Google confirms Jan 2016 core algo update was not Panda

[Robert Charlton]

Just to break this discussion out as a separate topic...

On Jan 13, 2016 in the Google Updates thread [webmasterworld.com...] ...Simon_H reported Google's Gary Illyes confirmation about the weekend update...

Gary has just confirmed (again) that the weekend update had no Panda in it. Or Penguin. The mystery deepens...

The update was confirmed as a core algorithm update, but Panda and Penguin were not part of it. In this context, the lack of Panda changes has been the cause of some confusion. Here's Jennifer Slegg's report...

Not Panda: Recent Search Changes Are Core Algo But Not Panda
January 13, 2016 by Jennifer Slegg
[thesempost.com...]

Google confirmed on Tuesday that the recent changes in the search results that webmasters have noticed since late last week are not Penguin related. So then the next assumption many made was that it must be Panda.

While Google had previously confirmed the changes were Core related, many wondered if that meant it had to be Panda. But Gary Illyes from Google confirmed that while the changes are Core – and Panda is now part of that Core – that these specific fluctuations we have been seeing are NOT Panda.

And Barry Schwartz also reported on the update in one of his articles on the update in Search Engine Roundtable. He was more focused on Penguin than on Panda....

Google Confirms: Core Search Ranking Update Took Place But Not Penguin Related
Jan 12, 2016 - Barry Schwartz
[seroundtable.com...]

We also discuss the question of "real-time" vs "core algorithm" here on WebmasterWorld...

Difference between "core algorithm" and "real time"?
Jan 13, 2016
https://www.webmasterworld.com/google/4786150.htm [webmasterworld.com]

I don't know whether the absence of Panda updates in this past weekend's update is relevant at all to the question of the reported Zombie traffic, but I suggest some re-examination of what we're identifying as Zombie traffic might make sense.

I also sympathize with everyone waiting for Panda and Penguin updates. I remember that somewhere along the line, one of the Google spokespeople said that they hadn't yet gotten it completely right, and they didn't want to release a flawed version that would only make things worse for webmasters. I can respect that without being happy about it.

It might be good to put the recent update discussions in this thread... and to note any changes that clarify what Google might be rewarding or not, and which of Google's several algo components might be operating here. Any sightings, eg, of changes that suggest how RankBrain is being incorporated?

[HCFGrizzly]

Thanks for posting this thread. I saw many discussions these days about the update. Because Panda was introduced in the core algorithm and also, there was a core algorithm update, people got confused. Was the core update Panda related? Was Panda updated? Is Panda now realtime?
I think that finally this thread clarifies all the questions people might have.

[deuces]

Its odd that Google says this wasn't Panda or Penguin yet in that other thread, plenty of people whose site was hit by one of those recovered with that update, including me. It makes no sense.

[frankleeceo]

What if the update was giving a heavier score to one of the signals thus lowering the weight of Panda or Penguin. Then that could explain while it wasn't an animal update, plenty of people's sites rise based on other factors.

[Robert Charlton]

What if the update was giving a heavier score to one of the signals thus lowering the weight of Panda or Penguin. Then that could explain while it wasn't an animal update, plenty of people's sites rise based on other factors.

frankleeceo, I was originally in agreement with your thoughts, but I've come to think that Pandalized sites couldn't rise that far without a re-computation of Panda itself, so I've revised this part of my post responding to your comment.

If Panda/Penguin work as overlay filters, which I assume they do, then I'm thinking that any algorithmic peaks for other signals would necessarily remain lower than the filter layer. That's the tough thing about Panda... you keep bumping into a ceiling until the problems are fixed and Google reruns the algorithm.


Here's an analysis which is the most thoughtful I've seen on the update thus far, based on more data than most of us have available. It discusses winners and losers in the updated results....

Google Core Update 2016: Analysis of Winners and Loser
January 13th, 2016 - Marcus Tober
[blog.searchmetrics.com...]

As I've read over the article several times now, it suggests that all of the content was high quality content and was treated as such... so there were no Panda adjustments... but other factors running through the results were shifting the rankings away from what we've generally come to expect of authority-weighted keyword matchings.

I get the sense, from reading through the observations... and this is conjecture on my part... that...
- Google is likely to be emphasizing semantic factors (in contrast to keywords) more than previously, allowing more reliance also on other signals, including freshness and perhaps on topically related links.
- RankBrain is probably involved in interpreting some of the queries.

From the article....

In particular in the USA publisher websites with pieces of content that rank with brand keywords and entities have witnessed heavy losses in visibility. Brands, on the other hand, have seen a boost in rankings.

The article notes in particular losses for sites "with old URLs that ranked for brand keywords and entities".

A reported big winner, which was both comprehensive and currently topical...

Top winner according to our most recent Suite data is gq.com. On gq.com the biggest winning URL is a comprehensive article about NFL star Tom Brady

Note that the "individual keywords" reported in the January analysis are quite broad, and would correspond very well to the kinds of keywords we'd see in the News-Wave Update, which we discussed here...

June 17 Google algorithm changes: "News-Wave Update"
June 22, 2015
https://www.webmasterworld.com/google/4753914.htm [webmasterworld.com]

(I should note, btw, that the article I discuss in the News-Wave thread also happens to be by Marcus Tober at searchmetrics. No connection between us...I was surprised when I checked out the thread. Worth mentioning that both of these algo changes involve "extreme QDF").

Again, as with News-Wave rankings, it's likely that prominent named entities would be most visible and reliable as signals. I don't know the degree to which these factors might also apply to less prominent brands or entities on less prominent sites.


Also, to note briefly from what I think is the most interesting winner in the study that's an anomaly from traditional SEO (though it might be explained by inbound links to the sites)... the educational game sites were a fascinating set of winners...

User intent instead of content

One group of winners is particularly suprising for SEOs: educational games. Domains such as brainpop.com and mathplayground.com with landing pages such as this:

From a classical SEO perspective, these rankings can hardly be explained. There is only one possible explanation: user intent. If someone is searching for "how to write a sentence" and finds a game such as this, then the user intention is fulfilled.

These results are where I'll speculate most and guess that RankBrain may have had a lot to do with it. RankBrain is the newest major component of the Google algorithm, and I've been expecting to see it come along in a big way.

From Danny Sullivan's Oct 27 article on RankBrain [searchengineland.com...]

What Exactly Does RankBrain Do?
From emailing with Google, I gather RankBrain is mainly used as a way to interpret the searches that people submit to find pages that might not have the exact words that were searched for.

I think that we're likely to see more such departures from the expected traditional keyword link patterns as Google correlates expectations from various aspects of the algorithm.

PS: I should add that I'm also seeing some reports of a rollback, and I've seen one graph that looks like a rollback.

(Edit note: My opening comments in this post changed as noted above).

[vlexo]

Starting to see signs of a potential roll back, but it's almost as if things are going back and forth.