Panda 4.1 Rolling Out

[netmeg]

Based on user (and webmaster!) feedback, we’ve been able to discover a few more signals to help Panda identify low-quality content more precisely. This results in a greater diversity of high-quality small- and medium-sized sites ranking higher, which is nice.

Depending on the locale, around 3-5% of queries are affected.

Google+:
https://plus.google.com/+PierreFar/posts/7CWs3a3yoeY [plus.google.com]

(sorry, it's https so it won't link)

Searchengineland:

http://searchengineland.com/panda-update-rolling-204313 [searchengineland.com]

[edited by: aakk9999 at 11:40 pm (utc) on Sep 25, 2014]
[edit reason] Made link clickable [/edit]

[JD_Toims]

I don't think we matter anymore...

Sadly, I think you and jmcc might be on to something...

It used to be we had GoogleGuy, Vanessa Fox, Adam Lasnik, et al posting, explaining, asking for input, sharing tips, etc. and not only did what they say make sense, the tips they shared worked -- Fast-forward a few years and they've long-since left here and the tips we do get from other places where some still share often only seem to work for a very limited subset of sites, if at all.

Google has burned through a lot of webmaster trust and destroyed businesses. It had better start working to regain some of that trust.

IMO They've got a long, tough road in front of them if that's a direction they ever decide to go.

[dethfire]

Why would they care about webmaster trust. As long as their stock rises they are happy. If it drops, then they pinch us even more.

[JD_Toims]

I can't argue with your question/position, and it's why I doubt they'll ever go down that road or give us info like they used to share here, which to me means *everything* they say should be questioned and taken-with-a-grain-of-salt rather than just "drinking the Googleaide" the serve up to us on a platter.

If they are really after quality results and that's what the algo serves to their visitors, they could simply define what "quality" is to them and the only way for us to "game" the algo would be to increase quality, which means it *should* be an all-around win, but they don't/won't tell us and that makes me question what their real goals/motives/objectives are for some reason.

[aok88]

All I can say is that this is the worst converting traffic I've ever seen...but SUDDENLY that can swicth and traffic turns HOT and converts...many times within minutes of the first conversion....then off for another 12 to 24 hours...not a peep. Like a swingin' carrot.

samwest: I'm have the EXACT same problem as you with one site. It is mind-bendingly frustrating because when the 'spigot' is turned on, which is only for a short time every couple of days, the orders come flooding in - and they're big orders, then boom, there is very little and they are much smaller orders. I can see it very clearly happening. And the crazy thing is, there is no traffic increase or decrease whatsoever.

When zombie traffic persists for a long time, IMO this suggests that Google is seeing something about the site as marginal enough that the site becomes a subject of frequent Panda testing. I hope that doesn't happen to you, as members who have gone through that find it extremely frustrating.

Robert Charlton: Yes, it is rough. What you suggest certainly makes sense - what makes you think that? Have you seen anyone fix it/come out of it? And if so, any ideas as to what was done?

[JD_Toims]

I'm not RC, but skipping to Page 15 @ 30 Posts Per Page of the Zombie Traffic thread linked by RC and then continuing to Part 2 (linked in the last post of the thread RC linked and below) might give some insight as to how at least part of it can happen and why it may have nothing to do with actual visitors to your site.

[webmasterworld.com...]

TL;DR

Zombie Traffic could be some sort of previews being intermittently tested again.

[EditorialGuy]

If they are really after quality results and that's what the algo serves to their visitors, they could simply define what "quality" is to them and the only way for us to "game" the algo would be to increase quality, which means it *should* be an all-around win, but they don't/won't tell us and that makes me question what their real goals/motives/objectives are for some reason.

Google has been saying for a long time now that site owners should build pages for users, not for search engines. Besides, building pages to fit a checklist wouldn't produce quality content; it would produce s*$! disguised as Shinola.

[JD_Toims]

Blah, blah, blah -- I wish I could say "great post" and mean it, but...
<snipped>

Theoretically, it might be better for everyone if you post the truth rather than arguing simply to argue, but of course maybe I'm just plain stupid -- I'll let the readers decide...

[edited by: goodroi at 11:23 am (utc) on Oct 6, 2014]
[edit reason] TOS [/edit]

[samwest]

Zero traffic all morning...stick a fork in me. I'm done.

<snip>

[edited by: goodroi at 3:00 pm (utc) on Oct 6, 2014]
[edit reason] TOS [/edit]

[RedBar]

traffic spiked up 400% on every 2nd or 3rd Sunday

Yep I've been seeing this for months now, why are you doing this G?

[Martin Ice Web]

@samwest, i am on vaction in austria, 400 miles away. Open WLAN network, computer with gerneric users on it. Guess what: serps a totaly the same as at home. No difference. So i guess that what we are seeing is heavy personal search results on user habits and buying habits. Maybe panda is not panda as they say it is ( looking for good content ) but just a plain personal search factor that is beeing lifted sometimes.
I juzst had to think of how MC introduced Panda:

DON`T EXPECT ANY FREE TRAFFIC IN TRHE FUTURE FROM GOOGLE ANYMORE!

That is wher we are right now, but the just keep saying, do your best, make compelling sites. Why? They need input for their system to scrap. Nothing more.
Two days ago, i read an article abouit amazon. it was interesting because there was to read that all analyst sayed asmazon not to take part in google shopping while this will give google big data about shopping habits from poeple on amazon.
This makes the point.
What did amazon, that we did? Wew gave google all our data. Amazon did not!

[spreporter]

@martiniceweb If you have a brand like amazon,booking,ebay etc. you need no search engines you just type the brand name and do your shopping.

[dethfire]

If you have a brand like amazon,booking,ebay etc. you don't need this community. You should be on your yacht

[spreporter]

@dethfire.....I'm on a cruise right now yacht comes later...lol

[EditorialGuy]

For what it's worth, Google said this afternoon that Panda 4.1 is still rolling out:

[searchengineland.com...]

[Shepherd]

Wow, that's quite a rollout.

I wonder...

1. is it a size issue? Too much data, not enough computing power.

2. are they having issues with the code?

3. are they not getting the results they expected?

4. is there no problem at all?

[EditorialGuy]

Wow, that's quite a rollout.

Well, they said it would be slow. And it's actually fairly quick by historical standards. Back in 2011, it took Google a month and a half to roll out Panda for English-language queries worldwide and nearly six months for Panda to roll out in other languages (and not all other languages, at that).

[jmccormac]

If they are really after quality results and that's what the algo serves to their visitors, they could simply define what "quality" is to them and the only way for us to "game" the algo would be to increase quality, which means it *should* be an all-around win, but they don't/won't tell us and that makes me question what their real goals/motives/objectives are for some reason.

Actually one of the reasons that they seem to be concentrating more on "content" is that they need it as raw data for their AI processes. The more content that is available, the easier it is to summarise a site and even generate a coherent abstract (think knowlege/scraper graph "answers"). This could be why the approach to downgrading sparse content sites that were linked well was being pushed over the last while. Even with a modicum of intellectual power, the links issue could be solved but the emphasis on the AI approach has caused them to overlook the rather obvious point of how websites have a Social Network of their own. (Perhaps given how Google has banjaxed almost every social media venture it attempted, it is understandable.) The web isn't just about information. It is about people communicating with people. Take that human aspect out of it and you've got Google's approach to "quality". Being utterly cynical, perhaps what Google thinks is "quality" is data that fits into its algorithm and produces the results that Google expects. For Google, disclosing what it really considers "quality" would lead to its algorithm being reverse-engineered and exploited. Perhaps that post from Google on what constitutes a "good" website might be an indication of what Google considers a "good" or "quality" website.

Regards...jmcc

[Zivush]

Take that human aspect out of it and you've got Google's approach to "quality". Being utterly cynical, perhaps what Google thinks is "quality" is data that fits into its algorithm and produces the results that Google expects.

@jmcc
That's exactly what I speculated for more than 3 years.

Panda 4.1 is still rolling out

Yes it is. I wrote about seeing an update in action two days ago.

[jmccormac]

@Zivush
It is beginning to look that way.

Perhaps one possible approach would be to start reading up on abstract generation algorithms. This is one of the first of them:
[courses.ischool.berkeley.edu...]

Tweaking content so that it is favourably summarised/abstracted (the content, when summarised, produces exactly what you want it to produce) might be the best way for webmasters to deal with Panda. (Pure speculation of course. :) )

Regards...jmcc

[Zivush]

Thanks jmcc. I am going to read this paper.

As for Panda 4.1, I think what with seen this weekend was 20%-30% roll back / reversed.

[hasek747]

I have 12 different websites in 8 different niches, all 2 years old or older. Compared to this same time last year, traffic is up by around 350% on average (minimum is 200% maximum is 400%), income from websites is up even more than that. All websites in the US, all review/affiliate based.

Again, these are 8 different niches, so not just one that I got lucky with. To be honest whenever I read a thread over here in this part of the WebmasterWorld forums I almost feel like I live in a sort of parallel universe, because my whole experience has been nothing but positive with Google (speaking as a webmaster with a whitehat approach to SEO, at least as whitehat as outreach to other webmasters gets). My only theory is that it's e-commerce sites that are getting hit hard and that most people posting here own e-commerce sites, though I find that a little hard to believe.

[Sand]

Since Friday, traffic for me has been down 10% compared to the same day the week before.

To add some clarity, I gained about 40% when Panda 4.1 first started rolling out. So I'm still up, just not quite as much.

[RedBar]

US traffic way, way down so far for me today, something's still happening.

[mrengine]

"Tweaking content so that it is favourably summarised/abstracted (the content, when summarised, produces exactly what you want it to produce) might be the best way for webmasters to deal with Panda."

My interpretation: Word and structure content so that it can be scraped and displayed in a Google knowledge product.

If this is what Google is becoming, we had best hope for a search product to evolve that satisfies the consumer while also recognizing the rights of website owners to control how their content is used. Google's market share makes that impossible at the moment. More competition is good for everyone.

[JD_Toims]

Word and structure content so that it can be scraped and displayed in a Google knowledge product.

Well, if someone wants to build the "Star Trek" computer based on AI, to have a "knowledge engine" rather than a "search engine" it makes the job quite a bit easier if people do that for them, doesn't it?

Sad as it may be, that doesn't seem too far-fetched to me.

[Zivush]

@Sand
I can tell that the ups and downs are still going on in my site.
Yesterday, 4.1 update continued and it brought the site back to its top ranking position.

[getcooking]

I don't normally complain about the SERPS (other than wishing my site ranked better!). This morning when I was checking some of my keywords I actually laughed at some of the results. Pinterest had jumped up in the rankings, and Buzzfeed was even above that - and for this keyword it's never been on page one (and totally doesn't deserve to be for that keyword). Even funnier was the "in the news" link that sometimes appears. It was showing right below the top listing (which it doesn't normally do for this particular keyword) and it linked to what else?.... a well known content farm.

Seriously hope that Panda 4.1 isn't done! Oddly, my traffic is up quite a bit today despite a pretty serious flux (and some drop) in keywords since yesterday.

[rbarker]

Pinterest had jumped up in the rankings

I saw that to for several terms. In one case Pinterest dominated half of page one and almost all of page two...

[ChrisWilson]

I was checking on of my top keywords and noticed that instead of being number one, like I have been for years, the number one spot is now held by a Pinterest page.

So of course, I go to the page and take a look at it. I would say about 60% of the links on that page show content from my site. Yet the Pinterest page out ranks me.

This love Google now has for Pinterest is getting absurd. Makes me laugh though, as traffic has not changed, my only guess is people see the majority of links pointing to my site, and just go to my site.

I simply laugh because in my case, the user actually wants to see what's on my site, but Google is making them stop by Pinterest first. lol

[snippet]

So of course, I go to the page and take a look at it. I would say about 60% of the links on that page show content from my site. Yet the Pinterest page out ranks me.

It makes sense in some search contexts that they would rank above you. Perhaps their collection of content is more complete than your 60%. The other 40% can give the user more options that they might not see if they just went to your site.

I think this is also why a lot of top ten lists rank over the source sites.