Forum Moderators: Robert Charlton & goodroi
Over and done with?
All done all through?
mmmm... Is it possible that G can analyze a sites recip links and figure whether a certain percentage of the links link back and then downgrading those links. Say for example that a site has 100 links and 75% of those links link back. This could be downgraded in the algo while a site that has 100 links and only 50% link back, no downgrade. It could even be on a sliding scale in which the higher percentage of "true" reciprocal links, the more the effect of the downgrade.
Of course, if this is true, this along with the many other factors considered in ranking may make it difficult to pin down.
Just some thoughts :-)
(I am not targeting them as I have many good friends there)
The quality of work from 95% of those I have tested out is terrible. They copy content from other sites and claim to write their own unique content, they are copy replace professionals, this leads to Google penalties for sure, I would know.
>>>>>>>>>>>>>>>>>>
I think all companies whom you met are 1 man companies claiming to have a big team for SEO and doing all work themselves,
i am from an indian company and we have 6 copywriters to write stuff for our client sites, All copywriters are Postgraduates in masscomm, viscomm, literature etc, We are a team of 30 members all dedicated just for SEO, we do only SEO with a 30 member team, We are much appreciated by lot of U.S companies when it comes to quality of work, So I recommend you be more specific in blamming Indian companies, We know what we are doing and well trained to do that,
widget term0, widget term1 , widget term2, widget nextterm
our title read "widget"
we dropped of to # 4 ..
The I just did the same as him:
widget term0, widget term1 , widget term2, widget term3
and #1 we still are on this term .. even after Jagger .. so I dunno possibly a chance event who knows.
Still would like to know where my traffic is ... I removed a description tag: saying blabla bla: widget
Whereby the blabla bla always was the same: whereas widget changed.
and I put rel="nofollow" on the reciprocals between the main two sites ..
Let's hope G gets it over time..
By speaking up, you have just validated Hollywood's point.
On topic, has anyone seen any more movement on the J3 rollout? I have only seen everflux today, as G picks up the changes I have made over the last month.
Back to watching
WW_Watcher
Edited to add,
Most of the sites in the serps for the terms I tested, other than mine have been static, and not moved.
Yes, I have noticed those sites requiring a recip link and those having a high % of recips are hurting.... I know many people will not like that if it's true, but it was only a matter of time before the agressive recip linkers got caught. If I can map out recip links on sites/networks and such, I'm sure it's easy enough for Google to do it. I don't believe they are penalizing those sites - they are simply discounting those links and the sites rank where they would without them. In a nutshell if 98% of your links are recips - you lost 98% of your links. With the toolbar PR update I think they are trying to discourage link hunters.I'm not hurting on any site as I never got involved with recip/hunting links (I don't spend the time and wouldn't trust anyone else with that job).
Same here. I was waiting to hear what you had to say on this one I have to admit. One reason this isn't more obvious is that certain type of link schemes evaded Jagger, and others didn't. My suspicion is that this came from the initial google spam directory seeding that happend about 1 month before the first sign of jagger.
Plus I think Google also pulled back significantly on their backlink degradation that was more visible in jagger 1 and 2. Probably changed the serps too much is my guess.
No site I know of that has had no artificial link building done, including reciprocals or anything else, has dropped at all. I really hope that Google focuses more and more on trying to automate as far as possible fake link building scheme detection, that will severely cripple the seo industry. Of course it's hard, and seos will find workaround, but those involve more and more emulating real sites, which is more time consuming long term.
This is only my opinion, but I believe far too much attention is being paid here to canonical, duplicate content, etc issues, that is, onpage factors under your direct control, and far too little on backlink spamming, in whatever form it appears. The reason for this is obvious, it's easy to fix onpage stuff, it's very hard to fix broken link schemes.
Execution is everything.
In particular those not under your control.
You be good now?
You remember that question you asked me about if I thought Google got caught up in the links coming and going?
I think my reply was along the lines if Google didn't have enough sense to age the links it could cause problems and they would get what they deserved.
Well I think they don't, and that it does, and the pay back is a coming.
I looked at a site earlier today and between possible major duplicate content issues and massive IBL churn it is totally in limbo land.
There are all kinds of interesting things going on.
3 link farm networks, two run by folks in Austrailia, one is hiding behind a privacy screen, all three networks are hosted in the US. They are also all using the same software.
There is no way off of their pages without clicking an ad, half of the ads are Google ads, some of the others terminate at other real useful sites but go through plenty of indrect redirects picking up click pennies as they go.
If that isn't enough one of the end sites produces a list of pages every link of which goes to a get your medically induced 4273 66 type of site.
One of the other sites is just pages delivering Google's home page at the end of every link. Must be trying to get PR10 or something.
Me thinks the bit twiddlers need a new game.
Within some of the Datacenters, some of the servers have it, some do not, so it is spotty at the moment, but rolling out. With it distributed on so many subnets, IMHO I would think it would not take too long now to finish the rollout.
Edited to add
64.233.161.99
Back to watching
WW_Watcher
[edited by: WW_Watcher at 10:20 pm (utc) on Nov. 13, 2005]
You seem to be right!
I did a check on 64.233.161.104 my defualt, and it ain't happened there yet :-)
For my 4 keyword combo 1-2-3-4 "free spinning widget plans"
NOT rolled out. Has rolled in and out 4 times. Now it's "OUT."
Combo 2-3-4 "spinning widget plans"
Rolled out 3-4 days ago, stable.
Combo 2-3 "spinning widgets"
Has not rolled out ever, new changes in J3 seen as late as last night.
I can't make heads or tails of these differences I see with the different keyword lengths I watch.
That is one of the datacentres with very very ancient supplemental results in it, and that set of ancient supplemental results was also in 66.102.7.99 and 66.102.7.102 all through the Jagger updates (and is still there now).
Each of those IP address goes to a load balancer, not a single box, some of the boxes in the server pools have J3, and others do not. The least busy box in the pool responds to the query, that is why it shows up one moment and not the next.
For my test I have been watching a 3 word combination, that in J2 had a single result for the keyword combo in place 6, on j3, it has indented result and takes place 6 & 7. I would have not reported it, if I had not seen it.
Back to watching (football at the moment)
WW_Watcher
Edited to remove an "e"
;-)
[edited by: WW_Watcher at 11:33 pm (utc) on Nov. 13, 2005]
The lack of supplemental results for you may be an *algo* thing - the supplemental results simply not actually *ranking* for anything, but they are still there in the database.
I have some specialised searches where *all* of the returned results are supplemental - and in that experimental DC there are a LOT more supplemental results and their cache dates are all much older than elsewhere.
Across all DCs there is new data being added all the time. I see cache dates less than 24 hours old everywhere, but some DCs are not always flagging them with an attached fresh date in the SERPs.
[edited by: g1smd at 11:35 pm (utc) on Nov. 13, 2005]
This is a mom and pop small hotel in podunk, foreign country. This site has been online since 2001, I took it over and launched a new site on it in May 2004.
At that time, all of the new was cleanly spidered and the old pages 404'd and went away. Ah, the good ol' days.
However, now here at the end of Jagger3, the old pages have come back in the site: command on 66.102.9.104 with cache dates of April 2004
How far back can G be reaching for this non-existent content! Geez! That is more than 18 months!
And yes, in Jagger3 this site has dropped from page 1 of serps in this podunk town to page 2.
How far back can G be reaching for this non-existent content! Geez! That is more than 18 months!
I just found a supplemental for a page that hasn't been on my site for over 5 years! I couldn't believe it. Where are these things coming from?
>> 66.102.9.104 is an OK DC, now it's time to roll out and stop playing around IMO :)
I was very glad to see one of my usual book related sites coming back on 66.102.9.104 while other DC would surprisingly wipe it out (it's really a major payer on the market) for no reason on 1-only related keyword.
I understand the need of testing and bug fixing and QA and bravo to GG for all the follow ups and everything but I don't know for you guys but I think it's time to move on...it's been almost 2 months now.
You can't run and you can't hide.
Well actually what they think was your site (there are some disconnects but who is keeping score?).
This would also include nonsense entries (seems edits were not strong suit from some things I've seen).
Heaven help you if those entries foul up the process (my number one favorite word to despise).
Finding pages in same position DC after DC as you listed -- with 11/9-10 cache dates.
Question: Should we expect new crawls/cache dates over the next few days, or is this rollout most likely "it" for the Holiday season?
The everflux, adding of new data, and mixing it with the old, never stopped, and will not stop. G-Bot & Fresh bot will continue indexing old, new, and modified pages and feeding it to G to be put in the index, constantly changing the results.
Only big G knows when the next update will be.
Back to watching,
WW_Watcher
It is great to get an answer that exactly & comprehensively answers the question.
Thank you.
kperr
Always remember, this is IMHO, as I do not work for G, and have to make my best guess, based on real world experience with datacenters & load balancers, and what I have picked up lurking around here. I have almost never posted in the past, just been watching and learning from others here.
I do not feel qualified to comment on most issues dealing with webmastering, but big datacenters, with lots of boxes, all of which have to be updated all the time, that was my past life.
Back to watching
WW_Watcher
Guess it was a case of asking the right question of the right person.
Btw, I'm a career advertising wonk fascinated by this most pervasive of media.
kperr
These pages just have links on their sites to my page - similar info content. You know like "Look here for more info".
Is this fluxing 9 or another pulling of the results?
In most cases, I agree, what you are seeing is right. But in that, I would have the people who give the projects to share responisibility.
1. More often than not, companies approach Indian SEOs for sake of cheap prices. I have seen being in Bangalore, most offers are less than what we charge Indian companies. You offer peanuts, you will only attract monkeys.
2. Many Indian SEOs work as contract jobbers. Infact I heard some US SEOs call them "workers" :) They are given a task, and they merely execute.
Its not about having something in black and white, as saying This one is Good, That one is Bad. Its all about making right, informed decisions.
Ok, back to watching Jagger3 :)
I see you still talking about Jagger3 :-)
Now why they keep Jagger3 in Jail and wouldn't allow it to move on with its life?
Has Jagger3 been very nasty and needed to be taught a lesson by limiting its freedom of speech and movement to only few DCs?
Or are we gonna see Jagger3 spreading first at X-Mas Eve :-)
Who knows?
