Forum Moderators: Robert Charlton & goodroi
Have been waiting for google's pending update but didn't get any on that. From last two days I have noticed extreme changes in search results for same google domain (like .com or .de or .nl etc.) with different interface language selected (like hl=de or hl=en or hl=nl). When you change the interface language the whole bunch of results are different.
Is this the actual update this time? Enabling the geo centric data which google wanted to implement for better localization? If yes (and if someone else is also seeing this change) then may be we can all identify the parameters of this new change, like language, servers in geographical locations or extensions (.de, .nl etc.)
-- rishi
216.239.59.103 - Based on what I've seen with my website, this is where I think the results are going.
Gee, I hope you're right ... in some cases I see slight improvements in positions I had last week.
My sites are all strong on content, but are "optimized" and well-linked, and if the earlier speculation:
reseller, it looks like a filter. The order of the SERPs are the same, with few sites yanked out of their original positions. The intent is clear. The sites that try to influence google ranks, whitehat or otherwise, are punished. But looks like whitehats affected the most. Nothing to do with on-site factors.
I continue to see variable results ... at different times this morning, I was getting "back to last week" results for most sites I manage or monitor, and at other moments, they are no where to be found in the top SERPs where they've lived since I recovered from the Florida update ...
Oh, isn't this fun! :p
is correct, I don't expect to be happy. If indeed, this filter is about "filtering out" sites with strong recip links ... which I consider to be white hat ... then I will have some work to do ...The line between white and black hats is badly faded, broken in spots, and moves irregularly. There is a difference between a wedding reception site trading links with a local dj and a php site trading links with an Istanbul Hotel Affiliate. One is a legitimate reciprocating vote while the other is an attempt to manipulate the rankings. Three Way links are almost always a way to manipulate rankings, but many would argue their validity for a number of reasons. If you ever do find the line then you have to deal with the fact that Google really doesn't care about the line. If the result is relevant, it doesn't matter how it got there.
There is a difference between a wedding reception site trading links with a local dj and a php site trading links with an Istanbul Hotel Affiliate. One is a legitimate reciprocating vote while the other is an attempt to manipulate the rankings.
Indeed. And I only trade with similarly themed sites that will benefit my site even if there is no benefit in the SEs.
Even the first instance (wedding-DJ) could be a tactic aimed at manipulating position. Nonetheless, even if that's the aim, there is still a legitimate reason to trade that link, as someone visiting the wedding site might well need the DJ and vice versa. So, even if Google discounts the value of that link trade, it remains a viable strategy to drive traffic to a site.
From what I saw in the SERPS of the "bad" DCs, many sites that are otherwise strong in content and should rank well even without the benefit of link strategies, were not there. It *looks* to me as though a penalty/filter is being applied to sites that are well linked reciprocally despite strong content ... but I trust that whatever the outcome of the current shuffle, Google will adjust those knobs until the sites that deserve good position will eventually have it again.
More will be revealed.
Cheers, MJ
In the technology news this weekend, alarms have sounded regarding a cross-scripting vulnerability some company discovered regarding G's AdWords. Cross-scripting is basically the ability to enter server-side or client-side script via regular entry form and calling up the page that displays the data. Hence executing the code. In this situation, the script revealed the viewers G login cookie to the culprit. VERY dangerous for G. Trust with our info has yet to be violated, at least for me.
This discovery happened on 9/20-21 which coincides with the catastrophy we all encountered. Do a search on the web for such news (10/9 and 10/10) and see what you can come up with as I cannot post the link. And now that the problem is fixed, we are all watching the DCs dance.
Now, here is how you can test this theory. Go to the goofed up SERPs in G, and enter a power keyword you would normally appear high in. For us, the keyword is music charts. As it appears, none of the SERPs on the first few pages have AdWords on them. So, to minimize exposure to this threat, they appeared to yank all the pages with AdWords for powerful keywords. Now, this concept hurts all 3 parties involved: the publisher, subscriber, and the ad network, but it does maintain the integrity of the organization.
As for duplicate content, conical url's, or the bacon potenta, I never bought into those concepts. Too much RSS and syndications going on to attempt to nail down duplicate content at such a finite level. Besides, if that was the case, their index would max out at 1 bil pages (which is not good for competition right now)... In my mind, I am now calling this bacon propaganda. Matt Cutts did good in diverting our attention and getting us all to check our sites and suffer a reality check. It was basically turned on us as if we stole our own cars. My suggestion is that tweakin your site now will throw a big flag. Hold still... I think the DCs are back dancing for an utter total recalc.
[edited by: Yippee at 5:41 pm (utc) on Oct. 11, 2005]
For the Google on the IP address you mentioned Google is serving up images along with the SERPS...never seen that before.
Try:
[66.102.7.105...]
Just a general phrase, but take a look at the results....
Mick
Interesting observation textex. Do you have enough sites that have dropped to be able to confirm this with little doubt in your mind?
I've had a similar suspicion. How far did the sites drop, 5 places, 10? or more? Did you have sites that dropped, and sites that didn't?
Interesting observation textex. Do you have enough sites that have dropped to be able to confirm this with little doubt in your mind?
Sorry textex for answering the question that came your way.
Yes. Can say that with enough evidence. More the commercial value of a keyword, bigger is the drop. On-site factors just don't come into the frame. Only to do with, if the links are achieved naturally or artificially.
While the intent on the face of it looks good, I fear Google has just compromised on the quality of SERPs. In the absence of smaller "optimized" sites, SERPs are full of sub-pages of big sites that have an oblique reference to the keyword. It happens such that sites with naturally attained links form a SMALL FRACTION of the pool of sites that is available for Google to choose from, and they have to fill-in for a LARGE FRACTION of search queries, for Google dislikes majority of smaller sites that don't have the purest of naturally achieved links.
So, what you have effectively is a [Total Google index - Sandboxed Sites - Sites without natural links - Sites suffering from Canonicals]. There could have been better ways of dealing with the matter.
I really hope the above equation does not stick.
Google is really setting itself for its own demise.
Yes, there are always people that are going to say they like the new results, but those that 'know the know' can easily see the junk be chalked up.
Not yet.
Wait until MS makes a deal with AOL.
Wait until IE 7 is released.
Wait until Windows Vista is released.
In the meantime, your option is pay-per-click. :)
Is this the canonical url problem?
Also, do you think i should abandon my link campaign altogether? i linked to related sites, in my opinion, but who knows what google thinks. my home page still ranks OK in the new update (the bad 1/2 of the update anyway..we'll see), its all other pages that have disapeared. Will everything in your domain get penalized, if google doesnt like the relevancy of some of your links?
The similarity, active link exchanges."
I am seeing the exact same thing here. Especially for the big name keywords. The more strong recip linking involved in tandem with the higher money keywords, the more the drop. This is consistent amongst all of my sites. This sets up an interesting scenario should it prove to hold water
Put in context, this could be called devaluation more than a filter perse if it holds to be true. Google may very well not be penalizing sites with recips, only not counting them or devaluing them based on some %. Hard to say fully.
This to me is going backwards, as "natural linking", for what it is, is a tough sell. Suppose an authority .edu site links to me via an article they wrote praising my service or product. Now suppose I wish to show my visitors that an authority site has recommended me and I link back. Is this reciprocal linking, sure. Does this mean this link would be devalued? Possibly. Slippery slope to be sure. :P
Were there any other questionable things done, like site wide keyword linking to home page for example?
If you're right, google made a rough map of these networks a month or two ago, and even if the stuff is cleaned up, it's still working from that map.
In the case of multiple domains, is the registrant information the same, or does it use the domain privacy function of the registrar?
These link farms have been an obvious target I have to admit, they are really getting out of hand as an seo tool, so it's not surprising that google would start targetting them.
CainIV, you might be right, that would account for the variable drops, if link farms are being dropped, and the credit for those links discounted, that would drop a site by x number of serps, depending on the mix of natural and link farm links. Makes too much sense.
I am seeing sites that don't actively look for links and definatly do not participate in link farms and don't have a network of sites all owned by the same person disappearing.
It seems as though the key to survival, at least in the sectors we watch, is to be 'big gun'.
What we see is a black and white issue. Link exchnages/pages = gone!
>>yes, but i do see large sites with extensive link exchanged in the top of results.
So link exchange is ok for large older sites but not for smaller sites.<<
Maybe the "NATURAL GROWTH" of inbound links is an important factor too.
Older sites might have had enough time for natural growth of indbound links.