Forum Moderators: Robert Charlton & goodroi
These were all dates when Google applied their Filter Updates (Allegra was also an algo update). Some sites dropped out of sight, some reappeared.
I'm very tempted to say "where have you all been for the last year?". There have been threads that have discussed this phenomena after each filter update. I don't think anyone has figured out exactly what the filter is (we all have our pet theories) but every thread has talked about the devastation to traffic of dropping 100+ places even on your sites own name.
If this is the first time you have been hit then you have been very lucky up until now. Some of us have been buried for a long time and have come back into Google over the past few filter updates. Whether our re-appearance is related to whatever actions we have taken is uncertain.
How about May 20th?
I see new counts on:
64.233.167.104
216.239.57.104
216.239.57.98
216.239.53.99
Nevermind, numbers went back as soon as I posted this, checked them 4 times before posting too!
Google has had multiple technical failures the past few years, some catastrophic and some small. In some cases, Google employees were clueless about the problems until webmasters more or less forced them to see, with Google Guy's unfortunate comments about 302s six months ago as the best example. They didn't know how screwed up they were, despite it being obvious.
The Supplemental index was a stupid idea that has some merits, similar getting out of dusting the living room by setting fire to the house. However, unlike other things, the Supplemental index is seldom seen by the public, so the humiliation it brings Google is limited to people in the search-conscious community. Eric doesn't tell cnet about how proud they are to have listings for pages deleted two years ago; that a webmaster has even told Google to delete from its index more than once because it doesn't exist; that Google has crawled to and seen as 404 literally 100+ times, etc.
Google's problems include mistaken notions, but its mostly ineptness. No one at Google sat down before the recent mess and said "hey, let's look incompetent by making stupid decisions that anyone can see by adding &filter=0 after our search page URLs."
This isn't an easy job they have, and sometimes they do a terrible job at it, even if they do better than their competition.
There will always be tin hat conspiracy posts here, but the fact of the matter remains that dropping a glass on the floor is not "updating" it.
[edited by: steveb at 6:47 pm (utc) on Sep. 28, 2005]
Of course everyone affected has been searching for the answer. 302 redirects (everyone has a few pages with 302s pointing at them), www/non-www issues and/or duplicate content, adding large numbers of pages, Adsense, affiliate links.
I did examine closely 3 very good, non-spammy sites hit as well as my own. In all cases (including my own sites) I could see a quite large proportion of links coming from a few domains. So, as a veteran of the PR0 cross-linking penalty I tend to favour linking patterns as the trigger.
There are a few other threads that has discussed this. Even one with a note from GG.
[webmasterworld.com...]
[webmasterworld.com...]
Anyone ever recovered? Any hints at what could be causing the penalty? (Ex: Add the "&filter=0" onto any term which you used to rank puts you back in the right place.)
My bet is on scrapers causing this issue.
so you think this filter isn't related to all those other filter updates? Certainly I've had one site come crashing back into the SERPS with this update after being missing since February. I just think that this regular pruning of search results has to be deliberate and stems from a filter (probably with multiple factors applied differently each time) and that this update is the just the latest in the line. Having the &filter=0 to spot the filter is useful though
what do you mean causing this issue?
do you think these filters are designed to catch scraper sites and other sites are getting caught?
if so, how do you think these filters work?
To those fellow members who keep posting: This is not an update
Would you be kind to post your own definition of an update?
And lets discuss it.
Thanks.
I dont know - as you know though I think it is another attempt at a fix for canonical urls - they still have not got it right though :(
Seriously - this should be the main priority at the plex at the moment (IMO - of course)
Listen Google - I will say this only another 96 times. The canonical url for my site is the homepage with the www - I have done the 301 - this is the page with the most backlinks - it is the page that should rank for the company name search. Etc.
if so, how do you think these filters work?
---------------------------------------------------
Scrapers are rampant. They copy titles and meta-descriptions. If I look at one article of ours, there is at least 100 different scraper sites with the same title and meta description of the article. On our site, the meta-description is the first few lines of article. We also have index pages that link the articles together. These index pages are made up of titles and meta descriptions of the articles.
I imagine that Google use some type of Bayes classifier for the filters. In order for these filters to work - they must be manually fed a list of bad sites or sites that they would like to penalize or rid the SERPS of.
Since the scrapers contain the same title and meta description tags, it not a stretch to see that our site would get caught up in the filter.
1)google has said they prefer algorythmic ways to detect spam
2)that idea seems like something a building full of phd's would see right through, as well as any half-decent spammer.
>>Reseller
I dont know - as you know though I think it is another attempt at a fix for canonical urls - they still have not got it right though :( <<
You mightbe right!
However, I see the problem in more simple manner:
Google keeps listing 1000īs of duplicate pages while at the same time removing "original pages" as part og deduplication process
And its hurting...
This is not an update tread.
This is not a "let's define what an update is" thread.
The thing is long and cluttered enough without this extra crap.
Start a new thread, or contribute to one of the other similar threads, if you want to discuss what an update is.
