote the business so you can actually get some organic links the correct way. I am sorry if I sound a bit harsh, but this overemphasis on something that will have most likely negligible effect is baffling to me.

[Storiale]

To add to one comment about "toxic" links being ones you created yourself like a link network.... if your site is being scraped or was scraped and therefore leads to hundreds or thousands of links because of scraping, you should disavow those as well.

I also agree about the online competition in California and UK are likely different, therefore less links are needed for UK - especially if using hreflang

[supafresh]

@MartiniBuster - Its hard to think google created a tool to placate webmasters. That was 4 years ago so they may have changed a few times since then.
To your point I have seen lots of crap links flow in with no penalties.