| This 242 message thread spans 9 pages: < < 242 ( 1 2 3  5 6 7 8 9 ) > > || |
|Major Ranking Losses on 2008 June 4 - 5|
I can now join the catastrophic loss club! UK site, PR6, rock solid rankings, 400k pages indexed, fast loading - 90% loss of traffic since 2 days ago. Last time the site went AWOL was early February for a couple of days.
Looking up what I used to rank for approx 20,00 medium to long phrases then all top listing just disappeared BUT results can now only be described as awful - in fact on one query results 11-20 contained sites from Poland, Germany, Czech, .info and not a single .com or .co.uk - when will Google admit that they really screwed up big time this time?
As my site is a .com hosted on uk servers then it seems that I have now moved to the index for Mars. Time to move to the dark side methinks rather than stay in the Google lottery.
[edited by: tedster at 7:08 pm (utc) on June 6, 2008]
[edit reason] moved from [webmasterworld.com...] [/edit]
My search traffic doubled on June 4! Recovery from a -950 penalty. See the 950 thread for more info. One webmaster there posted being hit with the 950 at the same time I recovered (after being in 950 hell for over six months). That also makes me wonder if it was a 950 Penalty Algo update. It's also possible the main SERP algo was updated and new weighting in one area or various areas OVERRIDE the 950 penalty.
Time on site or bounce rate, IMO, has nothing to do with it. (Mine didn't change.)
But here's a question. If Google were to consider bounce rate, how does it create an equation to factor in advertising... which affects bounce rate? If you put ads in the heat map (top of page), you're going to have a higher bounce rate than those who don't.
Unless and until Google can access and use the data of who clicks on ads, and who doesn't, using bounce rate as a significant factor for SERPs is ridiculous. If Google itself says an update had to do with bounce rates, I'll believe it. But not until then. Right now it's just too noisy a signal and much too unreliable to be taken seriously.
|That also makes me wonder if it was a 950 Penalty Algo update. |
The thought has crossed my mind, too. The negative relevance multplier always seemed too heavy handed to me, and if a graduated multiplier is more of the way it's working now, then you may have a good idea there.
Another possibility is that the various thresholds for tripping the penalty may have been recalculated. That's just speculation, by the way - I don't have nearly enough data to be definitive on the topic.
But all the reports here to do not include going to the end of results - and several sites report that their rankings are now frequently bouncing around. The rankiing tilt-a-whirl was not a part of the -950 in the past.
I do not believe that update 'Hecate' (well that's what I am calling it!) has anything to do with a -950 penalty but I do have a theory that could explain the catastrophic changes on the long tail and it goes like this!
We all know that Google will spell correct your search even when there is no spelling mistake for a word that does not happen to be in their spell correction dictionary so the basis of my theory is that original searches are being amended in some unknown way. From looking at the results appearing where I used to appear i.e 3 or 4 word searches then it looks like if I submit a threee word search then instead of running a three word search where the proximity of the words on the page has some importance then it looks like a two word search is run and the proximity of the third word has a lower importance. The loss of the importance of on page proximity then leads to less relevant results being returned.
Clearly, if you had something like Brand, Product Type, Model Number as the search but the results became more focused on Brand and Product Type but not Model Number then my ability to compete with say the monstrous comparison sites would be diminished as these are well known for stuffing Google with as many ways as possible to get those combinations on different pages as opposed to my approach of being very specific so in effect the only way that I could compete is to go from a compact targeted approach to filling Google with as many possible spammy on site mash ups of the same data as possible.
The other evidence I have is that when my pages are found then the depths to which people have gone (say, results 40-200) is quite amazing but when I replicate their search by clicking on my latest visitor stats then I can see why they chose my page as the results above are not specific to their search but are dominated by Brand and Product Type.
Just a theory based on observation.
I have now had the chance to go over my logs for the last few days and it is worse than I thought. Of my current residual traffic then about 80% is purely image search and it seems that the quantitative value of image search visitors is the same as it was before 4 June.
Of the remaining 20% then approximate 80% of this is NOT from Google uk or com but from countries wide and far, and once again that seems to be in line with levels before 4 June. So that leaves about 4% of my CURRENT traffic coming from where it used to or put another way I have lost 99.6% of my traffic! Wipeout or what?
Funny thing is Google has spidered approximately 300,000 pages in the first 11 days of June and this spidering activity is pretty constant day by day. So, yet again, everything looks completely normal activity for my site.
My other smaller sites have seen an increase in traffic and are achieving some sales but to be brutally frank when I track back to what the Google visitor was seeing then the results sets are pretty abysmal and my sites do certainly not deserve to rank for the terms that I am being found for. What a mess. I just hope that this effect will cascade to seriously affect some of the monster size sites and then I think that we may see some reaction from Google.
I wonder if age protection that some old sites had against 950 hell has been lifted. I say that because some of the sites hit sound like they are older. Also, because many sound like they are very large sites (which often are fairly thin on content or are borderline dupe content). Google may have turned the dial on the value of old sites wrt their age so they are no longer immune from the 950 penalty, or turned the dial against thin sites. My site had thin pages but the 50% traffic boost came after they were developed with more on-page text.
You may be onto something. Many of my long-tail searches came back on June 4-5. But the competitive two-word phrases didn't (mostly). They moved a bit, generally, but not much.
Anyone else notice anything interesting about long-tail SERPS after June 4-5.
I am sticking to my on page long tail proximity argument! The changes you are seeing, in my view, is not a release from a -950 penalty but a traffic increase because of an algo change that affects the long tail. You have gone up and I have gone down!
A few of the observations seems correct. My site is hit and it is:
1. Mainly a forum with content relevant 301ed urls.
2. 99.9% of traffic is long tail ( 3 - 5 word phrases). Not a single targeted competetive keyword.
3. Over 200k pages indexed ( ie. it's a large site) and googlebot traffic is constant. No change in PR.
4. Very few or no outbound links in pages. Even if one there it's 'no follow'ed.
hey Everyone, please, be honest.
Could the pages from the huge sites look like templates generated for internal linking to google?
I just checked our traffic and they seem to be climbing up, not quite up to earlier levels but its a god sign!
Also our percentage google traffic was down around 50% during this dark time but now its 84%.
I'm not counting my chickens yet but somethings changed-hope it won't reverse tommorrow.
Anyone else seeing anything?
6000 unique visits a day has dropped to 3,000 since June 4/5.
The loss is entirely from google natural search.
Page rank still 4 for homepage.
Still indexing for many major phrases in pages 1 (including the top) 2 and 3.
We are still getting traffic from long tail phrases of 4+ words, but a lot less than we were
Can't see any link to bounce rate either
It is specifically one section of the site that has lost the traffic.
That content has more dynamically generated elements that the parts of the site still getting traffic.
Does that inspire any ideas?
I have seen a slight improvement in the results in my niche but nothing to shout home about.
MattT - welcome to the discussion and WebmasterWorld! I could have written what you have said, albeit, I would have to change the PR, increase the uniques and the drop and say some major phrases but basically your site footprint is similar to mine. The key point being the loss of traffic being directly attributable to a loss of the long tail searches.
Where is your site located? UK? US?
Thank you for the kind welcome
The site is UK based
I am still getting lots of long tail traffic to parts of the site
But not to specific types of page that used to do well.
It could well be that those dynamically created pages do not have good word proximity.
or it could be something else about those pages
Our site's traffic from google has not changed. I've been trying to nail down some more specifics.
[edited by: tedster at 6:59 pm (utc) on June 12, 2008]
|I do not believe that update 'Hecate' (well that's what I am calling it!) has anything to do with a -950 penalty but I do have a theory that could explain the catastrophic changes on the long tail and it goes like this! |
The idea that potentialgeek and tedster are bouncing around is very valid when you think of the infrastructure (and it's application of sorting, filtering and displaying search results) as an evolutionary entity.
As elements evolve, they tend to have periods of stasis, followed by periods of "quantum leaps", in other words periods of the greatest flux when various calculations that were previously calculated on a larger, macro scale (such as the 950) get graduated into smaller and smaller segments.
It could very well be that this is a layering concept of Trust applied not only as 950, and what we have seen as graduated amounts (-30,-60. X), but now has evolved to be able to be even more graduated and specific. We started with large scale drops, more defined B+W (website lost too much trust, -950, lost some trust, possibly no effect or small effects)
At the same time, if the entire process was updated, this would account for websites leaving 950, and would also account for some websites only dropping 70 positions (as opposed to dropping 950, or 400.)
It would also account for losses of major keyword rankings and associated traffic along with associated long tails. And the pillar support those pages provide in the website both from an 'interlinking' and 'semantic theming' standpoint could cause a domino effect across the domain, for some websites.
I understand fully your explanation BUT my observation is that the resulting impact is more chaotic than structured. So, I can accept that the effect is graduated but I still believe that this is not an empirical drop. By their nature the -30, -60 and -950 drops were largely quantified as empirical drops which was my main reason for concluding that this is not a -950 drop, if that makes sense. This appears to be a new phenomenon that as yet I do not think anyone can relate a specific cause to a specific effect in a specific geo location. An interesting issue so far and we have not yet run what I consider to be the Google 16 day re-spider cycle that I have quantified for my particular site.
Yes, my pages could be confused by google as being created with a template for internal linking. They link internally and most are similar. They're all hand-made, with different content and photos, and similar links. I don't redesign a web page each time I create one. I use my own template page. Very straight forward html.
The craziness hit here. Ours are dot coms, hosting in the U.S. on the web since 2002, targeting mostly U.S. visitors, but some sites get alot of visitors from around the world. Nearly all sites with Adsense now. The rankings are wild, but I am seeing your pain.
One keyword we ranked #1 for on one site, was out of about 18 million pages. Today it is ranking #3 out of 478,000 pages, which I guess accounts for all you missing in action. But, it is ranking for the hyphenated domain we own, not our main domain AND it is not for oursite.com it is for our-site.com/index.html. Traffic to this site is off 66% from Google.
Rankings for relevant keywords that normally came into our oursite.com are all gone for one website, but not others.
For another site, I used to be listed number #1 with the six most popular pages on the site listed below for our #2 keyword, now I am just #1, with a basic listing.
Adsense revenue has dropped 20% to our network as whole since June 4. With the dog days of summer right around the corner, it couldn't happen at a worst time.
Late in May, one site, our main earner, #1 keyword went from a 950 penalty to ranking #18 (after a recent revision of our main page), bounced around for a week between #5 and #18, now has landed firmly on page 4 of search results, along with a good number of my competitors who were on page 1 or two before our cages began to rattle. On page 1 and 2 search results are numerous sites that were nowhere to be seen a week ago.
It was crazy. On a major product keyword, a CNN news story was ranking #2 out of 219,000,000 ... for a PRODUCT... above all the companies who have been selling the product for many years. That is just insane. I have never seen this before.
Somebody please tell me this is not long term, is not going to kill our summer and ruin our peak season, fall/winter. I am in the middle of my summer renovations to our major site, to be uploaded next week, which is always a nervous time, wondering if we goofed when we did the renovations, now I am scared to death to upload the changes.
We had just recovered in recent months from an almost year year long beating by Google and were back to our highest levels, 2006.
I had just uploaded within the past two weeks site maps for all our major sites to google webmaster tools so that all our pages could get indexed. It should have made things better, not worst?
Ok. I am up to speed now. Had to read through this all.
Yes, ONE of our sites, the majority of the traffic is suddenly coming in on our background image from Google image searches for the #1 keyword for that site. It is a sudden change we never saw before. I was just looking at that yesterday and it jumped out at me that it was odd.
Well, just when I thought it couldn't get any worst, I went back in and looked at the rankings for another clue, the keyword that had the 950 penalty, came back and floated around from #5 to #18 rank, then settled on page four is gone. Looks like it is back to a 950.
I also noticed that none of my sites have the OTHER PAGES FROM THIS SITE links below the listings.
A SCRAPER SITE that had been competing with me, gone for the past two years in search results, forced to advertise on Adwords to get any traffic Google traffic when Google cleaned most of the scrapers out, is back ranking, on page 4.
If it is an update, it a step back in time for the worst when scraper sites are back ranking.
I have a theory and it's a wild one. Maybe Google didn't change their algorithm at all. Maybe this is all related to the wordpress hack (http://www.techcrunch.com/2008/06/11/my-blog-was-hacked-is-yours-next-huge-wordpress-security-issues/ [techcrunch.com])that is spreading like wildfire. Spammers are taking over wordpress blogs all across the Web (mine included), placing links to their sites, and displacing yours in Google Results.
It would be interesting to see which sites replaced yours in the results. A quick check of their back links could confirm if they are getting link love from the hacked blogs.
If I'm right, this would be largest example of Google Bombing ever, resulting in the shift of millions of dollars from thousands of "small" site owners into the hands of spammers.
[edited by: tedster at 12:41 am (utc) on June 13, 2008]
[edit reason] to make the link clickable [/edit]
*** uploaded within the past two weeks site maps for all our major sites to google webmaster tools ***
Not the first time I have heard of someone uploading a sitemap and seeing the site tank a few weeks later. There was a lot of that a year or two back. No idea if it really is related or not.
*** But, it is ranking for the hyphenated domain we own, not our main domain AND it is not for oursite.com it is for our-site.com/index.html ***
Note that having both our-site.com/index.html and our-site.com/ is a Duplicate Content issue that you might want to fix. Internally link to the shorter URL, and make sure that requests for the longer URL redirect to the shorter URL too.
Have you got any www vs. non-www issues too?
Actually, I am concerned that you have two domains. Is the content on both related, or similar, or identical?
see my previous post from page one of the discussion.
Almost all of our G organic traffic came back on june 9 and then disappeared again on june 11. This is getting wild.
I'm hoping more people can point to the type of rankings they lost rather than just traffic drops. Traffic reports alone don't help us understand what shifted, and I'm certain something rather significant shifted with Google last week. I just can't see what it was, at least not yet. From the reports here, it was a pretty hearty pop against most of a domain's rankings, but it was not a total ban.
Up to now, sites that were hosting parasite links got totally dropped - and often a message in Webmaster Tools as well. I'd like to rule out that idea - because I think it's more of a major change to the way rankings are calculated.
Can I assume most people in the thread have run link checking software on their domain to spot any outbound links that don't belong?
|his appears to be a new phenomenon that as yet I do not think anyone can relate a specific cause to a specific effect in a specific geo location. |
My explanation for the reason why it is happening now is that a new variation of this particular filtering was applied just now.
|I'm hoping more people can point to the type of rankings they lost rather than just traffic drops. |
It is definitely long tail traffic that I have lost. I had one four-word phrase that was a big performer, and it has moved from #1 to #11, bringing in virtually no traffic now. I was at #1 for the last couple of years.
Over the past year my main (one-word) search term is really the only one that was bouncing around in the top 5. It has been in the top 5, usually #1, for more than 10 years. It is now #15. All of my long tail traffic was very stable until this update. I don't keep track of all the different phrases, but here is an example of my new rankings for the ones I could think of off the top of my head. My previous ranking is listed first:
Thank you, sahm - that helps refine your site's profile. You also mentioned these factors earlier
1. you lost over half your traffic
2. sites now above you are younger and smaller - many link to you
3. hosted in the U.S.
4. many incoming and outgoing links
5. many on-topic, one way backlinks
6. publish original and reprinted articles - no products
7. is updated several times a week with new content
8. extensive use of article distribution
9. your two smaller sites are not affected - with similar backlinks
10. most competitors with similar backlinks are not affected, but some are -950
Have I got the entire picture correct? I'm now wondering if other affected sites:
1. Ecommerce or informational
2. Monetize through sales or advertising
3. Distribute original articles
4. Reprint other articles
5. Buy links
6, Sell links
7. Target country-specific search or international
Well, we lost about 10-20% of our google traffic, now as of this morning june 13th we are losing 80% of our google traffic so things have got much worse.
We still have sitelinks for a couple of generic keyword terms so im guessing we have not been penalized and its just a massive algo change.
- 2.5 years old full page advertising site contains affiliate links on most pages
- uk site hosted in the uk
- massive amount of one way backlinks to main page (non paid for)
- lots of one way on topic backlinks to inner pages (non paid for)
- 3000+ pages of unique content on many different topics
- Was getting 10,000+ unique visitors a day
- New content / pages added daily
- Target country set to uk
Forgot to mention we have a much smaller site which is completely uneffected, infact both sites shared the top positions for high traffic advertising keywords now only the small site is still there and our big site is nowhere at all for those same keywords.
Hello again guys.
I am wondering if my site is penalized. I want your honest opinion.
Here's the thing. I have a web site which received an avg of 4000 unique visitors daily. For 5 days now it's been "hit" and my daily traffic barely crosses the 400 visitors mark. However I have noticed that new pages are indexed in less than 12 hours, even sooner. For example I created a page 7 hours ago and now I see it indexed.
My question is: if my site was penalized, would Google index my new pages so fast?
For the kind of changes we're disussing here, no one has mentioned a slow down in either spidering or indexing speed. So I'd say you might be seeing this same phenomenon hit your website.
But you should definitely keep an open mind about other more established types of problems. It might be important that the trouble kicked in for you five days ago - that's a couple days later than the reports that started this thread.
| This 242 message thread spans 9 pages: < < 242 ( 1 2 3  5 6 7 8 9 ) > > |