Panda Update 21 - 6 November 2012

[klark0]

<moved from another location >

Panda update #21

[searchengineland.com...]

[edited by: tedster at 9:20 pm (utc) on Nov 6, 2012]

[HuskyPup]

I'm in the UK, Google.co.uk has been an almighty mess here for months for my widgets and I don't mean they are not returning my results in the SERPs, they are simply showing loads of completely irrelevant supplier results from companies that would not have a cat in hells chance of supplying.

Meanwhile Bing's results look very clean and relevant.

[superclown2]

I'm in the UK, Google.co.uk has been an almighty mess here for months for my widgets and I don't mean they are not returning my results in the SERPs, they are simply showing loads of completely irrelevant supplier results from companies that would not have a cat in hells chance of supplying

I'm in the UK too and I have to agree. This last update has produced results which are the sort of junk we got years ago with many completely irrelevant sites with most of the search term words absent in either the text or backlinks and shoals of sites riding high on comment spam. Unbelievable.

Surely this cannot be the finished product. Now where did I put that copy of Xrumer?

[7_Driver]

No improvement on Panda hit site here (UK) - despite getting rid of lots of pages I thought Panda might not like.

UK Serps a mess? Perhaps that's why Google market share has just dipped below 90% for October. That's the first time in 5 years they've had less than 90% share.

(Perhaps it's Windows 8). Either way, I'm not shedding any tears for them.

[diberry]

The SERPs are getting pretty strange in the US as well. I've been using Bing to check election results, but this morning I used Google to search for whether a candidate had won and got nothing but 5+ day old articles about the candidate. Back to Bing, and the first article answered my question.

Pretty strange for a search engine; even stranger if they want to become a one-stop shop for answers.

[Lorel]

My traffic has been down for several months and steadily dropping even more for the last month and suddenly took an upturn on Saturday the 3rd by about 30%. This occurred on another site that had traffic steadily going down with a 30% increase also starting on Oct 3rd. No major changes on either site other than occasionally adding a new page.

[Nostalgic Dave]

It looks like Google was just teasing me with the traffic spike over the weekend, because it's all back to "normal" now. Does the Panda algo test a site? Throw some traffic at it and see what happens before lifting the penalty for real? Probably just a side effect of fluxing from other sites moving around... bummer! My ads are ALL far below the fold, I've got rel=canonical on all my attachment pages (pointing back to the parent), I've removed thin pages. Guess it's not good enough for the mighty Panda.

[xcoder]

Traffic is generally up for us (over a number of websites) but it is a non converting foreign traffic in most cases.

This is beyond "incomprehensible"...

[MrSavage]

I have done a couple massive, and I mean massive overhauls on two or three sites. I would be hard pressed to do much else frankly. Given the time frame of most of these changes, I'm just not sure if Google has systems in place as to not be too transparent. In other words, if I was just to do what I did and on the first Panda rerun my site returns, then that's too dangerous. I'm looking for a possible next Panda or 2 months from now change in fortunes.

Think about it. If making tweaks over a couple weeks saw bounce backs on the date of a next update, wouldn't we all be here advertising that fact? I know that the last thing an algo needs is for people to figure it out. So I fully expect a return to glory to be buried enough that I won't be able to say exactly what it was or which update specifically allowed me to return.

Trust me, I have one or two throw away sites which I have been able to cleanse without hesitation. If in the next 2 to 3 months my rankings can't return, then this is far more tragic than I first thought. I know I'm not the only one with disposable sites that have allowed for mass de-SEOing, so please advertise the success in doing that. I'm all ears.

[Sand]

I saw no change over the weekend. Then yesterday was a pretty good day, but it was dampened some on account of the US elections. Today things have really picked up for me (+20%) over previous Wednesdays so far.

I was previously hit by Panda in April, and then mostly recovered this fall. Now I'm getting really close to a full recovery (despite removing about half of my site's content).

Anyway, there's no real point to this post aside from sharing an observation. Some people reported seeing gains over the weekend, but I didn't start seeing them until Tuesday.

[MikeNoLastName]

I noticed the same effect that ichthyous indicated, in fact that is why I came in today, to see if anyone had an idea what might be going on. The WMT HTML improvements report showed significantly fewer duplicate titles and descriptions which we admittedly DO still have some of (archived 10+ year old weekly columns with the same column title and author), although the rest of the page is all different. Not sure if it is related but in the last few days the number went from around 250 on each report down to just under 50. Except we haven't done ANYTHING about them in the last couple months, and based on the list I printed out to work off of, there should still be around 235. Maybe they no longer consider these as a negative? I remember reading or listening to a very recent interview with Matt where he said something like: one shouldn't worry too much these days about duplication within the SAME site. Maybe they've just switched to ignoring them rather than penalizing?
I also noticed that a lot of these disappearing pages and those closely related are STILL ALL indexed and found by literal strings from them, but are now coming up on the Google PR tool as "Not Ranked" or 0's where they all used to be 1's. On the other hand, many of our other pages improved in the SERPS and the photos on these "duplicate-title" pages are as popular as ever in the image searches. We haven't made any significant changes to the site in months, except removing a single link to an offsite page from a navigation bar from about 50% of the pages.
Overall, between this heading into a annual normally slow period for us, the election yesterday, days of the week, etc. I just don't have enough steady data yet to tell if it was good or bad in total.

[gouri]

We haven't made any significant changes to the site in months, except removing a single link to an offsite page from a navigation bar from about 50% of the pages.

@MikeNoLastName,

Can you describe the link that you removed from the navigation bar? Was it a keyword phrase that the page that you linked to is trying to rank for? Was it a url link?

[gingersnaps]

Does anyone know if its done rolling out?

[Whitey]

.. We made purely cosmetic changes. With a more modern look we saw a quick recovery.

The site I referred to above was ... dated, ..less so than the first site that was hit. .... cosmetic changes to its appearance and ... also improved navigation, which had been .. weak. This time we had to wait longer for recovery, which I hope won't be short-term!

In both cases, we made no changes to core site content, which we believe to be strong.

@Doug10 - Recovery posts like this are gold. If only we had more folks with good news stories. And you have 2 sites out of the Panda's grip.

How long did it take for you adjustments to be recognised and worked on?

[sunnyujjawal]

Does anyone know if its done rolling out?

As per my rankings..earlier they got down (from page 1 to page 3) in Google India but now they are little bit back just 1 or 2 rankings down.
I think still rolling out

[Martin Ice Web]

Also I donīt know if other languages are updated, we see some good and converting traffic since Saturday.

But I also see that the google referrers come from >10 ( means page 2 ). That means users donīt find matching sites on page 1.

[Splugged]

@Martin

I just looked at my stats: 50% of G referrers come from page 5,7,9 and 10 (#52,#76,#91 and #101). Never seen visitors come from pages so "deep".

At this point... who's wrong?

[menntarra 34]

My sites have been hit by several panda, but in the last 3 pandas i always recovered some 3-6 percent.
I checked my google traffic today:
Saturday: highest traffic in 3 months
Sunday: highest traffic in 3 months
Monday: traffic is as ususal(back to the same level as in the last month)
Tuesday: traffic lower than the last 3 Tuesdays
Wednesday: i don't know if i can rely on the wednesday traffic stats, but it seems google traffic is down by 20%

[Martin Ice Web]

splugged, it is a wonder that poeple still using google.

I realy appreciate that google wipped out a lot of black hat sites with cloacking, redirects ... but the collateral damage is way to big. That panda is all about content or UE is imo wrong. I think its way larger like 404s, html coding style/compatibility...

[klark0]

Does any one notice that sites hit by Panda usually have a high "not selected" figure in Index Status. Usually higher the number of indexed pages.

[renomart]

Yes klark0, my site has been hit by Panda. I have 700 pages indexed according to the Index Status and an outrageous 790,000! urls "Not Selected".

[MelissaLB]

yes Klark0, I beleive my site has been hit by Panda and we have approx. 7000 pages indexed and a whopping 2.5mil urls not selected.

This is all from a fix we implemented over a year ago to sort our erroneous urls that were somehow being generated, by using 301's to have only the 'good' urls indexed.

hrm....

[zeus]

"Not Selected" what do youguys mean by that

[backdraft7]

I curious if anyone else is seeing a odd number of referrals from 172.27.72.27. Traffic is on a roller coaster this morning, sitting at zero for 10 minutes, then a rush of 10 to 20, then back to zero. Not a typical pattern. Oh, that's right, it's THURSDAY!

[MelissaLB]

zeus>> if you go into your google webmaster tools and select Health>Index Status, then select 'Advanced' and you will see a row of checked boxes. It should show you Indexed, Not Selected and Blocked By Robots.

[chalkywhite]

I have 800 pages indexed and......800 pages not selected, I though it was canocial but after a full check ove no issues. Well spotted klark0 we may be seeing a trend here

[bloard]

A point that might or might not be relevant. My site was hit by October 2011 Panda. Just prior to that time my ratio was 1:1. I had roughly 8,000 indexed and 8,000 not selected. As part of my Panda "recovery" strategy late last year I started deleting poor quality sections of my site.

Now I have 11,000 not selected and 2,000 indexed. My site has continued to loose rankings. Initial Panda cost me maybe 50% google traffic. Now I'm down 90%, and the one constant during that period is that this ratio has continue to increase.

[diberry]

I just looked at my stats: 50% of G referrers come from page 5,7,9 and 10 (#52,#76,#91 and #101). Never seen visitors come from pages so "deep".

I've been seeing this recently, too. And it's not always for long tail phrases, either - it's like people are just going through a lot of pages to find what they want.

[backdraft7]

@ diberry - but are those deep searchers converting? It almost seems like there's a robot doing the digging rather than a human. I can't imagine a human going through that much trouble.

[zeus]

hmm 101.000not indexed and 700 indexed, but I guess that just be cause that removed A LOT of pages from my site, but ok never in that range, I think my site had before 25.000 now 1000.

[7_Driver]

A year ago (6 months after Panda) - I had 195,000 pages indexed and 380,000 pages not selected.

Then Not Selected exploded up to over 1,000,000.

It has come down in stages, and is now back to 350,000.

All these figures are WAY more pages than I actually have. I've deleted loads of pages - and total indexed is now 170,000 (so "Not Selected" is still about double).

If Google would provide a list of "Not Selected" or "Total Indexed" pages - that would be useful, and would help us clean up their index so they only see real pages. Which is presumably why they don't do it! Without that info - the graphs indicate there may be a problem, but give you nothing useful to help you solve it.