Google Panda 4.2 Rolling Out

[netmeg]

According to Searchengineland:

Google has pushed out a Google Panda refresh this weekend, many of you may not have noticed, but this roll out is happening incredibly slowly. Google said the update can take months to fully roll out because it will slowly roll out through your site. The Panda algorithm is still indeed a site-wide algorithm but some of your web pages might not see a change immediately.

Emphasis by me, because WTF.

A Google spokesperson confirmed with us the update did being rolling out this past weekend. They also noted it can take months to fully roll out. Google did not share with us how large of an impact this was on their search results, but they did imply it was a fairly small impact.

I haven't seen any signs of it in any client sites yet, but maybe it'll be "months"

[searchengineland.com...]

[AngelosKyritsis]

@Itanium thanks for the tip. I couldn't find any links to my own site with the user agent switcher, but I did find some hidden links that point to a couple of other sites, I guess this is what you meant.

<snip> (removed screen capture w/details about domain linking to poster's site)

Do you think it would be any use to contact those webmasters, or should I directly disavow those sites with the thousands of links to a specific page?

----
Mods note: Public site reviews with specifics are not allowed, per TOS and forum Charter. This means no screen captures with specific backlinks, no details about other site, no domain names, and no specifics about site's own content. Sorry, but that's what we need to do... for everyone's protection... to avoid outing and promotion, and to keep the forum clean.

[edited by: Robert_Charlton at 8:33 pm (utc) on Oct 1, 2015]
[edit reason] removed specifics, per forum Charter [/edit]

[Itanium]

Yes, I meant those hidden links. They can be anywhere on the site. Sometimes they are pushed back behind a div, or with css out of sight etc.

I'd disavow the domains with links as these and contact the webmasters afterwards, just to make sure. Better safe than sorry....

[JesterMagic]

I would submit a Disavow file just to cover your end of things. I am not a fan of things like the disavow file and nofollow links but I do understand that Google is fighting a loosing battle here. I waited a while before doing these 2 things which was a mistake because it did help me regain some lost traffic after I submitted them (at least IMO).

[Globetrosky]

Heard that Google looked at 18 months of historical data in Panda 4.2, so if your site was hit a year ago, and you worked really hard to fix your site, you likely will not get out (at least not fully) because your old site data is hurting you. This is not how prior Panda updates worked, so may be responsible for the lack of dramatic Panda recoveries.

[aok88]

Heard that Google looked at 18 months of historical data in Panda 4.2

Where'd you hear that?

[Globetrosky]

@aok88 Lunch last week with a friend at Google.

[AngelosKyritsis]

Thank you @Itanium and @JesterMagic, I have disavowed 53 domains that had multiple links to a single page. Let's hope this fixes things up.

What's worse is that this sabotage had me fooled that my website had actual growth, and I have no idea what my "real" statistics would be.

<snip>

[edited by: Robert_Charlton at 10:27 pm (utc) on Oct 1, 2015]
[edit reason] Removed search terms poster is ranking for. [/edit]

[Mentat]

Well, another Panda push this weekend, another low point for me.

[aok88]

What makes you say it was a Panda push?

[Mentat]

Curves, downward spirals.

I saw the in the last 2 years each week, so Now I can spot them quickly on graphs :)

[AMC4x4]

@AngelosKyritsis and @Itanium - Oddly enough, I run a very small business site, and I saw the EXACT same traffic profile - a boost toward the end of July, and then a dramatic hit around September 23/24 or thereabouts. Our impressions from Google fell, as did our page rankings. For most of our common expected search terms, we were consistently ranking around 5-10, and now we've gone to 20 or lower, to completely invisible. The loss is much greater than what we were at before the boost at the end of July.

Just thought it was odd that I saw the exact same pattern as @AngeloKyritsis.

[AngelosKyritsis]

@AMC4x4

For me, my rankings and traffic started picking up again after October 6th. They still haven't reached their previous level, but it's looking good.

[i.imgur.com...]

Disavowing hacked domains pointing to my website seems to have helped a lot.

[AMC4x4]

@AngelosKyritsis - that's great to hear. I didn't have any unnatural links, but I did have some "suspect" ones, so I have disavowed all those. I run a small site, so even though it was about 200 links, it was a decent percentage by ratio. I think perhaps these links were artificially inflating our page rankings before we were apparently slapped down algorithmically speaking, so it will probably take a while for us to recover. Just gotta build that content I guess. The issue for us may or may not have been these links. They were "mirrored articles" from other subdomains of the same corporate site. This company apparently takes articles and then reprints them on multiple subdomains. Those mirrored articles had backlinks to our site.

We had a lot of impressions, but not many clicks based on those impressions. I guess the good news is that since the slapdown, the visitor quality has improved. Our rankings might not be anywhere near as high, but when people do find us, they are looking at more pages and staying longer. I guess that's what it's all about, so maybe Google knows what it's doing after all. :)

Cheers.

[adder]

There might be some clues as to what happened in mid-September during the unconfirmed Google Updates.

Have you seen this Twitter conversation involving Glenn Gabe, Gary Illyes (Google employee) and Barry Schwartz:
[twitter.com...]

There have been no tweaks to Panda 4.2 thus the sudden drop on 16th (or 12th) September can't be related to Panda.

What is it: Phantom or something link-related? I think I'd vote "Phantom"

I posted this message to the current "Google Updates and SERP Changes" thread but, alas, I couldn't envisage that it would quickly spiral out of control and turn into a place where people discuss Conspiracy Theories, so if Moderators wished to cut it out from the original thread and move to this thread, I'd have no objections whatsoever :)

[webmasterworld.com...]

To cut a long story short, there are two unrelated websites both with history of Panda problems. Both had a big drop in organic traffic around mid-September following an impact from the Panda 4.2.

Here are the screenshots:

http://imgur.com/RJDInTc
http://imgur.com/1BAKSEO

I can almost certainly exclude a backlink penalty. Has anybody else experienced this drop and do you have any evidence that suggests it may be related to Phantom?
Thanks

[aristotle]

Posters keep mentioning different dates for when their sites were hit. One possible explanation would be that Google is making various adjustments to various signals day to day, sometimes reversing a previous adjustment depending on the results it produces. If so, they are perhaps trying to gradually optimize their algorithm, while at the same time gradually tightening the clamps overall.

So we might be looking at one long gradual process that affects different sites at different times.

(Note: I put this in the Panda thread because we know that Panda is rolling out, but I suspect there might be more to it than just Panda.)

[SuperT]

Figured we were hit with a panda refresh late last year. Been fixing a lot of pages and it looks like on 10/08, SERPs started to gradually drop with a larger drop on 10/14. Anyone else seeing these drops around those dates?

[MrSavage]

It's been a while since I last checked this thread. I can speak on my own results and at this point there is still a flatline. No bump. Not a great investment of time it looks like. No pot of gold for me. I'll check back in a couple months to see some bumps from others. Afterall, this is a slow rollout. Months even. Or perhaps never for a person like me.

[Nutterum]

I think this will be the last panda we will see. Once this thing fully rolls out I'd suspect it will already be incorporated into the algo just like Penguin (will be hopefully). Panda does roll, make no mistake. I`ve seen several site hit by the algo in October alone. I'ts just not in the same way as previous updates. I think Panda is moving at the pace of the recrawl rate, which is understandable.

[Globetrosky]

@Nutterum - why do you think this will be the last Panda? Matt Cutts said back in March 2013 that Panda and Penguin would "soon" become part of the regular algorithm, but 2 years later they are still running manually.

[Mentat]

Well, I don't know if it's Panda or something else, but since last night my traffic is sinking like a rock!

[aok88]

Well, I don't know if it's Panda or something else, but since last night my traffic is sinking like a rock!

Same here. Since 10/25 visitors have dropped big time. Ugh!

[RedBar]

Well, I don't know if it's Panda or something else, but since last night my traffic is sinking like a rock!

Something's happened plus Statcounter seems to be almost totally fubard as well!

[mrengine]

The traffic sinking could be one of any number of factors. In addition to the slow panda we have zombies, rank brainless AI and who knows what else they will come up with next. All of these very close before the holidays, which usually sees adjustments by google also.

[Mentat]

Watch out for Halloween, Veterans Day (11.11) and Thanksgiving Day (4 days long weekend).
November will be like a storm!

[Martin Ice Web]

Has someone experience with panda and put a nofollow on low/similar/dup content sites? I wonder if panda does count them and if they still would raise a panda flag?
My thought is, that if i have a 90% copied site and do only 10% good compelling content, why should a SE think this is an overall good site? Still google says donīt do exclude any pages in robots.txt. This could harm our sites traffic and ranking?
It is not that i want to copy something but to put some widget pages with only manufacturer description online. All on nofollow. Still this pages are usefull content but similar to many togther pages out there, I donīt want google to have them in index nor to rank them put i need to have this pages as a big portfolio is the way to go this days.

Any thoughts, ideas, experience?

[Robert Charlton]

Has someone experience with panda and put a nofollow on low/similar/dup content sites?

Martin_Ice_Web, I'm guessing in the several instances where you said "nofollow", you were thinking of using the robots noindex meta tag, which also has options of using a "follow" or "nofollow" attribute. The meta tag defaults to "follow", and (for Panda control) should be used that way.

To clarify the distinctions among the meta tag, nofollow, and also robots.txt would take this thread way off topic for a long time, but let me suggest you read this long thread which eventually covers them all. You may have to read it twice. It's a very meaty discussion....

Pages are indexed even after blocking in robots.txt
https://www.webmasterworld.com/google/4490125.htm [webmasterworld.com]

For Panda, it is the noindex meta tag that's commonly suggested to keep Google from displaying low quality pages. I've always regarded the use of noindex for Panda, though, as kind of a temporary band-aid, and I agree with you that it is not an adequate long term solution for dealing with a lot of inadequate content.

i need to have this pages as a big portfolio is the way to go this days

I'm not sure I agree with this. Most sites I've critiqued have more pages than their inbound linking can support. You certainly don't want to have empty templates sitting on your site... but I think it's a mistake to think that "a big portfolio" with bad or mediocre content is going to help you.

Recently, and I don't remember where I saw it, Google's Gary Illyes suggested adding content to thin pages rather than noindexing or deleting them. I feel that if you have a lot of thin pages, though, it's likely that many of them are superflous. I would look carefully at site structure, carefully prune, and also add content to pages that are necessary for site structure. Gary's concern, I think, is that many of these decisions are made hastily... and you don't want to remove necessary pages from your site without careful consideration.

[Martin Ice Web]

Thanx Robert,

i donīt use robots.txt for "noindex". i meant that i out the meta "noindex" on them. Thanx for this in regard to panda. I will read this thread.
So in short: noindex means no Panda issue?!

I need to have this pages because many customers are looking for widgets i donīt have online because i did not have the time to write adequat descriptions for them. But i see users searching form them ( with EAN, manufacturer number, manufacturer description ... ). This is why i need to have them online but donīt want google to count/index them.
I will slowly rework them an then "index" them. There will be only "nofollow" links to them.

[Robert Charlton]

...There will be only "nofollow" links to them.

NO! Very simply, I would never use nofollow on internal links.

In this case, use the meta robots noindex tag by itself. Note that it defaults to "follow". Syntax is...

<meta name="robots" content="noindex">

Do not use the above in combination either with robots.txt or with the "nofollow" link attribute. Again, read and study the discussion I link to above and also the references suggested.

[Robert Charlton]

But i see users searching form them ( with EAN, manufacturer number, manufacturer description ... )

Martin - A PS re your post above... and that is to point out that "noindex" will prevent these things that searchers are looking for from appearing in the search results. Google doesn't want to display thin pages in the index. So, your users won't find these by Google search, at any rate. You might try an independent site search, or a chart that guides to them.

SKU numbers in the US present similar problems.... what do you do if everything is identical about two pages except for, say, the color, and the SKU numbers are searched for?

To include the page content on all identical products, that differ only in color, is an invitation to Panda, particularly if you've got many products that are identical except for a single product attribute. It's a problem that I haven't solved.

This is a particularly difficult situation for manufacturers.

[Martin Ice Web]

@robert, i think u donīt understand me.

"You might try an independent site search". this was the first intention of my first post.
And yes, i use <meta name="robots" content="noindex">
And yes, i see my users searching on my site search for the mentioned search terms.
I never said that i want to have this pages in google index. I just want to have this pages only for my customers, who allready bought something and now search for a certain item.

All I wanted to know, if someone has some experience with having a lot of pages with <meta name="robots" content="noindex">. ?