Google confirms Penguin recovery requires algorithm refresh

[Robert Charlton]

Barry over at SERoundtable [seroundtable.com...] noted today that Google's John Mueller had confirmed that Penguin recovery is possible, but not until after the algorithm is refreshed and/or data is updated... and that has not happened in quite a while.

John made the statement, quoted below, in a Google Product Forum Help discussion...

Has Google ever definitively stated that it is possible to recover from Penguin?
[productforums.google.com...]

Yes, assuming the issues are resolved in the meantime, with an update of our algorithm or its data, it will no longer be affecting your site.

Seems we've discussed this any number of times, and I believe that John has also confirmed this on one of his Google Hangouts, but I don't have the reference handy. Marie Haynes, who tweeted the announcement, has also previously discussed the issue.

To be emphasized... ranking recovery will happen after disavow only if you also have a sufficient number of high quality links, beyond the links you've disavowed.

[Broadway]

< moved from another location >


Over on SERoundtable today there was a mention about Penguin and how getting released from it will take an algorithm refresh.

In the comments section, someone was discussing WebmasterTool's Disavow Tool.

My site has not had any problems with Penguin penalties, nor has there been any link building done by me or for me of the type that Penguin seems to penalized.

At the same time, I did look through my backlinks and found some really spammy, off-target, and even risky-looking linking to my site.
I thought I would just nip things in the bud and place these domains in my WebmasterTools Disavow file.

I've done this piecemeal over the last 6 months, only adding several domains to the list at a time, as a way of testing things out.

So far, I'm under the impression that things are going just fine.
I've seen no hint of ill effects to my site because possibly I disavowed an important domain.

Now, after reading the article today, I'm wondering: Does anyone have any evidence that Google has ever implemented the use of their Disavow lists?
Am I safe in the sense that google has already used my Disavow list, and in doing so I've caused no ranking problems for my site?
Or, will my Disavow list only be used (for the first time) when Google finally reruns the Penguin algorithm.
(And therefore the appropriateness of the sites I've chosen to Disavow is still a total wildcard. And doing it piecemeal really served no purpose.)

Have I put myself at risk for a fall here?
.

[edited by: Robert_Charlton at 7:31 pm (utc) on Sep 8, 2014]

[Saffron]

I don't know if you're asking the same question as me here Broadway. I was looking at my links yesterday and there's some dodgy looking ones. I haven't done any link building and they are in no way there because of anything I have done. So, I thought about using disavow to get rid of them. But is that a declaration of guilt in Google's eyes?

I don't believe I've ever been hit by Penguin, but want to protect myself. However, am scared if I start removing these links I'll put myself in the line of fire.

[itsjustme2]

Penguin recovery continues to be a confusing topic.

I've had a disallow file of over 2,800 domains up since mid December 2013 (urls taken from Webmaster tools, Moz and Majestic). The disavow files are approximately 90% Chinese spam. It's a case of my domain being on a DMOZ like category template, and the template becomes one of many in the pack uploaded to new domains on Chinese servers. The anchor tag for the link (my site name) is in Chinese characters, and I noticed that my site name in Chinese characters has been the #1 or #1 keyword in Webmaster tools for 5 years (well before Penguin).

By my readings, Google has not updated Penguin since October 2013. So, technically, is it correct to say my site has not recovered from Penguin?

Nonetheless, it does look as if my site recovered from Panda 4. From out of nowhere most of my keywords returned to page 2, although with a 13 year old domain, use to page 1 results, the return to page 2 looks strange, along with the fact that the Panda recovery did nothing to increase my traffic.

If Penguin finally gets run again, would that possibly mean my page 2 results?

It looks as if the continuous adding of this template to Chinese servers means my site will be continuously hit by this spam. In the language of Majestic Trust flow, every time Google picks up one of these spam pages the trust flow and citation flows of 0,0 keep getting added to my site's aggregate trust and citation flows, pulling them constantly down.

On the down side, since I am only able to disavow domains rather than disavow any domains with a link to my site written in the Chinese characters, i.e.,

disavow: anchor text name
disavow: domain

It looks as if I'm stuck in a catch 22. My couple thousand of really good in content URL links acquired over the course of 13 years were what gave my site its SERP strength.

If you saw my link profile, you'd see no forum spam, link buying spam, article directory spam, Squidoo spam, etc. I always received anonymous dofollow links, until the Chinese spam hit.

So, if I started a new domain, I'd have to start with 0 links and lose the 13 years of anonymous dofollow incontent links (votes) I received.

If I do not start a new domain, I'm forever stuck in this awful Chinese spam loop, albeit I can understand why Chinese webmasters might want to put their own version of DMOZ up on their domains.

Forgot to add, I lost 95% of my organic SERPS traffic over the past 3 plus years, and I'm still confused as to why it happened. Panda, Penguin, or Google preference for alternative quality sites (however they might define and operationalize "quality")

[FranticFish]

Have you tried contacting the original data source (you say it's DMOZ-like, but not DMOZ) to get your link there removed, in the hopes that the scraped versions of that page will no longer include your domain?

[Mentat]

I believe that we always knew that we need an Penguin update in order to apply the disavow file.

The problem is that Penguin is so late and this created a lot o problems.
Clearly Google knows there is a problem so the logical conclusion is that the Algo is broken OR it's creating huge collateral problems.

From my point of view, since the past Penguin, the disavow file have grown dramatically, so the amount of processing power can be huge...

We do not know, but I feel that Panda is a drunk animal, with a lot of problems and probably some dependency on Penguin...

So... we are waiting each day for a new Penguin.

Maybe it will never came and they will prefer instead to disable the huge links networks one by one.

[netmeg]

I think the "recovery" part is a little misleading. I'd be amazed if anyone actually recovered all of their previous rankings - not only will their whole link footprint change, but it's been so long between Penguin refreshes that the entire ecosystem may have shifted. I wouldn't be holding my breath on this one.

[Awarn]

I think you can give it any black and white animal name you want. The simple fact is Google is screwed up and they obviously do not have the power to fix it. The way they keep going back and forth on old data sets tells me they don't know how to fix it. I find the fact that page 1 changes so little very strange. Almost like the things are pre selected. The only way this will be fixed is if an outside party brings a negative light on this so Google gets a black eye in the general public. It needs major media attention.

[aristotle]

A lot has happened since the last Penguin refresh on October 4, 2013, more than 11 months ago: Specifically:

-- Many people already affected by Penguin have submitted disavow files and in some cases gotten backlinks removed.

-- Some people may have attempted negative SEO attacks by creating toxic backlinks to competitors' websites.

All of these people have to wait for the next refresh to see if their actions have any effect.

I've been saying for months that Google may be delaying the next refresh because they realized that a lot of sites would be affected by negative SEO unless they made some major revisions to Penguin. So this exceedingly long delay could be due to their attempts to fix it before they do another refresh.

[Zivush]

For me, it doesn't make sense that people who have submitted disavow files will gain any advantage (in a Penguin update) over people who haven't.

[GreyBeard123]

So this exceedingly long delay could be due to their attempts to fix it before they do another refresh.

Or, perhaps they’re busy merging Penguin’s data into the Panda algo

[Awarn]

Has anyone really seen the disavow help when they had an algorithmic penalty instead of a manual penalty?

[itsjustme2]

Have you tried contacting the original data source (you say it's DMOZ-like, but not DMOZ) to get your link there removed, in the hopes that the scraped versions of that page will no longer include your domain?

No. I'm actually more inclined to email President Xi Jinping, in order to explain my take on the state of the global IT industry. It's grounded in a Frankfort School of Philosophy as cultural critique, noting my opinion that we are currently entrenched in a Post-cold war era of the glorification of anarchists, including economic and information anarchists.

Being a Neo-Waltzian and cognizant of the fact of market failure, I'm also of the opinion that agents can have too much power, however the concept of power is defined and operationalized. In that sense, China's move to a more distributive IT policy is understandable, just as a similar policy move by Latin American and European states would be understandable. Reasoning by analogy, one might liken the information industry to the electricity industry. The value of distributive electricity, especially renewable distributive electricity can easily be defended on both cost and security grounds.

The language and logic of the first Secretary of the Treasury, Alexander Hamilton, provides additional justification regarding the the value of distributive policies. After all, he used the infant industry argument to promote the growth of the political economy and national security during colonial days.

While this policy opinion might not be in Google's financial interest, it goes without saying that I'm neither Google's lawyer nor their PR point person.

Regardless of whether one takes a classical realist or Neo-realist approach to power, it comes across as reasonable to suggest that today's glorification of anarchists culture presents challenges, because, as the saying goes, anarchy is problematic.

In closing, I would note that I highly value the team concept inherent in Confucianism, I also value the individuality inherent in Daoism, especially the idea that the Daoist individual need not clutch so tightly to the concept of rights as to be inclined to make a Federal case over every perceived or real harm to self.

From that perspective, I'm also more inclined to remove my domain from the virtual world, and wait until the anarchist dust settles.

[netmeg]

Thank you for that; you made my day.

[Jez123]

I think the "recovery" part is a little misleading. I'd be amazed if anyone actually recovered all of their previous rankings - not only will their whole link footprint change, but it's been so long between Penguin refreshes that the entire ecosystem may have shifted. I wouldn't be holding my breath on this one.

I agree. My site recovered 80% of keywords but some have not and stubbornly remain affected no matter what. I am not complaining. Business has never been better and if my 20% of keywords do get released I might be swamped. Nice position to be in. My link footprint obviously changed but I think that the replacement links gathered over the last couple of years go some way of filling that void.

[Planet13]

According to John Mueller (emphasis mine):

"In theory: If a site is affected by any specific algorithm or its data, and it fixes the issue that led to that situation, then the algorithm and/or its data must be refreshed in order to see those changes."

So is a Penguin "update" an update of the ALGORITHM? Or is it an update of the DATA?

Can penguin DATA be updated between Algorithm updates? Or can it only be updated in conjunction with an algorithm update?

[Planet13]

Another quote from the Gospel of John:

"For example, if a site is strongly affected by a web-spam algorithm, and you resolve all of those web-spam issues and work to make your site fantastic, you're likely to see changes in search even before that algorithm or its data is refreshed. Some of those effects might be directly related to the changes you made (other algorithms finding that your site is really much better), some of them might be more indirect (users loving your updated site and recommending it to others)."

This makes me believe that things like Panda are less of a "filter" (that absolutely prevents a site / pages for ranking) and more like a "weight" to the algorithm.

So you aren't prevented from driving in the race; you just have to carry around a boat anchor which will pretty much prevent you from doing as well as other racers, unless you really improve your engine, brakes, and steering.

[aakk9999]

So is a Penguin "update" an update of the ALGORITHM? Or is it an update of the DATA?

Can penguin DATA be updated between Algorithm updates? Or can it only be updated in conjunction with an algorithm update?

You raised a good point, Planet13. I am guessing that updating algorithm would be changing the logic (or the number or value of parameters) the algo uses. Updating the DATA would be applying the algorithm on a collected data set.

Two may or may not go hand in hand. My guess would be that the data is collected first, algo applied on that data and then depending on results, the algo may be further tuned (or not).

[Robert Charlton]

A fuller quote from what John Mueller posted in the above Google product forums discussion, which suggests that a site that has really begun fixing its spam issues might see a benefit before the data/algorithm refresh....

Direct link to John's new post...
[productforums.google.com...]

In practice, a site is never in a void alone with just a single algorithm. We use over 200 factors in crawling, indexing, and ranking. While there are some cases where a site is strongly affected by a single algorithm, that doesn't mean that it won't see any changes until that algorithm or its data is refreshed. For example, if a site is strongly affected by a web-spam algorithm, and you resolve all of those web-spam issues and work to make your site fantastic, you're likely to see changes in search even before that algorithm or its data is refreshed. Some of those effects might be directly related to the changes you made (other algorithms finding that your site is really much better), some of them might be more indirect (users loving your updated site and recommending it to others).

... (it) helps to keep in mind here is that you shouldn't be focusing on individual factors of individual algorithms, it makes much more sense to focus on your site overall -- cleaning up individual issues, but not assuming that these are the only aspects worth working on.

[edited by: Robert_Charlton at 8:59 pm (utc) on Sep 10, 2014]

[dvduval]

If some of the best search results are penalized, seems to me the penalty is the problem. Maybe Google is too focused on link building instead of indexing quality sites.

[Awarn]

See this actually seems to go along the lines of what I see. It is like having a -10 or -20 penalty. The issue I have is I can fix everything but if Google doesn't update they are penalizing me and not penalizing sites above me that sure look like spammy links. This whole blog comments being valued is total garbage. They could fix things by putting zero weight on blog comments and only penalizing spammy sites using AdSense. It would get them close.

[rish3]

only penalizing spammy sites using AdSense

I've heard more than one person say that. I guess it depends on what niches you have experience in. From my standpoint, the biggest spammers aren't running sites with adsense.

They are running sites with "squeeze pages", selling products/services either directly, or via affiliate links. Usually high margin stuff. Diet, payday loans, designer purses/watches/shoes, etc. The money is better there, and these guys are generally more sophisticated than the spammy adsense guys. They have the churn/burn cycle down pat, and don't care if their site gets penalized...because they have spare sites waiting in the wings.

They are also often the sites that the ads on those spammy MFA sites take you to :)

[particleman]

I managed a domain that was hit hard by penguin when the algorithm originally was implemented. It was hit not because of back links (we thought) but keyword stuffing and over optimization. After revising a considerable amount of content over a year some pages recovered but the site never regained its page 1 position for the main keyword. Interestingly we had a backup domain name (without dashes) that served as a blog for many years. We decided to redirect the main site to this domain name. In a matter of less than a month the site regained its page 1 position and sits at #3 to this day. I want to stress however, a lot of the "long tail" optimization pages have still not returned, so perhaps we escaped some portion of penguin by redirecting the domain, but no others.

[fathom]

Has Google ever definitively stated that it is possible to recover from Penguin?

Yes, assuming the issues are resolved in the meantime, with an update of our algorithm or its data, it will no longer be affecting your site.

Seems we've discussed this any number of times, and I believe that John has also confirmed this on one of his Google Hangouts, but I don't have the reference handy. Marie Haynes, who tweeted the announcement, has also previously discussed the issue.

The problem with identifying what's up is:

1. Identifying recovery is problematic since disavowing unnatural links isn't a guarantee that those links were the correct ones to disavow.
   
2. Recovering from PENGUIN requires not just a 1:1 replacement of links as unnatural links (those that are keyword oriented) and far more powerful (rank inducing-wise) than any natural link that does not target a keyword.
   
3. To rank a domain without unnatural links requires a reconstruction of the website so that those keyword links are indeed 'somewhere' in the internal navigation paths.

Most websites, even if they have cleaned up 100% will never see a recovery due to points 2 & 3, and the fact that acquiring natural links eludes them.

[Hollywood]

Itsjustme2 & EVERYONE else.

Re "Penguin recovery continues to be a confusing topic."

The REAL problem here is that Google, the 1 trillion pound gorilla they truly ARE should be more transparent and ethical. The way Google is going about ALL of this is VERY unprofessional, unethical, misleading, manipulative and unbusinesslike.

Truly, something really should be written, stated, reported and admitted about this 'Penguin' & 'Panda' situation, etc.

If I worked at any large Gov't agency relating to fair practices, etc, I WOULD be doing SOMETHING about the way this is being handled. PERIOD!

Google is massive and due to that ALONE much more should be done to regulate this penalty stance on the entire search market, because people, they control most all of it, we all know this, thus, it IS the Gov't responsibility and Google's responsibility here to explain what ALL of this crap is about, really!

Tired of it!

~HOLLYWOOD

[fathom]

If I worked at any large Gov't agency relating to fair practices, etc, I WOULD be doing SOMETHING about the way this is being handled. PERIOD!

Oh come on.

It's 100% free! How unethical is that?

[EditorialGuy]

The REAL problem here is that Google, the 1 trillion pound gorilla they truly ARE should be more transparent and ethical. The way Google is going about ALL of this is VERY unprofessional, unethical, misleading, manipulative and unbusinesslike.

How so? They're being more transparent than they need to be. (What obligation do they have to even mention Penguin? They'd get a lot less crap from unhappy SEOs and site owners if they just kept their mouths shut instead of talking about algorithms and penalties.)

And don't forget: The Google Search target audience consists of end users, not people who are looking for free organic traffic. In terms of serving users, it's more important to get the next Penguin update out the door when it's ready for prime time than to get a half-baked Penguin update out the door ASAP--regardless of whether further delay is good for site owners and SEOs.

[fathom]

Agree EditorialGuy.

If you are that desperate for Google returns... disallow Google into that domain so it can still gain Bing, Yahoo, Ask and direct referrals and start a dot.net domain (duplicated sister site) for Google alone disallowing Bing, Yahoo, Ask, etc.

...And don't do stupid unnatural links or over optimizing tricks... now you and Google are in perfect sync!

I'm sure if your favorite TV station stopped showing your favorite show but it was playing on FOX... you wouldn't worry about it not being on NBC... you would CHANGE your pattern instant of *****ing about it.

[Hollywood]

LOL ok guys, love it, have fun... it's ok you don't get the underlying problem, wooo, oops. To the rest, have a good weekend!

[fathom]

LOL ok guys, love it, have fun... it's ok you don't get the underlying problem, wooo, oops. To the rest, have a good weekend!

Not sure you understand what the underlying problem was.

If you happened to fake your votes and got devalued, some people claim they had top results for 10 years or more... waiting say 1/4 of that time still means you stole 7.5 years of value from some other poor sap.

That was the underlying problem.

<added>Just a further note: long before Google developed PENGUIN they nailed me 4 times for almost a million dollars in domains.

Once in 2009, this 2010 one was published [productforums.google.com...]

A year later, 2011 Google nailed me in March & then again in April which made those practices unsustainable.

While Matt Cutts noted:

"I just say thanks for this report. We got a lot of useful data from it--in addition to (just recently) taking action on things that violate our quality guidelines."

I also got a lot of useful data from their actions in return.

It all equals out in the wash!
</added>