Penguin 2.0 is upon us - May 22, 2013

[viral]

Matt has announced Penguin 2.0 (Penguin 4). Either way it is out there and affecting.

Is anyone noticing much movement in the serps? I personally haven't seen much flux but Mozcast seems to be feeling something.

[mattcutts.com...]

We started rolling out the next generation of the Penguin webspam algorithm this afternoon (May 22, 2013), and the rollout is now complete. About 2.3% of English-US queries are affected to the degree that a regular user might notice. The change has also finished rolling out for other languages world-wide. The scope of Penguin varies by language, e.g. languages with more webspam will see more impact.

This is the fourth Penguin-related launch Google has done, but because this is an updated algorithm (not just a data refresh), we’ve been referring to this change as Penguin 2.0 internally. For more information on what SEOs should expect in the coming months, see the video that we recently released.

[edited by: Brett_Tabke at 12:12 pm (utc) on May 23, 2013]
[edit reason] added quote [/edit]

[tedster]

@tedster is there any info on how often this new penguin internal engine will update?

None that I've seen.

I assume that Penguin can be re-run, or some of the details that are used by Penguin can be updated without the essential code itself (the "engine") being changed. Otherwise why differentiate this new version from previous updates?

[ZydoSEO]

They could have differentiated this version because the algorithm itself had a major rewrite. Cutts did say that that the previous version "looked" at the home page (while it could still have "affected" the entire site) and that the new version "looks" at inner pages. That would likely require a major overhaul resulting in a major release rather than a minor update or data refresh.

I agree that there were likely not just algorithmic changes, but also infrastructural changes to allow them to roll out updates in real-time and more frequently without requiring days to roll out. This update rolled out in a matter of a few hours. If I recall correctly, previous updates rolled out over several days.

Search Metrics did post some interesting data on their blog about which sites they are seeing as most affected by the update... #*$! sites, gaming sites, and even some major brands like Salvation Army and Dish.

[tedster]

I'm assuming that Panda and Penguin are both types of Artificial neural networks [en.wikipedia.org] and they have some kind of self-learning component. There are many reasons to assume that, based on Google's public statements. In fact, I'll bet Bing also has some such component and possibly Yandex, too.

That's the key reason I say we cannot reverse engineer the thing. Unless we can work with the same data set (or a substantial portion of it) the job is computationally impossible.

That doesn't mean we can't understand something about Penguin, especially if we share observations with precision. I strongly believe we can do something, just not accurate reverse engineering the way we could back in the early days of search engines.

[Whitey]

the feedback I'm getting from a range of well established sites, sometimes brands, mostly all sites [ from a large sample ] have been affected.

Things are stabilizing into OK positions for most folks in that range now. Not sure about the one's I watch closely. Requires more analysis.

Search Metrics did post some interesting data on their blog about which sites they are seeing as most affected by the update... #*$! sites, gaming sites, and even some major brands like Salvation Army and Dish.

[blog.searchmetrics.com...] - probably a little early, given the reported flux. Not sure.

understand something about Penguin, especially if we share observations with precision

I'm curious as to how Penguin 1.0 fits into the plan of Penguin 2.0. Was it insufficiently engineered, or was it part of a strategic multi tiered approach which can provide a clue about how it operates?

[fathom]

I'm curious as to how Penguin 1.0 fits into the plan of Penguin 2.0. Was it insufficiently engineered, or was it part of a strategic multi tiered approach which can provide a clue about how it operates?

It simply outlived its usefulness much like Google Checkout is folding up now that Google Wallet is stabilized.

[ibmethatswhoib2]

Remember it takes days to roll out throughout her data centers. I thought I was good, got hit from page 1 10 keywords to page 2 for all. Quality site, no spam, not sure what it is.

[ibmethatswhoib2]

I meant I thought I was good yesterday and just saw it dropped today.

[diberry]

I don't understand this idea that Penguin 1.0 just looked at the home page. I hardly had any links to my homepage, and the pages it affected were not the homepage. What am I not understanding?

[fathom]

I don't understand this idea that Penguin 1.0 just looked at the home page. I hardly had any links to my homepage, and the pages it affected were not the homepage. What am I not understanding?

Well "hardly" does not imply you had none.

So maybe the "hardly links" you had actually had the PENGUIN problem on the resident domain and that translated to less link juice to your domain... and that would have a similar impact.

[Whitey]

It simply outlived its usefulness

Sure ... but why? e.g. Was the platform subsequently found out to be unable to scale, was it intended as a stage 1 project/algo, did it hit major "unforseen's" , etc etc

It's a tough question, but some perceptions and perspective of thought about what Google faced in it's task might lever open some answers. [I'm not expecting much, but some folks might have some ideas to go deeper].

[fathom]

Why go left instead of right is an equally tough question if you don't ever have the context of what precisely is to the left and the difference right presents.

[Wilburforce]

I don't understand this idea that Penguin 1.0 just looked at the home page.

I agree with what atlrus said yesterday (I posted to the same effect last year): Penguin 1 was both page- and term-specific.

I have one specific internal (fourth-level) page that was obviously affected for its subject-term. It has a large number of unsolicited backlinks using the term as anchor-text, and I am as sure as I can be that this is the reason for its disappearance from Google SERPS. It was on Google page 1, and is still on page 1 of all other engines.

[Whitey]

@fathom - I presume they had the best of plans with a vision at the outset. What do you think?

[fathom]

@fathom - I presume they had the best of plans with a vision at the outset. What do you think?

For the amount of revenue everyone makes off of FREE I expect they only want overly compelling results at the top and not just webmasters that figured out how to save a buck.

If you don't wish to make "a GREAT WEBSITE," there is always Adwords! :) (or Google shopping.)

[ColourOfSpring]

After 24 hours, my observations:-

- losses on some keywords - losing 2 to 3 positions on first page, yet traffic UP about 10-15% across a lot of my sites - likely wins on longer tail keywords (haven't confirmed yet). No clients complaining at all yesterday.

- domain clustering not fixed yet

- an SEO company I know that got TOTALLY obsessive with cleaning up their backlink profiles got hit pretty badly by Penguin 2.0. This is a company that I know well - speak to on the phone regularly. They tell me they disavow and also have spent a LOT of efforts cleaning up their backlink profiles. Anyway, I was always interested to see how they'd do with Penguin 2.0 - on 3 sites they actually run themselves (not their client sites), all lost 4 to 6 positions on many keywords - meaning heavy demotion on first page, or relegation to second page.

- one site that sold links up front and centre on their home page and also had a spammy backlink profile got hit by Penguin 2.0 (and survived Penguin 1.0 somehow) - a great example of Penguin 2.0 hitting a target square on and killing it - Google do get it right sometimes. This site occupied no.1 spot on a popular 1 keyword phrase. Not inside the top 100 now.

OVERALL: no real damage done, some gains made.

[fathom]

I agree with what atlrus said yesterday (I posted to the same effect last year): Penguin 1 was both page- and term-specific.

I have one specific internal (fourth-level) page that was obviously affected for its subject-term. It has a large number of unsolicited backlinks using the term as anchor-text, and I am as sure as I can be that this is the reason for its disappearance from Google SERPS. It was on Google page 1, and is still on page 1 of all other engines.

But you're assuming that only link recipients were the target(s) of PENGUIN 1.0.

Let's assume for the moment Matt Cutts is correct that PENGUIN 1.0 only targeted the homepage of domains... how can your fourth level page be impacted?

In a link network that uses say Wordpress all the heavy link juice links to you are conveniently on the homepage (as a default setting even if it paginate to page#2 or page#10) thus PENGUIN naturally deals with your sub-page from the other side while still only targeting the homepage.

We don't discuss it but I can't fathom why Google would devalue link recipients and ignore link generators (or maybe they cover both ends and domain owners that offer links are reluctant to discuss the selling of links or don't notice the issue for what it is).

[Whitey]

I know that got TOTALLY obsessive with cleaning up their backlink profiles got hit pretty badly by Penguin 2.0

@ColourOfSpring - interesting thread here, with a small number of response saying something similar. [webmasterworld.com...]

Looks like a lack of good links can also play into this as the counterbalance. What do you think?

[muzza64]

Looks like a lack of good links can also play into this as the counterbalance. What do you think?

I'd agree with that. A few sites I monitor that have been doing well and don't do link building (let it happen naturally) have all lost significant amounts of traffic with this update.

I wonder if they've given good links more power as opposed to punishing bad links (or perhaps a bit of both). So sites with weaker backlink profiles could have lost out without actually doing anything wrong.

[ColourOfSpring]

Looks like a lack of good links can also play into this as the counterbalance. What do you think?

Whitey, it's probably the case. They were hoping for an "attaboy" from Google in regards to their "cleanliness" of their link profiles and they put a lot of stock (and hope) that Penguin 2.0 would reward them. Clearly it didn't. To be 100% fair though, none of my sites that got hit by Penguin 1.0 recovered after Penguin 2.0 - if anything they lost a few more spots. I still think Google are being very harsh in not giving sites a clear recovery path. There's a lot of very good sites out there that searchers are denied.

[Wilburforce]

Looks like a lack of good links can also play into this as the counterbalance. What do you think?

I don't think so (at least, not in my sector): for Key Term none of the top ten results are above PR3 (the #2 spot is PR1!), and running backlink checks on them shows a backlink profile consistent with their PR value.

[aristotle]

- an SEO company I know that got TOTALLY obsessive with cleaning up their backlink profiles got hit pretty badly by Penguin 2.0. This is a company that I know well - speak to on the phone regularly. They tell me they disavow and also have spent a LOT of efforts cleaning up their backlink profiles. Anyway, I was always interested to see how they'd do with Penguin 2.0 - on 3 sites they actually run themselves (not their client sites), all lost 4 to 6 positions on many keywords - meaning heavy demotion on first page, or relegation to second page.

When you remove and/or disavow backlinks, then the site loses the rankings boost that those backlinks had previously provided. This could explain the drops of 4 to 6 positions on many keywords.

[SEchecker]

Yesterday traffic was 50% down same for today... I duno what i should think of that... i c a punch of new keyword sets are assigned in GWMT with 0 impressions... Penguin was supposed to push us up, now we 50% down that’s a disaster :-))))

[Whitey]

@SEchecker - why did you expect it to go up for you ; what did you do ?

[muzza64]

I don't think so (at least, not in my sector): for Key Term none of the top ten results are above PR3 (the #2 spot is PR1!), and running backlink checks on them shows a backlink profile consistent with their PR value.

Wilb, how did those results change due to this Penguin update though? I mean, if the PR1 site was top but is now second and has been replaced by a site with a stronger link profile, that says one thing. But if the PR1 site came straight into the top 10 in second position yesterday above lots of sites with stronger backlinks, that says something else.

[Wilburforce]

Wilb, how did those results change due to this Penguin update though?

Not much. The only major change is the PR3 site with the strongest (but spammy) backlink profile has now vanished completely. That site knocked my site off the #1 spot pre-penguin, and has held on throughout.

My site vanished with the first Penguin, and in the last year most of page 1 followed it. A couple of weeks ago there were three ex-page 1 sites including mine on the same results page (from memory it was page 46), with several dotted about in places like page 38. Google seem to be modelling their behaviour on Henry VIII.

I'm wondering, however, whether some Panda-Penguin interaction is involved. The site mentioned above follwed a very similar trajectory to mine: rock-solid at #1 for quite a long time, then a gradual slide to the lower-end of page 1, then gone. Possibly some page- or site-based Panda threshold is taken into account when Penguin runs.

[SEchecker]

@ Whitey

Since 1st penguin that was hitting us: We deleted "bad links" made by seo consultants, sent reconsideration requests, used disavow tool, played by G rules, all natural! In this last year we collected much natural links what should actually push us but now we are 50% down. we have 3k sites indexed and sites in questions by Gs penguin where maybe 3 single sites beside the home page... so the other 2.99k sites are all natural linked... and actually we get the most traffic by those sites..now we are 50% down...

[Whitey]

@SEchecker - can you clarify "3k sites indexed". Do you mean 3 thousand sites?

If so, how can you manage these with any quality?

[SEchecker]

3k = 3000

"If so, how can you manage these with any quality?"

I don´t see a connection with Penguin > Quality

Anyway all our sites are unique and quallity content ...

[Whitey]

@SEchecker - that's a big task to manage so many, with QA, not something that folks I know could achieve easily. But if it works for you, good luck. Is there any interlinking that could cause problems.

[taberstruths]

3,000 websites! Yikes! I have a hard enough time managing 5. More power to you brother!