Welcome to WebmasterWorld Guest from 18.104.22.168
I just got my site rank #31 on its own domain name and bunch of keywords/phrases I usually watch were bumped from #1 to precisely #31. Those #2 through #10 are sort of all over the map but generally within the first 60 results.
Does anyone have some experience with this? What would the respectful audience here think a most likely reason for such penalty is? What do you suggest as the best strategy to fix this?
There has not been any major redesign recently, just routine adding pages here and there. Some unique, some syndicated industry-related content.
Thanks for any idea or comment!
duplicate content caused by bad content management system
big links campaign
errors in html
keyword anchor spam
Usually, you need at least 3 of these going on big time. I HAVE heard of people recovering, and its by fixing these issues.
If your site is one big affiliate site, and has a massive duplicate anchor site map.
eg red widgets
greeny blue widgets
bluey green widgets
widgety widgida widgets
You've tripped a filter
[edited by: tedster at 1:16 am (utc) on Oct. 17, 2006]
You actually have to have a kick a.. site to get hit with just 30. If your site is of a lesser Google-perceived value (younger, less PR, less unique vs. original content ratio), you will get hit with "minus 54". I just had three more of my sites bite the dust with "minus 54" laid upon them. It's actually even harsher than just that: "minus 54" is when you search for your domain name. None of the actual keywords are to be found anywhere in the first 100.
My site is 8 years old.
Basically I did implemented mod_rewrite, which then meant googlebot crawled thousands (eg 5000 per day) of pages, mostly duplicate content.
About ten days later I noticed the drop in ranking for a 1 word search term of about 30 spots - ~30 to ~60. Other search terms weren't really affected.
I then excluded those pages via robots.txt.
Changing URLs will drop your site in the rankings. It seems to me that it would be better to redirect from the old to new with 301s, not block the old URLs with robots.txt.
Try an "almost unique" search from your site on Google that returns Less than 30 results. For any such term your site should be always BOTTOM or thereabouts for this Serp?
thats a good idea for you guys, whats the results?
This thread is also very interesting to me too. I had no IDEA about this kind of junk and actually did some promoting with anchor texts like this weekend for about maybe 10 sites to mine, hope thats not excessive.
I had no clue, although I can't say I felt completely legimate doing it.
my site is #2 of google, only 2 competing sites. So this will be interesting, if $#@$ hits the fan, at least I've learned.
I've seen sites come in and go out, the guys never change anything on the sites.
I actually think it's something along the lines of google have placed a? over site and it needs to build up some trust to be allowed back up.
[edited by: UK_Web_Guy at 11:01 am (utc) on Oct. 17, 2006]
.. this is exactly what I wanted to ask! Let's please clarify the effects for a moment:
1. Are you >30 for every single phrase or just some phrases?
2. If your domain is xyz.com and you search for "xyz" ... are you #31 for that phrase as well?
3. Is it only for competitive phrases?
4. Or is it only for phrases you've specifically targeted and perhaps over-optimized for?
5. Is this the same thing as "the sandbox" everyone loves so much?
1. Of 10 phrases I have just searched for where I would expect to be on the first page I was number 31 all 10 times.
2. If my pagee was xyz.com I am number 31 for "xyz"
3. It is any phrase even obscure ones, if there are less than 30 results returned I am last
4. No, all phrases
5. No, nothing to do with the sandbox. Mine is a 5 Year old site that used to be PR6 (now 5). It is still crawled regularly and PR hasn't dropped in over 12 months.
Its difficult to quanitify "did nothing" as you can not actually "do nothing"
50 sites may have added links to you, which tipped you back over the not trusted/trusted limit.
then in in the "did little" category
a slight content change may have fixed things.
a slight update may have removed a hidden link orphaned in code you were not aware of.
yes, its being going on for years, I had a site tank for 3 months(+30) 2 years ago, issues then were.
a. added 300 recip links.
b. home page sitemap with 400 links in it, all with one word the same.
c. hiddenish text.
fixed all, added 100 more links, removed the ones where no link back, site recovered.
Well... I would have to agree with the person above who said that suddenly something happened and they've tweaked and analyzed endlessly and nothing happened. That's clearly a penalty, not a algorithmic shuffle. Now, the argument I guess is whether it was manually imposed or not. I believe it is Google's *INTENT* to automate everything but I can speak from personal experience on my own sites and tell you that when my algos are not ready or not yet up to snuff, I often fill the gap with manual edits ... otherwise my community will falter. So I do not think manual edits are out of the realm of possibility, regardless of what the Google press releases may say.
...this is very unfortunate. I really do wish Google would provide better webmaster relations on such things. Such a heavy handed approach needs to be used very delicately or it can ruin what people have worked very hard to create. Same complaint as always about Google in that regard.
That is exactly the point. Many of us webmasters have worked for years to build the "trust factor" with Google only to be nailed with a penalty and we have no idea what triggered it. Yet Google keeps saying that they are trying to open up lines of communication with legitimate webmasters, notifying them of reasons for penalties. Well I have been waiting for 10 months to hear from that higher authority and not one tidbit. I certainly have tried to communicate with them but they have not reciprocated.
It seems to me that Google can easily identify legitimate websites that have incurred a penalty and then communicate with them. For example, a website that 1) is at least 5 years old, 2) that never before was assessed a penalty, 3) with a true PR5 or higher and 3) an presently-assessed penalty would be an ideal candidate for communication from Google. That would screen out all the whiners and provide a tremendous service to the webmaster.
If the webmaster then did not remove the reasons for the penalty within a time certain, then they shouldn't just be penalized, they should be banned.
Seems easy and fair to me. I certainly volunteer to be the guinea pif.
Even as one of the people lobbying for hybrid results...
I honestly... not making it up... i whole heartedly Laughed Out Loud X)
Missed this on sept 1st when it was news ( original thread: [webmasterworld.com ] )... for i had barely joined and was preoccupied with our site's problems... but now it's on webmaster tools asking whether we'd like to participate with the images of our site allowing others to decide... well basically the fate of the pictures. ( Okay, exaggerating, it's only beta and all but... but you see, pictures are pretty much THE things we have on our website so it makes me wonder. )
I mean i, for one, have always thought there should be a fine tuning mechanism involving a board of people...
- Who don't know each other
- Who have no commercial interest in corrupting the data
- Who see the pages popped up random for them
- rank for design, ethics, usability
- and are reviewed by each other randomly in at least three rounds...
I've been wondering what it would cost to hire such personnel and how employment could be dealt with to achieve full anonymity... And never thought how EASY the real answer was.
It was right before my eyes, how come i never thought of this?
There're tens of thousands of BORED people surfing Google every other second! Let THEM decide! :)
( call them googlaholics )
( btw, read the sep 1st thread here on WW, it's funny )
I just really hope now that they make the same system for websites in general. Perhaps there aleady is such a system... perhaps Google has always been building its database manually ;)
Hehee... this is fun. Sorry this is news to me, i just noticed, don't know how i could've missed it so far...
It works 100% the way i envisioned the system for reviewing spammy sites only it's in reverse drive! And i've never would have thought it would be us, the visitors doing the job... i mean wow. I can't tell a good business model when it says "good business model" stamped all over it.
There should be a back-end to Google. Half of the world would be typing in search queries on one side, while the rest would be typing in the meta-keywords on the other!
( though the sept thread mentions this as well... oh okay, i'll shut my mouth now. But really... )
This is... The best idea... and at the same time...
...the best joke i've seen in years.
[wipes tears from eyes]
I've seen such stuff enough times on photo and art sites... always loved to play...
...I just wonder how long i have to click PASS until i see our pics come up.
No offense to the originator of this thread or any others suffering from the 31 dilemma. I'm just curious about this whole internal-link-anchor-text-paranoia-thingy.
Let's say you have a site with 500 pages of content and let's say the site is categorically about red cars. Let's further say that there's a page on the site that is about "cleaning tips for red cars". Lastly, let's say that every page on the site has a link to this one page and that the link says "cleaning tips for red cars".
Are there webmasters out there so paranoid to believe that there's a possible penalty for this (that was rhetorical as I know there are)?
Let's apply logic here. In this example, the site is about red cars. If you have a page on such a site that is about...the cleaning of red cars, why WOULDN'T you place a link TO this one page on every page of your site? If it's a well written page that offers lots of tips on red car cleaning and you think your visitors would like to read the page, what would be the prob?
And, in this example, why SHOULDN'T you use the anchor text "cleaning tips for red cars" in every link if, in fact, the page is ACTUALLY ABOUT cleaning tips for red cars?
Should a person worry about offending googlebot because it may get bored reading the same anchor text over and over and, thus, lash out at the site with a boredom-inspired penalty?
Perhaps one should appease the bot by "varying" the anchor text in the internal links to this page? I mean, it won't change what the page is about, but perhaps googlebot would like to see "red car cleaning tips" or "tips for cleaning red cars" or "cleaning cars the red tips way" just to break up the monotony of endless crawls.
Here's what I say: if I have a page about cleaning tips for red cars and I think the page offers good content to my site's visitors, I'm going to include a link to it from every page of my red car website
and I'm going to do it so they can EASILY FIND the page.
And I'm going to do more than that. I'm also going to make every link say "cleaning tips for red cars" because...that's what the page is about.
And lastly, I'm not going to vary my stinking anchor text because...why should I? Who am I trying to fake out?
And I'm going to do more than that. I'm also going to make every link say "cleaning tips for red cars" because...that's what the page is about.
Makes perfect sense. If I have a major section of my site devoted to widgets, I'll have a link with the anchor text "widgets." What the heck else am I supposed to call that section?
Links from external sources are a different matter. If I talk 100 Webmasters into linking to my Acme-widgets.com site with the word "widgets," that's going to look unnatural. To a human being, at least, it would make more sense for most of those links to read "Acme Widgets" or "Acme-widgets.com."
It's also worth remembering that factors such as anchor text aren't likely to be considered in isolation. (I think someone else may have pointed this out earlier in the thread.) Profiling doesn't say "fat guy is likely to be a drunk." It says "fat guy in a Bud Lite t-shirt with an open beer bottle in his hand, a Jimmy Buffet CD in his car stereo, and alcohol on his breath is likely to be a drunk."
With two sites hit, and having corresponded with others in a similar position, I am convinced it is a thin affiliate issue. As people have suggested, Google cannot hand index every site, but they could feasibly start with adsense publishers.......
My response is to steadily add more content and reduce the number of Adsense/affiliate links. It would be great if GoogleGuy could let people know the re-evaluation process when sites are ready.
However, I would not hold my breath waiting for Google to leak more data about AdSense - Google search correlation. Scare enough webmasters to remove AdSense code from their sites and see your revenue drop? I don't think so.
While I am still at #12 for my primary keyword on google....for the first time ever, I jumped 109 spots on MSN to #5 for a very comeptitive keyword (38,000,000 results in the travel sector) and I jumped 15 spots on Yahoo to #40. Both are the highest spots in 18 months.
Google takes a lot longer to react.
we shall see, good things so far!
Logic says, if Google says only go for about 100 links per page, and you place 500 links on one page with one word common to all, its going to raise a flag.
If your site is about red cars, you would not put “red cars” in every anchor for site visitors. That much repetition, can only possibly be for search engines, your visitors know your site is about red cars, no need to tell them in evey anchor.
Yes, I believe Google is sophisticated to see that’s what its for, and its for search engines, not people.
This, is not an idle opinion. I’ve now done an analysis of over 30 separate, unrelated search terms, checked out the top 50 sites, and am seeing minuscule numbers of sites with large anchor repetition in the top10, and the ones that are, are big authority sites.
I am not bothering to question overly the logic of Google, I am testing whether there is realistic cause and effect, and am seeing it.
If some would like to sticky me a top 10 site for a competitive phrase that has 50+ repetitions of the main keyword in anchors on the home page I’d be grateful, and a little surprised.
1st two sites of mien affected did not have adsense, or affiliates.
It took exactly 6 months for the penalty to be removed.
During that time an exact search for my unique company name listed my site way down (may have been 50+, I didn't count). Most-all sites that referred to my company appeared before mine.
Besides fixing the dup content issue i also greatly reduced my internal linking. Since then i've only brought the internal linking back by 50% of what it used to be... search results have never been better.
I would put "red cars" in every single link that led to a page titled "cleaning tips for red cars" because that's what the page is about. And if I wanted every visitor to my site to have an optimal chance of finding the vaunted "cleaning tips for red cars" page (because it's just so doggone good), I would place the link on every page of my site and I wouldn't worry about it. That would be building the site for visitors and not for search engines.
"uncontested phrase in brackets" - Result No 1/2 - page 1
without brackets - page 3
"heavily competed phrase in brackets" - page 3/4 [ approx 30 below ]
without brackets - page 3/4 same as above! [ that's strange ]
I'm not sure if we're classified as an old [ 3.5 yrs ] site being reintroduced to the SERP's after a few fixes, going through some filter checks
If we're continuing to trigger a filter that puts us into the -30 bracket for only some common phrases.
a combination of both
Somehow, i think this is a penalty by degree system which ascribes a score to a range of filters for promotion or demotion. What each filter check is worth, [ other than the obvious one's ] only Google will ever know, but i think we're getting close to a stronger gut feel as to what is the most severe and the least severe ones [ accumulated could be a problem ]. So some degree is able to be estimated in conceptual terms.
It would be good to have a dot point list of things that might trigger this, and a dot point list of *factual* things that oppose it [ ie Trust - IBL's in ; age ; brand ; fixing the problems! ; etc ].
Then from this, it may be possible to arrange those things into *severity* levels.
Then from this we might be able to see sites that are existing, recovering and new.
This is a sort of parallel thread with, Filters exist - the Sandbox doesn't. How to build Trust. [webmasterworld.com], where we've cobbled together some thoughts.
I think this would get us closer to understanding things.
However, I do appeal to our friends Matt, Adam , Vanessa and GoogleGuy to provide some light on this. Clearly, webmasters need to understand in broad terms what these triggers are to better quality control their sites, and they have *TRUST* at least from those who subscribe and identify themselves at WMC/Sitemaps.
We really need some systematic guidance beyond the existing guideline, on the application of filters, broken down and reportable, either here or eventually and ideally through Webmastercentral.
In the interim something spoken or written would help. We don't need secret soup.
[edited by: Whitey at 2:21 am (utc) on Oct. 19, 2006]