Penguin: Core, realtime and updated today

[Shaddows]

[webmasters.googleblog.com...]

Penguin is now real-time. Historically, the list of sites affected by Penguin was periodically refreshed at the same time. Once a webmaster considerably improved their site and its presence on the internet, many of Google's algorithms would take that into consideration very fast, but others, like Penguin, needed to be refreshed. With this change, Penguin's data is refreshed in real time, so changes will be visible much faster, typically taking effect shortly after we recrawl and reindex a page. It also means we're not going to comment on future refreshes.

[Ebuzz]

I've not seen ANY change in my Penguin hit sites. It might as well not even be announced. Absolutely no change. I don't even see any changes in the niches that I am involved in. Competitor sites, my sites, everything is the same.

I have a site that was hit by Penguin 2012, and I managed to improve its position after a number of disavows in the past. Whatever improvement was already limited to those disavows in the past, and no further improvement is being seen with this so called Penguin.

[Simon_H]

We know very little about P4, but we do know that it's going to work in a different way to P3 (i.e. 'granular' approach rather than site-wide approach). This suggests that even sites hit by P3 that are not due to recover will *still* see some kind of traffic change because P4 will treat them differently to P3.

Therefore, it follows (and I'm making some assumptions here!) that if a site already hit by Penguin has seen no or minimal change in traffic patterns, it hasn't been processed by P4 yet.

[Shai]

Just to add that we have also not seen any significant movement on sites that are affected by Penguin.

[aristotle]

Walt Hartwell wrote:
I do not know how Google would treat an empty disavow file, so I downloaded my last disavow, removed everything except one foreign language domain and uploaded my one domain disavow.

Just disavowing one domain most likely won't have any noticeable effect. You should disavow all of your site's low-quality backlinks plus any backlinks that have contrived anchor text intended to boost rankings for particular keywords..

[buckworks]

More intense crawling

Consider how much crawling you actually need. Consider blocking some parameters in the Search Console to reduce crawling for pages that change very seldom.

[Walt Hartwell]

Just disavowing one domain most likely won't have any noticeable effect. You should disavow all of your site's low-quality backlinks plus any backlinks that have contrived anchor text intended to boost rankings for particular keywords..

Re-read the post. I wanted to remove my disavow altogether, but not being sure how Google would handle an empty file, I left one domain and re-submitted. Effectively, I went from 100+ domains disavowed to 1 disavowed and haven't seen any difference.

[Shai]

Sorry Walt, I don't get it. What's the point of leaving a domain in there or even leaving an empty file? If you wanted to get rid of the disavow file so that you are no longer disavowing any domains, why not just delete the file?

Also, as mentioned above, I still don't understand why any of the recent announcement made anyone think they no longer need to disavow domains?

[aristotle]

I went from 100+ domains disavowed to 1 disavowed and haven't seen any difference.

With Penguin, up til now at least, you usually had to wait a year or more, until the next penguin update, to see any effects from submitting or changing a disavow file. So if you changed it sometime over the past few months, you most likely lost a chance to see any benefit.

[Walt Hartwell]

Paranoia. I know nothing much happens when you change the number of disavowed domains/pages, I don't know what happens or doesn't happen when you delete it.

I changed that disavow prior to the recent Penguin chatter, others are still in place. My decision was a philosophical choice as I moved into the camp that believes Google correctly values or de-values links most of the time.

In a circumstance where there were significant "bad" links, it would probably be quicker to delete the internal pages targeted by those links rather than going after the links. If the links are to the home page, then that was a really bad choice.

[aristotle]

Walt -- Two of my sites were hit by the original Penguin rollout in April, 2012. I was pretty sure that the reason they were hit was because I had submitted them to a lot of free directories back in 2005-2006.

I don't remember exactly when Google began allowing us to submit disavow files, but it was probably in the last half of 2013. Luckily I had kept records of all of the free directories my two sites were listed in. To my surprise, many of those directories still existed, with my sites still listed in them, although in the meantime most of them had converted from free directories to pay directories.

Using my old records, I created disavow files for both sites and submitted them to Google. I had to wait about a year for the next Penguin update, but when it finally came, both sites were released from the penalty.

So apparently disavow files can work, and they should work much faster now with "real Time" updating.

P.S. Contrary to what some have said, all of the past penguin rollouts and updates only took one or two days to be fully completed. Of course it might become more gradual in the future with the transition to "real time' updating. But in the past a whole update would only take one or two days and then be over.

[System]

The messages about Google's AI [webmasterworld.com] were move to their own thread ... because this thread is focused on Penguin.

[edited by: goodroi at 11:46 pm (utc) on Sep 25, 2016]

[NickMNS]

Number four on the list of things to understand is that statistical analysis is an old technology and very likely has been superseded.

By what?
1- Statistical analysis is a practice not a technology, it cannot be replaced. The methods used to perform the analysis may improve over time but it can't be superseded. Machine Learning is simply statistical analysis at scale.

[sench]

I think it's more likely that any lack of "useful information" for gaming the system is a side effect of Google's use of AI, personalization, etc., not the reason for Google's advancing beyond the relatively simple search algorithms of a decade ago.

This is most likely a correct way to think about it

Also, I just wanted to provide a different perspective on this - isn't the fact that Penguin is now acting with every crawl (and per-page) actually resulting in more potential useful information than before? Since before your entire site was hit and you couldn't pinpoint which links/pages caused the hit. I'm not saying I approve of such methods, but technically wouldn't it be possible now to make some test pages and bring batches of different types of links (and anchor texts) in order to see if you could pass under the radar? Again, I'm not saying this should be part of any type of method or tactic in SEO, but just that there's actually more potential information available now than with the previous Penguins. So that's one more counter-argument to tinfoil-hat about Google, which some other users on this forum are too used to imho

[Shai]

Sench I would assume that a wide use of time delay algorithms on any ranking changes; good or bad, would neutralise those type of experiments.

[Nutterum]

I am not sure about people not seeing shakeups. There were hundreds of local websites in AUS that were gutted in the past 10 days. Before that there was a big stir in .de , .fr , .gr and .dk especially in the local scene. Many small websites operated by SEO agencies using the good-old "comment link spam" technique vanished over the past two months. Saying there was nothing on the Penguin front is confirmation that you know nothing about the nuances of Google soft-updates done on weekly/bi-weekly/monthly basis.

[Peter07]

I believe all the major fluctuations in the SERPS during this month of September happened because of Google testing and implementing Penguin 4.0 update. I saw a few Penguin penalized websites recover around the 10th of September.

[Shai]

Peter, welcome to the forums. Just out of interest, did you see those sites go back into penguin around the 14th?

[Peter07]

Thank you Shai.

Between the 10th and the 14th we saw a few sites recover from Penguin penalty. Since then rankings stayed the same. So we believe that the major fluctuations in rankings during this month, were Penguin 4.0 testing and roll out.

That would also explain why nobody sees anything happen after Google announced the Penguin 4.0 update.

[NickMNS]

@Nutterum, we are not saying nothing happened. What we are saying is that whatever happened seems to have (or is) happening progressively over a wide time span. This is in contrast to past Penguin roll outs were the impact was large scale and immediate.

[aristotle]

From the announcement:

With this change, Penguin's data is refreshed in real time, so changes will be visible much faster, typically taking effect shortly after we recrawl and reindex a page

So maybe this means that the penguin part of the algorithm updates on a page-by-page and search-term by search-term basis as googlebot recrawls each page and the algorithm applies all of the currently-stored backlink info, which is also updated as other sites are crawled.

[EditorialGuy]

Also, since the Penguin algorithm is now "granular" (page by page, not site by site), the algorithm's impact should be less sweeping or obvious.

[Whitey]

I'm anticipating no visible changes to rankings for commercial search terms because I believe "brand" has been the key establishing factor around "relevant" pages. Signals are simply internet activities that support the validation by Google that a site is a brand.

I'll be watching these threads to look for any reports to the contrary.

[Nutterum]

What I noticed from over 600 different keywords and their related phrases is that now more than ever exact/semi-exact domain name match plays a bigger role on the local front. Your results may vary of course but this is what I can say for sure.

[dan3s]

Hit by Penguin again. Traffic dropped a lot starting mid September.

[Shepherd]

It really speaks volumes that after the long wait for Penguin, 6 days after it finally, supposedly, updated there are only 54 (55 now) posts in this thread.

[NickMNS]

@Shepherd would you have expected more?

[Simon_H]

If you haven't already seen, Gary Illyes had a Facebook chat with Barry S, written up on SEL. It appears that (1) Penguin will no longer penalise sites for spam links, it will simply devalue them (as the original communication seemed to suggest), (2) because of this, there is less reliance on the disavow file and (3) the rollout is still not complete.

[mboydnv]

No sales in 4 days. Increased rankings however. This sucks.

[Simon_H]

And... in a separate twitter chat, Gary has said that Penguin rolled out in multiple moving parts, but the last part which was the most significant just *started* rolling out around the time of the announcement on 23rd.

[rustybrick]

Gary was very chatty today. :)