| This 73 message thread spans 3 pages: < < 73 ( 1  3 ) > > || |
|Suddenly 950'd... what to do?|
Hi guys - I need your trusted, professional advice, fast!
After 13 years of sailing along fine, and after what I'd now consider only a few minor bumps along the way (in the last two years) - my site looks to have finally run aground, got struck by lighting and is now on fire! (yes, I know many of you are cheering burn baby burn)
I still rank for a handful of key phases, (and I MEAN a handful) but most of my good pages are now sittings at exactly page 9 or 10. I would think a total 950 would remove my entire site, but it seems only select pages got clobbered.
It's as if it's a selective 950. I'm hoping it's not the death penalty I now fear.
On suspect is the EMD, but come on, I registered this domain in 2000. Funny thing is, one of the only remaining terms is the EMD match term.
I've been combing the site for obvious web spam, but can find nothing.
I submitted a re inclusion last week and they already cam back saying "NO MANUAL PENALTY"
The only issue I DID find was this:
The images use the ALT and TITLE tag and I had those both set to the SAME thing for each image. Might that be considered SPAM or duplication perhaps? It was certainly not intended to be so.
I have since cleaned that up, shortened the alt to a concise description of the image and pretty much left the title to the longer useful description shown on mouse over.
The other issue may be length of the slide shows.
The main page slider is 20 images, but one on a sub page is 60 images. Would this be considered spam?
I'm tempted to take the 60 image slider, import in to Camtasia, make it a movie and feed it to Youtube then embed it on my page. Playing Googles game as it were.
My other (highly paranoid) thought is that some Googler's have tracked my often candid & crazy comments on here and busted me for personal reasons...but if that were the case, 90% of use would be 950'd and that would just be downright evil ;^D
Your thoughts and advice are greatly appreciated.
We've got 950 pages and I don't think it's manual, I think it's part of an algo that says "We aren't going to deindex this page but it won't see the light of day either.
The EMD problems you might be stuck with to some degree, one of ours was pretty level for a while and I'd thought somehow the domain was finally understood as a brand name but that's not the case. Matt Cutts released a video the other day about sitewide anchor text with KWs [youtube.com...] believe what you will; up till now what he says seems to be have been true for us but it might also be the source of the "breathing" we see from keywords fading in and out every 10 days or so.
petehall has identified a serious problem with my site, but one I was always well aware of and have fixed before. The old / vs /index.html canonical issue. It seem when I switched my expression web 4 over to a new faster workstation, as I imported the site files in, EW4 took the liberty of adding /index.html to my code...393 occurrences in all! Now G thinks I'm posting dupe content. The timeline matches the decline so I'm scrambling to fix the issue on my end.
My next question is, do I need to notify Google or will they pick up in the change themselves?
BTW - thanks again pete!
Turns out, it may not have been Expression Web 4 's fault at all, I was cleaning up my old dirty code using a website code cleaner, dirty markup and it was most likely the culprit. Beware!
No problem... more than happy to help. Very easy to overlook if it's something you've fixed before in the past! Happy coding... :)
I have always found they pick up the 301's purely by setting them up in htaccess or programatically. If anyone does know of a faster way I'd like to know myself...
Whew! Looks like changes are already happening...4 pages that were stuck on page 9-10 have now instantly moved to page 3 and 4, one to page 2. Also starting to see pages that were AWOL in GART now appearing with visitors. It's not a total recovery, but it's only been an hour...now I just have to wait and hope for the best.
You're the best pete! Thanks to everyone who has jumped in with suggestions, they've all been good and helpful! You guys are GREAT!
I am now finding pages that were gone, now appearing for page specific terms. For this term, my page is #1. For a very similar term that is also used on that page, it is at page 9. Goofy! But again, it needs to digest. All these changes have occurred within an hour of removing the /index.html links.
I'm beginning to believe the smoke & mirror mystery penalty theories less and less as this looks to be a simple problem that manifested with my web design software move and easily overlooked.
|removing the /index.html links |
Can you expand on that, I'm wondering if I might be having the same issue. Did you have links on your sites pointing to /index.html files? Off page links? Did the index.html files not exist? (PM if you want)
Glad to see you didn't give up.
Glad Petehall was able to diagnose, and the fix seems to be working!
Does anybody know of a reliable service that can find things like this for you? You'd think there'd be some kind of scan you could do that would quickly detect stuff like this, or even monitor your site and send you warnings.
There are some tools that are available; I can't really post them here because of the forum rules, but if you look at the SEO Tools post, and check out some of the ones I say I use, you might find something useful. In particular my third post in that item:
Netmeg, thanks! That's a super valuable thread. I'll definitely download that one and try it out.
@backdraft7 - Can you explain the problem and solution a bit better you found? (Congrats btw)
From my understanding you had pages that could be accessed by either
and you had canonical links on this page and versions of this page that pointed to either link but not just one?
*Blushing* Pleased there's a bit of light at the end of the tunnel. :)
Thanks again for all the help folks...
Unfortunately I spent the afternoon reading up on the END update and I'm not hopeful. I feel like someone kicked me in the teeth. I have a testimonial page, it is text and photos from actual customers through the years who have sent positive feed back to me. And guess what? they use my domain name and the two words that make it up in their positive comments. Over the years I've probably had enough complaints to count on one hand, yet in Google's blind eyes, I am a criminal of no worth. If anyone thinks those testimonials are phoney, keyword loaded spun content, go to my facebook page, which is linked to that page and read the long list of comments with photos of each customer result. That can't be faked. So, when I search for my testimonial page (and the rest of my site) and it's now at -1000, I'm somewhat incensed.
Granted, on my testimonial page (which rather sprawls the length of probably 4 normal page lengths) there are 56 occurrences of word 1 and 33 occurrences of word 2 of my EMD. Am I supposed to change those words to "widgets"? The EMD is tantamount to censorship and tears against the fabric of what makes us free.
Yes, spammy sites with long keyword loaded domains that offer no value, dead end quote or signup forms must go, but to pin everyone with a penchant for the verbose as a criminal or to silence the legitimate positive feedback of customers is wrong.
I have an old home town friend who is now very well known legal reporter for Fox News (GVS). She went to the same High School as myself and graduated in the same HS class as my older sister. I think it might be time to run a story like this past her to see what she thinks. Maybe with her firepower we could get a manual review or interview with the Great & Powerful Oz. Of course we'll have to wait for a slow news day...
Bottom line - IMHO, the EMD is blind censorship and so far, appears to be the most deadly of Google's "quality" updates to date.
As a joke, I replaced my key words with the names of colors. Didn't work so well, The page went from page 13 to 28 in just a few hours.
backdraft7, remember that patent for jerking your pages up and down the SERPs so you can't figure out what's working and what's not working. You may be making too many changes too quickly... or you may just need to let it settle.
In my experience, something I do to a page, be it on or off has RARELY if ever had an immediate (read hours) impact on serps. It just doesn't work that way. I think you're reading false cause/effect into googles recent schizophrenia between DC's.
What kind of a joke is it where you shoot yourself in the foot on purpose?
You didn't learn something from your action and the S/E reaction?
After making those changes I would leave it alone and sit back... don't keep changing things. I'd normally have to wait a couple of months to see an improvement, although I have no idea how fast Panda is working these days.
@blunt - it wasn't so much of a joke as an experiment. that page was clearly dragging the whole site down with the high EMD keyword density, so it has been demoted from my front page to a link on the product detail pages...I figure it's best to put any testimonials on the product detail page level than on the absolute top of the navigational hierarchy. Believe me, I've learned...the hard way. If I can't poke a little fun at myself, then I'll truly go nuts. Well on my way though!
@pete, agreed, no more changes. I have plenty of offline work to keep me busy for a few weeks.
In my experience (five or so trashed EMDs), it is the EMD that is the prime problem.
There must be other factors however because I still have one EMD which was not affected.
The EMD penalty by G was about as crude and blunt as can be imagined.
I gave up bashing my brain in over the reason, scrapped the sites and moved the content, pictures and videos to another site with good results.
This is slightly off topic, but a vital part of my "recovery" effort.
As I'm working on a new site, I'm wondering: What's the penalty for using the domain forwarding function to point one domain to another?
Why? because we spent a week searching for a brandable name that nobody else was using AND had a matching available domain. The one I found has a consonant in the middle that works either as D (the brand word) or DD which is what most humans would try (two separate words) so there is a possible confusion. I registered both. I plan on installing the site on the brand URL and forwarding the DD variant to the brand site using the domain forwarding function at the registrar level.
So, my question is: Will Google's AI bridge this gap? or I jump out of the gate with a penalty for domain forwarding?
I found topic: [webmasterworld.com...]
I do not care nor do I want Google to index the human variant, only the brand domain.
My suspicion is that if I just use the brand domain but use the two words on my site, Google will bridge the gap eliminating the need for the human variant, however, for word of mouth purposes, a certain number of people will surely enter the two word (the way it sounds) into search, and this is where I am hoping G can figure out the relationship and correct the search, pointing the user to the correct domain.
I'm not sure I'm answering your exact question, but I've registered and forwarded a number of domains that I thought people might "mis-hear" or misspell as my "real" domain. I've done this for years, and Google just doesn't index them, or maybe they index one page that you can only find it by putting in the full URL. I don't think it's a problem, if I understand what you're talking about.
There must be other factors however because I still have one EMD which was not affected.
The EMD penalty by G was about as crude and blunt as can be imagined.
I'm wondering if Google figures "theredwidget.com" should rarely have the words "red" or "widget" appearing in the site text, links and navigation because they're implied by the domain name, and re-using them a lot in the body of the site = keyword stuffing.
Spent the day combing my site for grammatical errors, keyword overuse and missing alt tags...my GOD! You know, over the years, my intention was to post what I thought was useful to my readers, trying to manage server issues, create detailed paid content and take care of my customers all at the same time, but after reading it again, and knowing that G now looks so closely at keywords in content, it does sound a tad keyword heavy.
Funny thing is, I was just writing in my style for my readers, but after long days of trying to manage site singled handedly and somewhat bleary eyed, I became blind to the possible ill effects of my own verbosity. (I grew up with those early, long winded full page Sharper Image ads by Richard Thalheimer and always loved the colorful catalog critiques of Seinfeld's J. Peterman). Now I'm being forced to remove my style and correcting to make the search engine happy...and that's wrong on both counts, now isn't it?
The bottom line is that it was never my intent to keyword load the pages, but it just happened over a dozen years of quick revisions and no thoughts of bad intent. With their divine powers of prognostication, it's too bad that Google couldn't detect it wasn't bad intentions, it was just that I was inept and a bit over creative, having no idea that the black box was already formulating my destruction.
A useful test can be to read it out loud. If it doesn't sound natural out loud, it probably won't when being read.
It's really surprising that Google does issue some sort of warning when on-site duplicate content issues arise. Google issues notices in WMT for duplicate meta tags and the such, so how hard would it be to warn webmasters of canonicalization problems? One could argue that duplicate meta tags have a limited impact on ranks while canonical results in serious penalties. And although Google does not classify it as a penalty, losing 900 positions is not a minor shift that will slightly impact traffic.
I've had the same problems with canonicalization. The funny thing is Google would warn about duplicate meta tags in WMT, but not those pages that were canonical related problems. I wonder why Google fails to report this in WMT when the meta tags are duplicates? Weird.
|Now I'm being forced to remove my style and correcting to make the search engine happy...and that's wrong on both counts, now isn't it? |
It is, but there may be a bright side.
I got the verbosity trained out of me by a wonderful high school English teacher. I thought she was so very wrong at first, but the more I read my essays they way she had "forced" me to write them, the better I realized it sounded. So you might find it actually sounds better in the new style - I hope it does, and I hope it converts into sales.
|The funny thing is Google would warn about duplicate meta tags in WMT, but not those pages that were canonical related problems. |
Actually, in reading through what I'm getting is:
1.) It wasn't a duplicate problem it was the links all being changed to /dir/index.html rather than /dir/ like they should have been.
2.) It doesn't seem like it actually solved the issue after reading the longer post a few hours later.
|So, when I search for my testimonial page (and the rest of my site) and it's now at -1000, I'm somewhat incensed. |
The results are all over the map now, unfortunately Google has no idea what they are serving. When I'm looking for my product (and there's only two other sites on the web that I know of that deliver similar) they keep feeding me thin overviews or synopsis rather than meat and potatoes. I used to cringe to see my competitors ranking near me, now I miss them dearly, cause they are gone too!
I suspect they hope visitors will have to trudge through the ad ridden pages and maybe click on my ad to find me. In my niche, the results are the worst I've ever seen, and not just because I'm not there.
Data sets must be cycling fast because traffic can be up one minute, then nil the next. I will say this, it was the best Sunday out of the past four, but that's not sayin' much. Time will tell.
BTW the way, I saw the Bing commercial today bashing G's privacy (or lack thereof) saying "don't get Scroogled", funny stuff! I laughed and then I cheered! Go Bing! I also see FB is kicking in with ads for FB Home. Good to see these guys stepping it up against Goliath.
My logs show Bing coming up on G fast! But with G dumping me, that's easy to explain.
I feel your pain, I really do.
Back when -950 was introduced, I was one of the first to be penalised with it. A great site, male health, at number 1 for years, very useful to so many people (millions of visits), gone overnight.
Never recovered, no matter what I tried.
I believe that was due to every page having an internal menu with same anchor text.
Since then I have been plagued by -950, having pages penalised on many sites for no obvious reason. (Except that maybe I didn't learn a particular lesson....)
It (the penaly, google, me?) reached its height of crazinesss recently with a new page on another highly ranking site being instantly (within days) whopped to the end of results - despite only having one link (anchored as click here) from the home page.
So after I realized I was playing to some degree into the definition of madness - doing the same thing over and over and expecting a different result - I sat back and thought about this.
A page on relationships on a site on relationships, well constructed and full of great content (at least in my eyes), with the page not featuring on the internal menu, or with any external links, instantly penalised for the keywords? What could that tell me?
At a keyword density of 0.8%, too.
I conclude the deadly trio are a meta title too similar to the H1 heading, and a ludicrously low level of acceptability for the keyword phrase. (0.8% ? Come on....) The file name or domain name may play a part too.
But....but....but. I see so many pages whopped to the end of results that have no apparent links, over optimisation, or keyword stuffing. There are some corkers there, too: pages on completely different subjects to the one I am looking at, but which happen to have the various keywords randomly scattered on the page.
Perhaps this is madness: an unpredictable algo with some combination of criteria which act together to produce what looks like a random effect.
Of all the things Google has done the EMD penalty and the End of Results (EOR) penalty are too of the most difficult to understand and get out of, in my judgement. The original discussion of the EOR penalty demonstrated this - in the random nature (which of course is not random at all, it's highly algorithmic) of the things people did which got them out of it - changing internal navigation, adding one link form an authority site, and so on.
In my experience, it takes 6 months to get out of the EOR penalty. I had a few EMD blogs, which were all sitting having a party at the end of results. As an experiment,I changed the anchor text on the inlinks (it was pretty well over-optimised) and left them alone. Several months later - there they were - at the top of the results.
So: what does it all mean? What can one do? I'd suggest reading the original thread on the -950 penalty. I think the criteria have been turned up or added too since then, but some of it is still true (eg the need to change incoming anchore text and internal menus).
I believe the EMD penalty is in fact the End of results penalty. I think the domain name was added to the criteria for this penalty.
Two final points: I am having some success with 301s to new domains. It saves a lot of time (I don't have to wait for 6 months!).
I also echo the poster who observed that natural reading aloud is helpful. I did this on som eof my pages that had been whopped, and I was struck by something - that the keywords didn't sound natural. (Not that there were many of them.)
That makes me wonder if someone somewhere with an astonishingly high command of English reviewed them manually and decided their intention was "impure". Or is it possible that the -950 penalty or End of Results, call it what you will, is actually the product of Google's efforts to enagage in semantic analysis, machine learning, or computer driven analysis of the language.
Who knows. But the results of this penalty are certainly thought provoking. And distressing.
Thanks for that Fred, it certainly is thought provoking.
Just wanted to report and verify that yesterday WAS indeed an improvement over past weeks decline and edged out the past Sundays back to March 24. It only beat out last Sunday by about 10 visitors, but it was an improvement, especially when you consider my blog traffic (or boat anchor) is now gone. It's not a victory by any means, but perhaps the bleeding is under control...for now.
That said, it looks like traffic today is lower than dismal. 1 hit every 5 or 10 minutes, (which doesn't even pay the light bill) whereas yesterday was seeing relatively decent volume (considering the circumstances) of 10 to 15 at a time...in spurts. I suspect many of those are my returning customers.
On Google, the site is down to page 1 results on about 6 keyphrases...from a max of 19,000 last year.
| This 73 message thread spans 3 pages: < < 73 ( 1  3 ) > > |