Google Updates and SERP Changes - February 2011

[tedster]

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

Related AdSense Farm Update < continued from [webmasterworld.com...] >


It's a new month and our regular SERP watching thread has been neglected a bit lately. Most of the posting right now is on dedicated threads focused on specific changes - especially the Scraping algo change [webmasterworld.com] and the promised (but not yet active) campaign against Content Farms [webmasterworld.com].

But Google's perpetual update machine keeps on turning. I'm particularly wondering about sites that publish a lot of legitimately syndicated content rather than a lot of original content. Did your rankings and traffic wobble with the "dupe content update"? If they dropped, did they rebound?

I'm watching one such website and though they are mostly republished content, nothing seemed to change.

[edited by: tedster at 6:00 pm (utc) on Feb 4, 2011]

[edited by: tedster at 8:38 pm (utc) on Feb 26, 2011]

[walkman]

Walkman , the same happened to me :
- when I search for site:www.domainname.com , www.domainname.com is on 5th place ( cache date is 26th Feb )
- the first 4 places are other pages www.domainname.com/a-page.html ( cache 23rd, 24th, 26th Feb )

Many years ago, and others can correct me, I saw this as a penalty. Some said it was links, some dupes...

Take a sentence from your index page and search for it, "in quotes." Maybe there's a dupe, somewhere?

[econman]

@tedster, robert and goodroi: I know it is a lot of work, but could you start a new thread for this update with a summary of your personal favorites from this thread to start clean?

I think it could be useful to start off a fresh thread with a sharper focus on what's really going on (as opposed to commiseration and misery loves company commentary).

Personally, I would like to see more brainstorming (and empirical observations) focused on trying to figure out what is going on behind the scenes:

What data is Google using in its new effort to detect "low quality"?

How has the algorithm changed in an effort to reduce combat "low quality"?

Is the focus primarily site-wide, or page-specific?

What is unique about sites/pages that are losing the most traffic?


If such a new thread were started, it might be helpful to include selected posts from some of the earlier threads (in addition to this one), including these threads:

Google Goes After AdSense Farms
[webmasterworld.com...]

Big changes promised shortly at Google
[webmasterworld.com...]

Matt Cutts: "chase after your best interpretation of what users want"
[webmasterworld.com...]

New Chrome extension puts content farms on notice!
[webmasterworld.com...]

and perhaps this one as well:
Matt Cutts: Google Algo Change Targets Dupe Content
[webmasterworld.com...]

It would also be helpful to gently suggest a few ground rules to help maintain focus.

[indyank]

One way by which google could have pushes the kind of sites I discuss is by turning off their authority. It looks likes google has retained authority for all high profile sites and they might have used the premium publisher list as a filter.

Are all the sites that people discuss here content based sites? Are people seeing any impact on e-commerce sites?

[krikomoush]

I know one premium publisher who called me and said they are down 30%... so it can't be that. I have checked a few e-commerce sites we work with, all seem to be fine, can someone chime in? Someone who actually sells and ships a widget and not just copy/pasting manufacture descriptions are you down as well?

[falsepositive]

It just smacks to me as something like this: individual page factors are involved, including a site wide penalty. Enough individual page factors are triggered and brings down your ranking -- which is evident say, when one page is heavily duped. Enough of this and your entire site is penalized and brought down. Some tests I did:

1. New pages I have, I cut off the RSS feed. That page shows up fine, ranks, no problem, albeit LOWER than I'd usually expect for keywords.

2. Old pages, heavily duped without my consent (copyright infringement). Scrapers are way way ahead, my article is much lower, sometimes nowhere to be seen. Difference between me and scraper? I have links out and in. Scrapers toss out all links and copy my work verbatim without links, or with minimal links (some with link back to my site). For my lost income, I am of the mind to SUE these people for the money I've lost. I don't believe I have seen a time in the past when a scraper outranked me. Well now I have.

3. I am knocked down a few ranks on some unimportant keyword on my domain name, but a main word, nevertheless. For example, if my site name was "The Content Farm Victim" (snark), my site used to rank #2 for "Content Farm". Now it's #6. So there's a site wide issue here. My guess is that because of the multitude of pages affected in #2, it triggered enough of a sitewide penalty to cause #3. If we fix #2, then #3 may be fixed as well.

4. Some of my best pages are just not visible anywhere. No scrapers are either. This is weird. They are just not ranked, although indexed, even by using the full title of the page. Many of these pages are my pride and joy, and I have no idea what's up with those. Again, perhaps because of the sitewide penalty, they are hit. Maybe because they are older pages, they could have stale or no links.

5. My huge authority site was hit. Small sites of friends in the same niche, NOT touched. My back link profile is excellent. I have links from outstanding places, and perhaps spammers alike. I am on DMOZ. Small sites from others? Not so much. Minimal link profile.

6. It may not look like overall link profile is a big factor, but possibly on page links are. I am totally white hat, always afraid I would be slapped by Google for a mere insinuation of a paid link. I've never gone there, so this is a bigger stab in the chest than it should be. It's truly a "punishing of the good".

All in all, a lot of work if we are to fix this. Starting with: go after your scrapers? I would think Google should already know what to do. They were supposed to release this algo along with the "duplicate screening algo". Maybe these algos work hand in hand. I know that the first phase of this thing involved clearing out the scrapers. I saw them disappear in a week or two. But I was not affected then. This time, I am seriously affected, and scrapers abound. So will this mean that the scrapers will disappear in a week or two?

Re figuring out the "quality" of a page via visitors. Scrapers have my content word for word, as well as the work of others. Wouldn't they have great user experience? What if they also have great site design, making them percolate to the top? There has to be other things involved like link profile, age(!), authority and those link backs I placed to my site, that some of these thieves still retain in the content.

Could this algorithm be one to sniff out the "original content source"? If an even bigger site copied you word for word, how would this algo recognize that?

There are certainly patterns you can see here, but not yet totally clear.

[ortelius]

I came across a blog post that has done a pretty good analysis of the top 25 sites hurt by the update so far. (Actually they have a list of 100, but just 25 posted in their blog.) The post is at [sistrix.com...]
The top 25 sites that they see as being hit the worst, in rank order:

ezinearticles.com
suite101.com
hubpages.com
buzzle.com
associatedcontent.com
freedownloadscenter.com
essortment.com
fixya.com
americantowns.com
lovetoknow.com
articlesbase.com
howtodothings.com
mahalo.com
business.com
doityourself.com
merchantcircle.com
thefind.com
findarticles.com
faqs.org
tradekey.com
answerbag.com
trails.com
examiner.com
allbusiness.com

Maybe we can see some commonalities here?

My site got hit -- 5 years old, all original content, about 5,000 pages, got pushed down, on average, about 5 ranking spots for every search term I can think of. Maybe a few just got pushed down 1 or 2 spots. But overall, a 40% loss in traffic, I'd say.

My marketing has included placing unique articles in sites like Ezine over the years. No more than 20 aricles, total. Maybe that has some impact?

My take on this is to try and get inside Google's head. If they want to bring down content farms like this top 25 list, what, in Google's opinion, would make a site look like a content farm? What factors would Google come up with that qualifies a site as a "content farm"? Maybe we can come up with a coherent list, and see if it matches the top 25 sites that have been hurt.

Two things on my own site I noticed that might make me look like a farm (even though I am far from it!):

1. A lot of my description meta tags, including an embarassing duplicate meta tag on a lot of pages, has buzz words like "tips, articles, etc."

2. My site is totally unique, hand-written content, researched and written by me. But I am trying to make $$, and the topics I choose to write about are good SEO terms. It is tightly-optimized. I also have a number of pages that are different spins on the same topic -- basically trying to capture all the search volume for related terms. Maybe I am over-optimized?

I agree with many of you that the update is multi-faceted. However, my gut tells me this might be more about on-page optimization than it is about links.

Oh, the Adsense issue -- I have a small amount of Adsense ads on the site, but mostly I run display ads from Ad Networks.

[falsepositive]

Ortelius, the part where you say "trying to make $$" may be a flag. I know I am. Colleagues who are less SEO savvy were NOT hit. I know of heavily SEO'd sites that are much smaller, in the same niche, that were NOT hit. I had prominent placements in highly competitive terms.

Some commonalities -- some of my content has been repurposed in labeled content farms (ezine, ehow, articlesbase). Has yours?

If your work has been regurgitated in these labeled content farms, maybe you get marked.

[edited by: falsepositive at 4:11 pm (utc) on Feb 26, 2011]

[ortelius]

I think it could be useful to start off a fresh thread with a sharper focus on what's really going on (as opposed to commiseration and misery loves company commentary).

Totally agree. We all got devastated, and right now the causes seem all over the map. Time to shake off the hangovers and get scientific about how we can analyze this and learn what to do.

I have no confidence Google is going to tweak this in our favor. More than anything, I think this was massive PR effort by Google to show the world, and Wall Street, they are serious about cleaning up spam. Clearly, it backfired to a big degree. But we are collateral damage. I doubt Google is going to put too much time into helping us.

Time to figure out the best approach to reverse engineer this mess ourselves.

[ortelius]

Some commonalities -- some of my content has been repurposed in labeled content farms (ezine, ehow, articlesbase). Has yours?

If your work has been regurgitated in these labeled content farms, maybe you get marked.

Yes, it has. Not a lot, maybe 2 or 3 dozen articles over 5 years in a few of those article sites, but I have done some of that. Haven't actually done it for a few years, because I now think it's a waste, but it is out there. The articles were "reworked," for sure. They are not dupe content by any means. But they are often on a similar theme, and often point to the same internal page or two on my site. Maybe this has a role.

[Murk]

Our main site has been around under same domain name and ownership since 1996 (orig. launched in 1994) and has lots of well-indexed legacy (aka evergreen) content. No spam, and the only "copied" content are a few PR releases that appear on other sites in the same vertical niche. Most traffic is from Google searches. On 1st page of Google for many keywords for years, and daily traffic has been boringly consistent (until this week, that is)

Here are the actual numbers [can you spot which day the algo changed?]:

Past Week - Total Pageviews - Unique Visits
Friday 25th - 12,640 - 8,152
Thursday 24th - 14,449 - 8,857
Wednesday 23rd - 18,209 - 12,010
Tuesday 22nd - 17,928 - 12,162
Monday 21st - 19,416 - 13,170
Sunday 20th - 18,504 - 13,013

[ortelius]

How about we form some kind of test teams, and after we wait a few weeks to see what settles, then we draw up a list of careful tests, and then divide up the tests among ourselves to see what we can figure out? With our common base of sites, we could implement carefully-controlled testing til we come up with the answer.

[Lenny2]

OK guys and gals here is my example.. I hope it helps... AND BTW, there are a ton of people asking for the positive side of the story... is there anybody out there?

We dropped 35% in total traffic, all google... which means we probably lost 50% of our google traffic.
We are an e commerce site
15,000+ pages
90% hand written content
Old website, been around since 2003

Our traffic came from both highly competitive key terms AND from long tail... we lost both...

Our site was built to help google determine which pages were what. We spent a great deal of time with making sure we were very clear to Google what each page was about... for both the traffic and for the user experience.

The insulting part of all of this is that google claims that the search results are better: "have you noticed better Google search results..." But, like many people in this forum are complaining about, there are many key terms in our industry that are ranking sites that scraped their content directly from us. In fact at least one that I saw even has a very subtle link posting back to our original content which is now on the 2nd or 3rd page.

[backdraft7]

If you are "trying to make $$" simply by plastering "original content" on thousands of pages filled with adsense ads and affiliate links, then that = content farm. The list above is a collection of classic, shameless self advertising efforts that offer nothing but spun content. When I go to these sites, they only give me enough info to piss me off and funnel me out of the site through their ads, to more relevant, complete informational sites or actual product vendors. This "extra step" or "middle manning" is what should be filtered. Google should come up with a new tab at the top left of the page titled "WGAS Expert Opinions" and filter all those "expert" commentary sites there.

[ortelius]

If you are "trying to make $$" simply by plastering "original content" on thousands of pages filled with adsense ads and affiliate links, then that = content farm.

You really ought to get more information before you sound off. My site is a tightly focused site on the topic of etymology, or the study of the origin and history of names. About 4,000 of my pages are name definition pages, each one carefully researched and written. In the offline world, it would be comparable to something like The Oxford Dictionary of First Names. If you are calling that classic work a "content farm" then so be it.

The remaining pages on my site are personally researched and written articles about names. I have put about ten hours a week into this site for the last five years. The fact that I make sure I write on topics that have some degree of search volume is, to me, fair recompense for the work I put it.

So in all, I have probably put in 2500+ hours on this site, which I love doing, and in return get a rather poor rate of return from ad revenue.

If that's a content farm, then I'm a farmer, and proud of it!

[indyank]

can you spot which day the algo changed?

It is obviously thursday and it does look like content based sites have been impacted in this update.

[ortelius]

If you are "trying to make $$" simply by plastering "original content" on thousands of pages filled with adsense ads and affiliate links, then that = content farm.

Follow up -- BTW -- I never ran a single affiliate ad on my site. The ads are from ad networks, display ads, just like CNN and everyone else.

[tranquilito]

I came across a blog post that has done a pretty good analysis of the top 25 sites hurt by the update so far. (Actually they have a list of 100, but just 25 posted in their blog.) The post is at [sistrix.com...]
The top 25 sites that they see as being hit the worst, in rank order:

ezinearticles.com
suite101.com
......

Part of the these domains do not rank with their home page on 1st position for site:www.domainname.com search ( freedownloadscenter.com, business.com, etc )

[chrisv1963]

If you are "trying to make $$" simply by plastering "original content" on thousands of pages filled with adsense ads and affiliate links, then that = content farm.

Niche websites with articles, written by professionals (e.g.doctors for a medical website), on one subject are not content farms.

[jk3210]

...there are a ton of people asking for the positive side of the story... is there anybody out there?

What time does the Ferrari dealer open on Monday morning?

[edited by: jk3210 at 5:00 pm (utc) on Feb 26, 2011]

[walkman]

Part of the these domains do not rank with their home page on 1st position for site:www.domainname.com search ( freedownloadscenter.com, business.com, etc )

Great find. Mahalo also doesn't even rank #1 for Mahalo. From my IP they rank #3 for it, after Wikipedia and after an angelfire site that has Hawaiian phrases, a very bad thing from trustrank. Maybe that's the difference between manual and algo penalties /demotions?

[dickbaker]

Just to chime in on others' observations: when I do a site: search on my top level domain (www.mysite.com), the home page comes in at #23. The first 22 results are third or fourth level pages (three or four clicks from the home page), and most of them are third-level pages that aren't as important or as heavily viewed as other third-level pages.

I'm searching now for scraped content, and finding a ton of it. I'm searching for the first 15 or so words (in quote marks) from my main content pages (the manufacturers pages), and I'm finding some instances where other sites show up for the quoted text but my site does not.

Regarding ezinearticles.com and articledashboard.com, I've submitted a fair number of articles to both, but I don't have the article content on my site.

[falsepositive]

dickbaker, we reside in different niches, but your profile and situation sounds very similar to mine. I wonder if in fact, Google knew what they were doing... could they be using certain sites as examples for a training algo? I mentioned I noticed visits from Mt. View yesterday and some bot activity ongoing. Perhaps they choose scapegoats and watch/track/monitor them to see how things change over time? Not that they choose the scapegoats manually (they could very well be sitewide automatic triggers that are imposed on some sites but that they notice and track).

And my traffic is recovering, but very gradually. But it can just be me, doing my own wishful thinking.

[tedster]

walkman posted:

Someone checked some data and
Mahalo
EzineArticles
Squidoo
Assoc Content
Buzzle
HubPages
have been hit with 1-100+ downgrade

Here's another big data crunch report - very similar to walkman's list but with a few others: [sistrix.com...]

[alecs]

Ok it all sounds scary, i experienced a 30 % drop in traffic, 50 % drop in adsense, a do not use google analytics. However there are a few things that make me think :

1) after year and years of tweaking their algo i refuse to believe that this is the best they can do(do you really think google has arrived to the conslusion that scrapers or twitter pages should rank better than the original sites?) so i tend to believe the whole bubble sort is not over yet, and the fact that crappy sites appear in top results is because we only see the bottom part of the piramid at this point.

2) there are simply to many people complaining about traffic and revenue loss, with all kinds of websites and from various niches

3) what is also curious is that the long tail keywords have dropped but the main keywords remained the same...at least in my situation.

There are too many things that simply do not make sense, so i think it is better to wait and see how it will all look like when it settles down.

[backdraft7]

You really ought to get more information before you sound off.

Do you sell a branded product or service? Or are you simply writing a little on a lot of subjects for the sole purpose of luring visitors to your ads? That's a content farm in my book, and they've become a dime a dozen. I know it's hard to accept when you've spent 10 years spinning content to lure in users and search engines, but that's no longer working. Bury you head in the sand and believe in your business model or adapt. It looks like way too many people bought that "get rich quick on Google" ebook.

I doubt you can compare yourself to CNN. Their original content is frequently the source of many content farms. That and true authority sites. If you're writing tens of thousands of one page articles and surrounding the with ads, face it, you've built a content farm and complaining about losing ground is pointless.

Why don't YOU define content farm to me...I await your reply.

[backdraft7]

@indy - for once I have not been affected. In fact I now occupy up to FOUR consecutive spaces (on keywords we discussed). Even though it's my site, I can't say I'm happy about those listings because it looks like Google is still broken...except this time (out of many) it winds up in my favor....for now.

Everyone needs to just relax...I've never seen so many WebmasterWorld new users / lurkers come out of the woodwork. You'll soon learn that what happens today will change tomorrow.

[Jessica97]

Quick update on what I'm seeing:

a: syndicated content searching for a title from duplicated articles still rank usually at least the top 1-4 on most searches sometimes 1-6 or more (even though the exact same content is on up to 15k sites ---word for word) But granted this is a different type of long tail search, done for an article title, it's not a product or something with high commercial intent.

b. different types of terms, items or products with commercial intent, so for example some of them mentioning specific product names -- one of which has been mentioned in this thread already appear to rank very well (#1) for spammy, auto generated pages - keyword stuffed that have lots of content on them but the content makes no sense. Shouldn't these be the types of sites that are targeted for low quality?

c. If it wasn't so sad it would be amusing. But I'm getting comments and questions from non techy users who don't know anything about all this asking me why they can't find what they're looking for now using Google. Just in the past few days...so if even end users who don't know how to get anything back from a search engine other than the "search term" that they put into the search box can't find what they're looking for--that's not a good sign.

d. I had a few personal sites that I have done nothing to for at least 6 months, some a year. Just noticed that one of these has gone from position #3 to position #1 for it's main keyword. The others have either maintained their results or gone up --no updated material and they were actually setup to post some info to be shared for meetings that I attended long ago, so more an informational base, but nothing new has been added, no links, nothing.

[edited by: Jessica97 at 6:06 pm (utc) on Feb 26, 2011]

[chrisv1963]

i tend to believe the whole bubble sort is not over yet

I agree. I think Google is still tweaking or re-calculating. USA traffic to my site improved today.

[FredOPC]

I have written here a lot before, but here's some more data for folks:

1. Our site is ~1000 pages, all hand written. Site is a blog

2. Some pages were "posts" designed for a short time that we never took down (e.g. giveaway announcements, weekend round-up style posts, etc.).

3. Many pages are highly researched articles with "how-to" information that was very detailed and useful (not eHow-esque).

4. We use Adsense , Impressions, and Affiliates to monetize depending on the page and what makes sense.

5. We are down an average of 5-8 spaces on rankings, sitewide (long tail and short).

6. We still rank strong for our domain name, and overall, we haven't been "obliterated" from the rankings, especially where competition is weaker in the search. Generally, scrapers aren't outranking us.


I have looked at tons of medium-sized sites across three big niches. I am at the conclusion that:

1) This is a site-wide penalty that impacts the "score" of the site. The worse the site gets evaluated for this penalty, the more points it loses in the rankings. This site-wide penalty (or negative "score") is factored in with Google's other traditional factors, probably all of which were not very affected by this change.

2) This penalty can grow with the size of the site. Smaller sites (sites with 5-20 pages max) are not affected in any material way by this change. In fact, sites with rather spammy link profiles or scrapers stealing a small amount of content can be doing better than other larger sites because of this. The larger the site, the bigger the potential penalty...

3) Google is looking to kill content farms. The "old" Google was of the mindset: "Upload 100,000 pages and we'll find what's good based on our algorithm". The "new" Google is saying: "The more pages you upload, the more you must demonstrate your site's overall value in aggregate." This resonates with a couple of things that came out on the official google blog release. Matt Cutts points out that Google has a responsibility to "encourage a healthy web ecosystem". What better way to encourage this ecosystem than to penalize sites for putting up 1000s of spammy pages? This forces responsibility among webmasters.

Matt also says that this will reward "sites with original content and information such as research, in-depth reports, thoughtful analysis and so on"

Note that Matt's comments are all about SITES, not PAGES. Sites that spend time putting together well-researched content with exhibit different factors than content-spinners. Content spinners rely on saying the same things 100 different ways but without adding any real value.

There are probably a variety of factors Google can examine to make this site-wide determination, but I think a fundamental factor is the % of "worthwhile" articles vs. "spammy" articles on the site. How google makes this determination: who knows... could be links, could be bounce rates, could be a mixture of these.

This is hard for sites that have focused on satisfying long-tail queries with really useful information but may not exhibit whatever ranking signal google is looking for. These sites (including my own) are collateral damage for this problem. But, I believe that Google now prefers sites with a much higher percentage of "valuable" pages. If you don't meet the percentage value determination, whatever that is, you get whacked.

Note that this is VERY different from the blanket Google approach of the past, which basically said that the more links you get into a domain or its pages, the more "authority" the domain has, and therefore the more likely new pages and content are going to rank. Content Farms lives on this approach. Google had to find a way to poison the well for these sites.

That's my take. I could be wrong, but so far, it makes sense to me.

[zoltan]

So, what is actually the definition of a content farm? Is a well organized (and of course SEO-d) niche classified ad website (pets or autos or properties or anything else, or even a general classified site) filled with tens of thousands of ads considered a content farm? If so, google base is also a content farm... wrong?