| 8:38 pm on Apr 1, 2008 (gmt 0)|
Update: Now every other click or so is moving around pages in positions 2-8. Some are dropping completely off page one and some are just swapping between two spots, like #2 & #3.
Interesting. I think I'll quit watching for a while before I get an ulcer :)
| 8:42 pm on Apr 1, 2008 (gmt 0)|
Mike, are the ip addresses switching when the ranking switches?
| 8:49 pm on Apr 1, 2008 (gmt 0)|
rdrysdale - I'd like a piece of that ranking pie.
| 8:50 pm on Apr 1, 2008 (gmt 0)|
Tedster - not that I saw. It's been 126.96.36.199 the whole time. At first I thought I was nuts with the fast changes, so I started using the IP to make sure I kept the same DC. Now it's not doing it though... :(
Maybe I'm losing my mind.... I was watching it do it on google.com and google.co.uk. I have sites in both places in the same industry and they were bouncing all over the place. Now it seems to have gelled up. Maybe it's moved on to a different DC now.
| 9:04 pm on Apr 1, 2008 (gmt 0)|
I see a visitor on raw logs come from one search string, try it myself, and, poof, nowhere.
This is one funky dance! Let's call it the Elaine! :-}
| 9:06 pm on Apr 1, 2008 (gmt 0)|
i like it!
"like a dry heave set to music"
| 10:18 pm on Apr 1, 2008 (gmt 0)|
35C in Cape Town today - and my sites went up in rankings!
Its been a good day :)
| 10:39 pm on Apr 1, 2008 (gmt 0)|
|the Elaine.. "like a dry heave set to music"|
|I like it too! (and love the analogy) |
| 10:42 pm on Apr 1, 2008 (gmt 0)|
My 4 year-old site was previously in the top five positions for the phrases I wanted. Some are still in the top five, some are now in the top ten, and a few have moved to page 2.
If I do a search on www.google.com, I'll come up #1 for many terms, which tells me that Google is watching what I search for. Probably a cookie or something, and it's probably something that's been covered in another thread. If I check the datacenters, I'm not #1 for those terms.
The idea that Google is giving lower-ranking sites a shot at the first page is an interesting one. That would explain some of the sites that have come out of nowhere.
| 10:43 pm on Apr 1, 2008 (gmt 0)|
|I see a visitor on raw logs come from one search string, try it myself, and, poof, nowhere. |
I am thinking there may well be a geo component involved.
PNW here. In one niche I am seeing many of the "new and odd" returns coming from west coast - including BC! The BC sites are ones that I have never seen rank in the top 50, and would only seem to have relevance if one takes into account our location.
| 1:42 am on Apr 2, 2008 (gmt 0)|
Well, I take it back. All those BC sites have disappeared again. I think I'll go back t watching the grass grow...
| 2:25 am on Apr 2, 2008 (gmt 0)|
>>I think I'll go back t watching the grass grow...
Great time of year for that too.
| 2:53 am on Apr 2, 2008 (gmt 0)|
|I think I'll go back t watching the grass grow... |
Sorry I don't see any grass but snow up here.
|Now every other click or so is moving around pages in positions 2-8. |
So Google has several sets of search rankings and shows them randomly each time a search is made.
| 5:50 am on Apr 2, 2008 (gmt 0)|
Hey all, I asked a few people to look into this and they weren't seeing many large differences in rankings between these datacenters. The issue with discussing on this thread is that specific urls/queries aren't allowed. If anyone wants to mention a search where they see large-scale differences, feel free to send feedback to Google in the usual way. I'm going to pick a random-but-pretty-unique keyword so that I can look up reports. Let's use "dewey" as the word. So if you want to mention a search where you think the results are very different at one data center compared to other data centers, use the spam report form at [google.com...] and make sure to include the word "dewey" in the "Additional details" text area. Or feel free to point out differences in other ways: do a blog post, leave specifics on the Google webmaster help group, or whatever way you want to point out specific searches that look different to you.
The usual rules of thumb apply: you probably won't get a personal reply, but I'll try to get someone to check out reports that get sent in. There shouldn't be much difference between data centers, so I'm curious to find out what queries people seem to be seeing different results on.
| 7:00 am on Apr 2, 2008 (gmt 0)|
Thanks for chiming in.
How about different data ceneters displaying much different page numbers (up to 30% difference) for site: operator? should that be reported too using "dewey" ?
| 9:12 am on Apr 2, 2008 (gmt 0)|
|I've had one site online since 1997 that had a lock on the top spot for my keywords on and off for four years on Google just go "poof", all that's left of it is two obscure pages I haven't changed since about 2001 ;~) |
This sort of nonsense has been coming from Google for years and they just seem to get worse. They are currently sending traffic to pages that have been 404'd for over a year. I tend to agree with the other threads that are going on about how Google is just on an all out war with natural rankings and SEOs because relevant natural results really damage their bottom line. They spend more time making their PPC results relevant.
Made the decision about a year ago to push for a No.1 ranking on Yahoo/MSN (who might still become the same thing) and to push for more referral traffic. Glad I did, I'm not part of those "99% of my traffic disappeared last night" anymore. This huge swinging update has barely affected the majority of my sites.
|So Google has several sets of search rankings and shows them randomly each time a search is made. |
Here in the UK rankings seem to change every 2-3 days rather than every search. As far as I can tell, there are six different results/algorithms that have been rotating for the past month or so.
| 1:16 pm on Apr 2, 2008 (gmt 0)|
I'm starting to see some stability (fingers crossed) in my sector although one of my sites is still MIA yet one of my .mobi has risen to 14th!
Still a few too many "shopping" type of sites for my liking but at least my traffic seems to be stabilising...at the moment.
There seem to be far fewer indented results with the SERPs going direct to the correct pages.
| 1:47 pm on Apr 2, 2008 (gmt 0)|
Seeing some god awful sites getting into top spots for competitive terms by using huge, irrelevant reciprocal link networks or by pointing links from all of their other sites to one domain (over 200 sites all visibly owned by the same company pointing to one site). Not impressed by this update so far.
| 1:49 pm on Apr 2, 2008 (gmt 0)|
This whole thing remineds me of Update Jagger-1-2-3 [webmasterworld.com] :-)
| 2:55 pm on Apr 2, 2008 (gmt 0)|
G's at it again. I'm at DC 188.8.131.52 and positions 2-8 are hopping all over for my niche. This time I have screen shots. :)
Someone mentioned naming this updated Elaine. I think it's perfect. Google's thumbs are in the air and it's legs are kicking all over.
| 3:03 pm on Apr 2, 2008 (gmt 0)|
Call me paranoid, but my first thought on this when it all started was that Google was trying to boost its bottom line by knocking off websites with high organic listings. This would force formerly top ranked sites to use adwords after their sites( and income)tanked in order to get their sites seen.
I banished the thought when I did a quick calculation as to how much revenue such a plan would have cost Google if these same banned sites had hundreds of adsense adds suddenly removed from circulation. Ah well, the joy of SEO, it's seldom boring and is always changing.....
| 3:34 pm on Apr 2, 2008 (gmt 0)|
reseller, feel free to send that feedback too. As I said, the folks I've talked to have not seen major differences between data centers, so that's why I'm curious to hear what people here are seeing.
| 4:36 pm on Apr 2, 2008 (gmt 0)|
Matt - Up here in Canada (using google.com), the rankings are dominated by .co.uk and .au sites - these were gone 3-4 weeks ago, but are back again. Doing this same search from a proxy browser out of the US, produces completely different results, mainly all these .co.uk and .au sites are gone.
I've sent in several spam/feedback reports on this, but nothing's changed. Since these sites are useless for users in Canada and the US I'm not sure why Google brought them back.
Just wanted to bring it to your attention!
| 4:42 pm on Apr 2, 2008 (gmt 0)|
I guess I'm not sure what "big differences" means. On my top keyword I'm just seeing sites in the top 20 varying from 4 to say around 15. The top three seem to be solid.
This is more like the old Google dances. Doesn't seem like everflux though.
| 5:07 pm on Apr 2, 2008 (gmt 0)|
|I banished the thought when I did a quick calculation as to how much revenue such a plan would have cost Google if these same banned sites had hundreds of adsense adds suddenly removed from circulation |
The made-for-adsense spam sites are all ranking just fine. Google's efforts are to fight good SEO, not bad SEO as bad SEO generally has to rely on Adsense for revenue. Good SEOs work on sites that can actually move to an Adwords model should their rankings be snatched away. That's why they had the big clamp down on paid links - people were paying to boost their natural rankings which had a better ROI than Adwords.
I might sound harsh on Google so just in case people think I'm ranting because I've been dropped, I haven't. My rankings are just fine and on the variant datacenters I show increases for nearly all my sites. I did lose a couple of places last month after removing Adsense from some of the inner pages of a site but that's it.
| 6:56 pm on Apr 2, 2008 (gmt 0)|
A new observation is, that now any freak can slash a bunch of embedded youtube clips with loads of adsense together to earn big time by riding solidly on No.1 for prime international keywords.
We had this No.1 spot since years with loads of selfproduced, expensive, exclusive content that has earned tons of organic inbound links from odp, yahoo etc.
Of course it is acceptable that other real content sites may take over that spot, fair enough. But now a few bucks for a domain and some freehost are enough to sink some of the top media brands.
Perhaps time to say bye to content production.
| 8:23 pm on Apr 2, 2008 (gmt 0)|
Another observation is a case in which a site with as little as 3 PR 0 inbound links of which one is even paid and labeled as "sponsored" at an otherwise empty domain grabber page.
Such stuff leaves now the ultimate content and authority site with even wiki inbound + loads of related relevant inbounds behind.
The nice thing is, that i know the story behind that newly successful site which had been somewhere well deserved around 100 before due to loads of errors and not much relevant info.
From a Google point of view: The owner of that new top 10 site spends a lot on adwords.
Would be great to see this among Google case studies :-)
[edited by: night707 at 9:19 pm (utc) on April 2, 2008]
| 9:05 pm on Apr 2, 2008 (gmt 0)|
"Dewey" .. I thought we settled on "Elaine"?
| 9:13 pm on Apr 2, 2008 (gmt 0)|
Matt Cutts chose "dewey" as the code word for communications about the changes, so we went with that name. Using Dewey helps cut down possible confusion.
[edited by: tedster at 9:14 pm (utc) on April 2, 2008]
| 9:13 pm on Apr 2, 2008 (gmt 0)|
|"Elaine, the pain! The pain, Elaine!" |
- Lyrics to a Korean karaoke video w/English chorus
| 9:33 pm on Apr 2, 2008 (gmt 0)|
And that reminds us of when Matt asked for feedback about BigDaddy Update [mattcutts.com]. It was also Matt who chose the name BigDaddy ;-)
The best place to do that is to go to [google.com...] . In the "Additional details:" section, I would use the keyword "bigdaddy" in your report.
Do you think that he is telling us that what we see is just a software update (infrastructure update) ;-)
[edited by: reseller at 10:00 pm (utc) on April 2, 2008]
| This 157 message thread spans 6 pages: < < 157 ( 1 2  4 5 6 ) > > |