Google Updates and SERP Changes - Jan 2012

[londrum]

< continued from: [webmasterworld.com...] >

i tried to sprinkle different variations of one phrase onto each page

if the target phrase was "big red widgets" then the page might have a couple mentions of each of these

big widgets
big red widget things
big widget things
big red things
red widgets

plus an exact match in the <title> and <h1>

[edited by: tedster at 12:31 am (utc) on Jan 2, 2012]

[Lenny2]

Lets not forget that this past weekend was a holiday in the US...

[Whitey]

I wonder whether they have already incorporated Panda different factors within the general scoring algorithms. I.e we are not gonna hear about a Panda update anymore ;-)

I think we're heading that way -sooner or later.

[AlyssaS]

Gshaughan - because those of my sites with really good links suddenly took additional #1 positions but the sites with not great quality backlinks dropped a few places.

I feel they've devalued some types of backlinks, but am still going through trying to work out which types

[SEOPanda]

I have a feeling all sidewide inbound backlinks are devalued. For example, if a link to your site is in a sidebar across the site.

[gyppo]

My sites with less robust backlinks are performing better than the ones with solid link profiles. I have no reason to believe so far what we're seeing is link based.

SEOPanda - Hasn't this been the case for a while already? i.e. a Sitewide not really counting for more than 1 link from a domain? If anything there's a possibility they're cracking down on users that have an inconsistent link profile (i.e. lots of links with the same anchor text from a handful of sites).

If they had rolled out something to improve link scoring & reward quality I would expect lots of the SERPS I monitor to have seen decent movement. Not really seeing that currently.

[SEOPanda]

It's been the case for a while, but I think we really saw this with Panda. I have seen sites with horrible off-topic backlinks rank very well, and almost none of these links were sitewide. For sites with good quality links, but sitewide, it looks like Google has taken the approach of not counting it just once, but not counting it at all with the assumption that if it's sitewide, 95% change it's an ad.

[seoskunk]

Feb Pandalised site partial recovery, no changes made since october last year. Just thinking aloud but could this be the affect of a more mature site in SERPS and the site gaining more trust because its over two years?

[londrum]

maybe they are just devaluing links, period - across the board. and focusing more on user-related stuff. we know that's where they are headed, with google+.
in a years time i reckon links wont be the big thing anymore. those 'votes' only came from a small segment of users anyway -- those who had access to their own webspace. thats why google would be clever to shift its focus over to user behaviour. maybe thats why sites with no backlinks are getting a bump up

[SEOPanda]

For those getting a bump in traffic on a panda site, check to see if your backlink profile changed.

[Gshaughn]

Social strategy was sound for my site that got hit. Naturally good. Doing way more than other sites in the space. Just FYI.

[SnowMan68]

First time posting here.

I have seen a nice increase in our Google traffic (20-30 percent) on one of our Panda'd sites over the past couple days.

[Whitey]

@SnowMan68 - can you describe what activities you have done to your site to restore it? It would be good for others to see if it is Panda related or something else.

There is some speculation that "quality score" elements are being, or will be phased or migrated into the generality of normal frequent algo adjustments. Your insight would be helpful to others.

btw - Thanks for sharing and your 1st post

[SnowMan68]

@whitey

We purchased a content/informational site in Feb. 2011. Within 3 days of owning this site it was hit by the infamous Panda update. Overnight we lost about about 70 percent of our traffic. The site itself was outdated and when we bought it we knew it needed to be updated. The site was what Google used to consider as an authoritive source on our subject. It has thousands of links and 100's of them are very high quality.

The site itself had a lot of glaring issues from a usability stand point. It was in need of an update and face lift. The navigation was hard to follow, it had a ton of ads above the fold, pages targeting misspellings, very slow loading blog and some duplicate content issues with articles targeting slight variations of the same keyword phrases.

What we have done over the last 10 months to fix the issues... We have completely redesigned the template, with 100 percent brand new code. Fixed the navigation issue with a new drop down menu. The flow is better by breaking the site into 5 major sections. Within each section is a section menu that is specific to only that category. The new template also makes it easier to share content (facebook, twiter, ect) Removed most of the ads (it was using in content Kontera ads and Chitika ads). We replaced them with one small chunk 250 x 250 of adsense ads under the main title. We 301'd the pages targeting misspellings to the pages with the correct spelling of the same subject. We also 301'd the pages that were targeting slight variations to the main subject page. We fixed the slow loading blog by going through the WP modules and removing any modules that were slowing it down. The blog loads about 4-5 times faster now.

[Whitey]

I have seen a nice increase in our Google traffic (20-30 percent)

Can we play with some speculation here - if you were observing a traffic increase due to a "quality score" element/s that is no longer part of the Panda iteration, or migrated into a newly introduced interim adjustment, what would you observe that the element might be.

There's a bit of discussion here about "Rolling Updates" [webmasterworld.com...] and I'm playing with the possibility that the remaining 70% of traffic is still trapped within Panda as part of the main iteration, which requires a further re calibration. What is of interest is that you have a partial recovery before any "announcement" of a new iteration.

Also do you think you've done enough to break the beast completely?

[cr1t1calh1t]

tl;dr

I think the changes we are seeing are panda related, but I'm not sure they've 'settled' yet. I was affected by the Oct. 13 panda iteration and after lots of work on the site am currently seeing what looks like at least a partial recovery, with unique visitors up 25% since Jan. 3 vs. same period in December.

October 2011

A site of mine lost a little over 10% of its traffic following the Oct. 13 iteration, coming off steady growth in the prior 2-3 months. Adsense revenue dropped by almost 20%. While the drop I experienced wasn't as dramatic as what other folks have reported, I am fairly certain it is panda related because of the timing of the change and the selective nature with which it affected pages on my site.

After reading SnowMan68's comment, I was reminded that I changed the adsense code on my site in the beginning of Oct from my personal account to my business adsense account. Can anyone speak to how google might treat sites that signal a change in ownership?

I spent the second half of Oct. and all of Nov. doing the following:

• fixing lots of technical and duplicate content problems related to Joomla URL generation (and the flawed way that I implemented Joomla in the first place)
  
• added canonical tags
  
• noindexed the entire forum
  
• removed ad units
  
• sped up my site load speed by 100%
  
• a host of other fixes, changes, and improvements.
  

I didn't rewrite any content on the site except for one page that was hit especially hard. I wanted to test Content_ed's rewrite strategy from this thread: [webmasterworld.com ], so I did a word for word, line by line total rewrite. I posted the rewrite on the night of 1/1/2012, while all of my other work was wrapped up by the beginning of December. I didn't remove any content from the site or noindex any pages except for the forum.

To give you an idea of the size of my site, prior to noindexing the forum a site:mydomain.com search showed approx. 23,000 pages indexed. Today the same search shows about 5,000 pages indexed, and this decreases every week. Aside from the forum, there are about 1,500 pages that range in length from 100 words to 3500 words, not including comments (some pages have as many as 500+ comments).

I've kept a detailed log of all of the work that I did to the site since Oct. 13, as well as GWT reports and other stats for this time period and am going to put it together in a report on my panda related efforts after I've removed any way to identify my site from the data... If forum rules permit, I'll post a link to it when I get it finished.

January 2012

I started seeing significant changes in rankings as well as a surge in traffic as early as Jan. 1, which was particularly surprising since this site's traffic falls off the map on Sundays and on holidays (New Years Day was on Sunday, fyi...)

On Jan. 3 this site hit an all time high for unique visitors, and has hit numerous milestones in other traffic related metrics since then - including a new all time high for unique visitors on Jan. 17.

Interesting Traffic Pattern Change

Normally the site's highest traffic is on Monday and each successive day's traffic is lower than the previous day's traffic. This site's traffic follows this pattern like clockwork, barring holidays. In the week beginning Monday Jan. 9, however, Thursday was higher than Wednesday, and almost as high as Tuesday. Beyond this being a rare and interesting deviation from a very reliable pattern, I'm not sure what to make of this.

I'm not ready to claim a victory or say that I've reversed the effects of Oct. 13 because some of the early January ranking changes are still in flux and I don't think these changes have 'settled'. On balance, I do think my rankings and traffic will end up showing a net improvement for January 2012 over December 2011.

It is very important to note that this site also hit a new all time high for unique visitors on Oct. 3, which I think gives some evidence to some of the theories out there about how Panda works.

Traffic Increase

Unique visitors are up by about 25% for the period of Jan 3, 2012 - Jan 18, 2012 compared to Dec 3, 2011 - Dec 18, 2011.

Position Changes

Using DirigoDev's methodology of summing keyword position increases and decreases:

Dec 12, 2011 to Jan 11, 2012
----------------------------------------
...........................Point Change...........Keywords
Down.......................-754.8..................541
Up..........................918.5..................645
Total Up/Down......................................1186
Total Keywords.....................................7148

In this period about 75% of my top 100 keywords had a change in position, with the range being from -3.0 to +5.0.

Thank You!

I'd like to give a big thanks to all the folks here that share their observations and theories and experiences. I've been following the panda updates from the beginning, but it wasn't until I felt the effects in Oct. that I really dug into it on a nuts and bolts level - and the discussions here at WW have been the best that I can find, and have been crucial to guiding my efforts to recover from panda.

cr1t1cal

[leeds1]

@cr1t1cal well done on your success - when you noindexed your forum how did you do it? block it in robots.txt or add noindex (or noindex/nofollow) to each forum page ?

[cr1t1calh1t]

@leeds1 - thanks, but I'm reluctant to celebrate any success just yet... To answer your question, I added noindex to each forum page.

I had previously blocked much of the forum in robots.txt, but removed all of those blocks after I added noindex so that the bot could access those pages to see that they should be noindexed.

For more on my noindex of the forum, check out this thread: [webmasterworld.com ]

cr1t1cal

[leeds1]

thanks for that - I have a forum much like yours so will do what you did and report back on any progress

[gyppo]

My main site which has been Pandalized since Feb last year is back today for the majority of it's major keywords.

[Whitey]

@cr1t1calh1t @gyppo - Are we able to make a call on rolling Panda updates or are things still too hazy with not enough "buzz" to confirm.

Maybe it's time for someone with connections to lift the phone to Google and ask [ if they will tell :) ]

[mhansen]

I can say with a high degree of certainty that a site I follow (Not mine, but I almost bought it!) very closely was negatively affected in serps late Monday night.

I know this for fact because I was in the middle of negotiating the purchase of the website. It was a -100 page consumer guide type site with a domain name of "keyword1keyword2keyword3.tld.

On Tuesday morning, I woke up to check email and start the money transfer procedure and the site is completely gone from previous serps. I waited a day longer to be sure it isn't a burp, but sure enough... its gone. Even exact quote searches on the homepage are not found in the top 10.

Just for the sake of record.

- Site was WordPress
- Had -100 pages
- Content was weak, at best. (rewritten, duplicate per copyscape)
- Had a single Clickbank offer on it (No adsense)
- Domain was exact match 3-keywords (+10k exact match searches per month)
- Domain is 3 years old

I am very happy the owner didn't sell early in the process. I would be stuck holding this bag.

MH

[gyppo]

@Whitey for me it's still to hard to tell, I've had 2 sites come back from oblivion this week alone. Both had characteristics of Panda (traffic drops during every iteration since Feb 2012) + 950 Penalty.

We've made so many changes, which I can document here if you would like. However, the most recent change we made to both sites was requesting removal of sitewide links on some blogs - coincidentally within a week of getting these removed both sites returned. We're not fully back yet, some stuff is still suppressed.

That's why I'm finding it difficult to know if this is Panda or not.

[Whitey]

@gyppo - hmmm .... it looks hazy. So many of those improvements may/may not play into the quality score. Who knows.

Feb 24, 2011 [webmasterworld.com]
Apr 11, 2011 [webmasterworld.com]
May 10, 2011 [webmasterworld.com]
Jun 16, 2011 [webmasterworld.com]
Jul 23, 2011 [webmasterworld.com] ( our monthly Update thread)
Aug 12, 2011 [webmasterworld.com]
Sep 27, 2011 [webmasterworld.com]
Oct 13, 2011 [webmasterworld.com]
Nov 19, 2011 [webmasterworld.com] ( minor update )
[webmasterworld.com...]

Something is happening in the background - i can't believe that Google would let a whole period go from Oct 13 to Jan 19 with only a "minor" recalibration of the data on Nov 19 . That's way too long in Googleland.

[reseller]

@Whitey and @gyppo

You may wish to consider the possibility that there hasnīt been a Panda update in January 2012 and would be no more Panda updates in future.

Panda could be dead and buried somewhere in GooglePlex :-)

[DirigoDev]

Panda could be dead and buried somewhere in GooglePlex

I'd be okay with this outcome. RIP you ugly beast!

Odds are the black & white guy still lives and is ready to pounce on some more hard working/honest webmasters.

For the first time in years I lost money acquiring new customers last week. This is exactly where they want us. Destroy the SERPs and make us bleed on advertising.

[tedster]

Well, it seems clear that some kind of noticeable ranking shift happened on Jan 4, and another one on Jan 14. Right now, I'm leaning toward a backlink weight shift - but I can't say I'm 100% satisfied with that idea, either.

[Whitey]

Panda could be dead and buried

That's the issue .... that's it's morphed into the overall mix of things.

But i sense we may get a few more iterations to go with reduced effects, whilst it's features are migrated over steadily into the main algo settings. The 19th is a good day to pull the switch if you look at the history of the iterations. Time will tell.

[DirigoDev]

@tedster Agreed! Something did happen on the 4th and 14th. Real bad/highly untypical sales on the 15th led to the red ink. Bot sessions on the 15th were well below the normal range.

[snickles121]

My traffic has also increased by around 20 percent around the 4th or 5th of January primarily in Canada and Australia. No change in U.S. traffic.

I'm going have to say I don't think these ranking changes is related to a Panda update and I don't believe that one has been done this month because every time there is a panda update my adsense CTR goes through the roof. This month my CTR has been low with no jumps.

[alhutch]

@tedster I agree with ranking shifts. Our rankings went down slightly on Jan 4 but on the 14th increase substantially. I think the 4th and the 14th might be two different things based on how our rankings changed inversely and the shift on 4th slight and 14th shift had a much higher variance something I have not seen since panda in feb.