Panda and The Incredibly Shrinking Internet?

[seoskunk]

In an effort to recover from Panda I have taken 7K pages off my site and stripped it back to basics. I was wondering how many other have removed content due to Panda and despite the automated blogs has the internet slowed or decreased in size at all?

[tedster]

I've seen no evidence of overall shrinkage. As much as our attention here has been on Panda, it's still been a pretty small sliver of all websites - and many online businesses have still never heard of "Panda".

I've been working with one site that really needs to remove 90% of their content - it was the kind of thing where they created one page for each and every keyword phrase they thought could bring traffic. Deadly stuff for Panda.

[chrisv1963]

I've seen no evidence of overall shrinkage.

I agree. Copyright violators are still very active stealing our content, creating new pages with it and getting good rankings ...

[suggy]

30,000 pages gone. We had followed a similar strategy to that described by tedster. Had worked; now killing us!

[Whitey]

Well in some verticals the reduction in content has been enormous due to those effected. Typically affiliate groupings, but many serving up quality, but stale content.

We're talking big no's - 60k+ networks of affiliates down to less than 15k producing, the others producing zip. Multiply that out across the web and you have some serious attrition which won't come back. And there's probably more to come.

In my mind this is Google upgrading "quality" with a view to concidentally forcing the distribtion sytem via it's own assetts eventually.

I can't put a number on it, but it is likely substantial in concentrated areas, which makes the magic 12% considerably higher in some or those verticals IMO

[george_1]

"
I've been working with one site that really needs to remove 90% of their content - it was the kind of thing where they created one page for each and every keyword phrase they thought could bring traffic. Deadly stuff for Panda."

on the other hand I monitor a competitor website that does that thing over 10 years now! They never have been touched by any earthquake updates like Florida, brandy, caffeine, panda. They monopolise a whole expensive niche with thousands of keyword phrases even with empty content pages!
PS: they never used adwords or adsense!
Probably they will finally get caught by the IRS.......lol

[george_1]

oh..by the way instead of shrinking the above website within 1 year explode from 100000 to over 250000 pages!

[rango]

I haven't really kept count, but we've removed 10s of thousands of pages. And noindexed millions more (individual photo pages mainly). Should have done it a long time ago really. Crossing fingers that it will actually help in the SEO stakes, but at least it's cleaning up the site a little which is certainly a good thing.

[realmaverick]

Several million. Though many were just junk index.php types.

Also noindexed over 2 million profile pages, most of which, were indexed.

[serpsup]

Tedster, could you elaborate on your technique for removing the 90% of the content? I'm in a similar boat, I'm an ecommerce (comparison shopping) affiliate in a very specific niche. Our search results pages are indexed which creates a lot of keyword rich urls. This used to be a great thing but I highly suspect this is a key indicator for Panda. I am thinking of coming up with a list of the top 500-1000 or so Category-level keywords based on what urls Google used to love and still loves and for all others just adding a meta noindex. The search results pages on our site are pretty much the main part of our site so I can't do a 404 / 410. Does this seem like a good plan? Any insights would be welcome.

[Planet13]

Also noindexed over 2 million profile pages, most of which, were indexed.

That seems like a good idea since indexed profile pages are a spammer's paradise.

[tedster]

The search results pages on our site are pretty much the main part of our site

Then IMO it sounds like you might need a serious redevelopment if you hope to compete again for Google traffic. Surely the search results pages are lists of content pages, right?

[tedster]

could you elaborate on your technique for removing the 90% of the content?

No real technique as such - just a ruthless "machete", tuned to Google's descriptions of what Panda was trying to accomplish.

[jmorgan]

I keep reading of people trimming pages of their site but never of people saying that this method has actually worked in terms of counterring the effects of Panda.

All I ever read is, "Got rid of X amount of pages. Now hoping it works ..."

Why do people keep doing it if nobody has actually shown any evidence that it can or will work?

[rango]

@jmorgan - there have been a few people in this thread [webmasterworld.com...] who have specifically mentioned removing lots of pages and being released from Panda.

Of course, they did lots of other things also, so it's not necessarily *the* solution. But since none of us really know what exactly is *the* solution, this has to be one of the things to try.

[tedster]

Panda is all about developing a machine intelligence that measures content "quality". So if you know you have some real low quality junk on the site (Amit and Matt described it very clearly) and you also got nailed by Panda - it seems pretty obvious to me that you either get rid of that, or ideally upgrade it.

Anything else is just looking for a way to "beat the algorithm" - and even if you find it, that's not likely to be very future-proof. The rule for me is "don't cut any corners."

[serpsup]

Tedster:

Then IMO it sounds like you might need a serious redevelopment if you hope to compete again for Google traffic. Surely the search results pages are lists of content pages, right?

Hey tedster, thanks for your reply. Let me elaborate - my site consists of product pages to buy a widget (which are affiliate links) and search results pages which are pictures and links back to the product pages. There are also some hand curated lists and ways to browse through the site by category etc.

Since the product in our niche can range across a variety of topics such as movies and TV shows it is common that we have something for just about any popular "keyword widget". So these pages are dynamically rendered but could just as easily be considered "category" pages for a typical ecommerce store. Also, if you take a look at some of the biggest comparison shopping sites it is common practice for their search results pages to be some of the most prevalent in the SERPS (at least before Panda). This may also be related to why comparison shopping engines in general took a huge hit b/c of Panda... but there are still a few giants that seem unaffected.

So at this point I have a ton of long tail keywords combos of search results pages in the index. I am considering something drastic at this point such as either noindexing them all and 301ing just a few of the most popular to category page equivalents (basically same thing by a different name), or just noindexing 95% of the current search results and just keeping the most common / popular. Any insights would be greatly appreciated.

Thanks.

[fathom]

A thin page may not be what you think.

Huge article websites lost massively due to PANDA where article pages have lots of text. The quantity of pages compared to the available link juice is the problem.

Thin is a metaphoric description.

WebmasterWorld itself has also been effected by PANDA... the difference... repetitive pages that don't deserve ranks are not naturally promoted.

e.g. "I lost my ranks" threads get indexed but are buried below the mix of WebmasterWorld quality threads.

An educated guess WebmasterWorld has be harmed just like you out of 2 million threads (a guess) WebmasterWorld gets traffic on say 10% of them... the fact that they always got that means they were pre-PANDAized and it wasn't a huge problem because they didn't rely on traffic from "I lost my traffic Jan 12, 2011" & "I lost my ranks Nov 1, 2009"... that's the difference.

All domains are the same we have all been PANDAized some just feel the effects more than others because of (for lack of better words) regurgitated pages to target every longtail phrase was the model of choice... but now it does not work.

That said... all the pages are still there (indexed) and if they deserved ranks they would each have a few links to them. Granted you can't develop 3 million links for 1 million pages but you don't need to... just pick your ponies wisely.

[chrisv1963]

The quantity of pages compared to the available link juice is the problem

Is that internal or external link juice? I have the impression that my main website is slowly recovering from Panda since I started doing intensive internal link building.

[fathom]

Is that internal or external link juice? I have the impression that my main website is slowly recovering from Panda since I started doing intensive internal link building.

Not easy to explain without visual aids but...

Try not to think of this as internal or external nor think in terms of PageRank... link juice propagation is best understood through architecture and as such internal links can both bolster and hamper page quality regardless of what value external links deliver.

There are two completely different thought processes commonly associated with SEO architectures.

Since PANDA (as Google explains) is distinguishing between high & low quality content let's stick with that and call these SEO high & low architectures.

Low Quality Architecture: is about repetitive anchoring on a specific page pointing similar styled anchors to other pages.

High Quality Architecture: is about succinct anchoring on a specific page allowing each new link generation (above or below) to gain prominence by compounding values of partial matches.

If the target is Search Engine Optimization (poor example but gets the point across)

Page 1 Search
Page 2 Engine
Page 3 Optimization

You can also do

Page 1 Search
Page 2 Search Engine
Page 3 Search Engine Optimization

But in the latter you limit yourself in the realm of longtail targeting which is where PANDA is killing domain pages off... and because of the loses associated with low level pages all the internal links back to top level pages also disappear which compounds the traffic loses.

BTW... IMHO PANDA is Supplemental Results in drag

[Uber_SEO]

I took a Panda effected website from ~2 million pages down to ~100,000. That was about 6 months ago, and it's made no difference

[petehall]

BTW... IMHO PANDA is Supplemental Results in drag

I agree with this entirely.

[enigma1]

they created one page for each and every keyword phrase they thought could bring traffic. Deadly stuff for Panda.

Yes, I guess they now need to merge the content back, only it needs manual labor to have some quality in it.

[suggy]

they created one page for each and every keyword phrase they thought could bring traffic. Deadly stuff for Panda.

This is probably the root cause of my own pandalisation.

Many of the affected pages weren't thin; they were just too closely related to other pages and lacking any real link juice.

[RP_Joe]

I find it hard to believe that people have over 10,000 pages of quality content.

[petehall]

WebmasterWorld has around 490,000 pages in Google at the moment. When your audience is writing and creating content pages for you, it's easy to see how the numbers rack up quickly.

[enigma1]

I find it hard to believe that people have over 10,000 pages of quality content.

Quality content doesn't mean 10,000 pages of unique articles. You could have 10,000 products and have lots of customers because of good prices. Or you can have 10,0000 images others find useful, or a forum etc.

[fathom]

I find it hard to believe that people have over 10,000 pages of quality content.

We need to put things in PANDA perspective. All domains are impacted including YouTube... YouTube has tons of low quality videos that Google does not want to rise to the top... but YouTube's value to the web isn't the millions or billions of videos but the 10K or 20K really exceptional ones and those are generally determined by natural links pointing to them.

A few pages of gems can easily support 10,000 pages of crap. It's the gems that drives your ranks and thus traffic. People got stuck on longtail because you can rank crap easily for 1 or 2 clicks but Google reset that to stop the crap from appearing just because you can manipulate with a few tricks.