| 12:07 pm on Apr 3, 2010 (gmt 0)|
As far as the suggested reduction in URL's is concerned, there has indeed been a >50% reduction in the number of results presented before the familiar message kicks in - "In order to show you the most relevant results, we have omitted some entries very similar to the 414 already displayed.If you like, you can repeat the search with the omitted results included." - this used to be in the 900's.
Little doubt that the numbr of pages now in "supplementary" is much greater than week or two ago.
| 1:39 pm on Apr 3, 2010 (gmt 0)|
Toolbar Pagerank update underway. On schedule ... unlike their Caffeine roll-out ;)
| 1:53 pm on Apr 3, 2010 (gmt 0)|
Google is surely updating its pagerank!
Two of my sites are dropped while one new site increased from 0 to 4 .
| 2:20 pm on Apr 3, 2010 (gmt 0)|
pagerank updated for one of my sites too 0 -> 3
< Note: PageRank Update thread [webmasterworld.com] >
[edited by: tedster at 12:28 am (utc) on Apr 4, 2010]
[edit reason] Add the link [/edit]
| 4:41 pm on Apr 3, 2010 (gmt 0)|
I have noticed a real ahake up, my site has dropped for a number of keywods it was high for, will the change revert back to normal as the sites which are up there make no sense.
[edited by: tedster at 4:48 pm (utc) on Apr 3, 2010]
[edit reason] moved from another locatioin [/edit]
| 4:52 pm on Apr 3, 2010 (gmt 0)|
re:SERPs Yes, it has reverted but it is more progressed than pre-Easter. i.e. the current SERPs are fixed more than pre-Easter - what I saw yesterday was the furtherest out the fix has gone in months, but they pulled back, not all the way, but back.
It's like, they totally went for it yesterday, tried to push out, say, 70% of the new datasets they have, it was chaotic, then by the end of the day, they pulled a lot of it back and have now settled on say, 15% of the new data being folded in. 0% -> 70% -> 15% is far from scientific, it is just a rollup of my own observations on my monitored keywords (2000+) across 6 different North American IPs.
I am risking condemnation ; ) by posting this unscientific observation because I am curious to see if anyone else finds my observations mirror their own. ; ) This is simply what I observed yesterday for my niches on my test IPs...
| 5:26 pm on Apr 3, 2010 (gmt 0)|
would the results be back to normal? I thought caffine does not change results? Why is this happening?
| 6:31 pm on Apr 3, 2010 (gmt 0)|
Caffeine is not an algo change, so it does not 'change the rankings' by increasing the importance of one variable and decreasing the importance of another. But, Caffeine is a change to the underlying data storage system the rankings are based on, so it has a direct impact on the final results through: data, speed of calculations, storage methods, etc.
One of the things I've read speculated they have been trying to do is keep the results generated from Caffeine 'essentially the same' as they are using the Big Daddy storage system, but any change to the underlying data can, and IMO most likely will, have an impact on the rankings to some extent.
So, Caffeine doesn't change what's important (algo), but does change the underlying data storage (Big Daddy to Caffeine), so Caffeine may directly affect the rankings, but not what's important to rank. If that makes any sense at all...
| 7:16 pm on Apr 3, 2010 (gmt 0)|
As of mid yesterday and today (saturday April 3rd), I'm seeing a few spammy sites make their way to the top.
| 8:40 pm on Apr 3, 2010 (gmt 0)|
There was a totally new group of results flickering in and out at about 2am pacific time the morning of April 1st.
If those results are what they are aiming for with caffeine, then they have nailed things pretty well. The results were caffeine influenced, but very different than the spammier caffeine-ish results on the different IP addresses. This is to be expected as the IP address results are never what actually appears on Google. The extra dose of secret sauce between IP address and display on Google has always seemed to be what caffeine lacked, but this batch that appeared briefly did have it.
Of course, whatever... they may not even be able to find their car keys for six more months, let alone this more mature bunch of data, but that date was the most significant development so far "after the holidays".
| 9:29 pm on Apr 3, 2010 (gmt 0)|
steveb: did you monitor them further on the 1st? Your characterization of what you saw matches up with what I saw on the 2nd, but I saw it last all day and into the evening. I mostly used an east-coast proxy, but some others throughout the US. Perhaps the difference was simply the nature of the SERPs we each monitor, dunno. I know IPs are a moving target, but where was the IP(s) you were watchign located? There has been some speculation that it moves from the west to the east... Curious. I have seen nothing shuffle today...
| 4:27 am on Apr 4, 2010 (gmt 0)|
Looking at IPs is useless. You never see actual results that way, only 3/4 baked stew.
The results I saw were on Google.com, both using an anon and not.
| 4:58 am on Apr 4, 2010 (gmt 0)|
steveb: sorry, I was ambiguous. I don't check particular IPs like that, I use proxies that get me to different parts of the country. I used "IPs" when I meant proxy, I didn't mean a different google IP, I meant my IP...
| 3:05 pm on Apr 4, 2010 (gmt 0)|
Not to go off-topic for too long, but how are those of you who are accessing Google through proxies doing that? G always refuses proxy servers when I try.
| 4:22 pm on Apr 4, 2010 (gmt 0)|
dickbaker - there are websites out there that allow you to browse the internet (including google ;) ) through their website, and in the process, browse using their IP address. A nice byproduct is that it is considered "clean and safe" web browsing:
"Anonymous (Proxy) surfing allows you to surf the net without having to worry about people knowing what you've done and where you have been. You can surf the web safely and anonymously by entering the url (website address) into the web address box and click the begin browsing button. The site will not know who you are, won't save any cookies or tracking codes, you are free to surf where you want when you want."
You will know what their IP address is, and thus, THAT is the area where you will be browsing from, regardless of where YOU are located. Get a proxy located in Chicago, and by browsing through their IP, you are getting results that are common in the Chicago area. And so on... ; )
| 5:27 pm on Apr 4, 2010 (gmt 0)|
Well I have to comment and share my experience about the above geo-targeting debate..
a UK site targeting UK customers and set in GWC as UK hunting long tails for a daily/seasonal long key-phrases.
Checking from my local IP (non US/UK) is giving me good SERPs for my key-phrases. Checking from a US IP giving me *almost* the same set of result. But checking from a UK IP (which is our target) is giving me totally different SERPs.
Even worse, checking .com and .co.uk is giving different set of results for each IP (non US/UK, US and UK ones)!
I don't see if that's something new due to the recent updates or this has always been the case, but with this insanity I am really considering purchasing a SERP monitoring service (not a software)!
Appreciate any useful input from UK webmasters.
| 3:40 am on Apr 5, 2010 (gmt 0)|
Has all the dust settled now or are u still getting those results.
| 5:41 am on Apr 5, 2010 (gmt 0)|
Those results disappeared after they briefly appeared April 1st early AM Google/California time. No sign of them since.
| 6:50 am on Apr 5, 2010 (gmt 0)|
exactly what i mentioned...
Google did the caffeine update but then due to total mess it reverted the changes gradually within a day :)
| 7:02 am on Apr 5, 2010 (gmt 0)|
Exactly the same thing I mentioned earlier on this very same thread on page 2 :)
|Pass the Dutchie|
| 2:01 pm on Apr 5, 2010 (gmt 0)|
Caffine is rolling out accross Europe. Seems to be sticking on google.de. No major shift in SERPS which look very similar to a SERP shift from last week.
So from .de pov SERP update happend before Caffine's bells a whistles were displayed.
This is more than just an update, this is Google 2.
| 3:49 pm on Apr 5, 2010 (gmt 0)|
|This is more than just an update, this is Google 2. |
Seriously? Just feels like a slightly "faster" Google with a few layout changes to me ... sort of a knee-jerk reaction to the Michael Jackson fiasco with a few distracting shiny things for good measure.
| 4:48 pm on Apr 5, 2010 (gmt 0)|
I see 0 changes over the past 36 hours. 100% static. Neither USA - East Coast nor Canada...
| 5:49 pm on Apr 5, 2010 (gmt 0)|
@Pass the Dutchie: Not seeing anything here for me on google.de - seeing the SERPs from the week before, not the flickering of Caffeine on 1st and part of 2nd April. Can you give any IP-addresses I might try?
| 5:56 pm on Apr 5, 2010 (gmt 0)|
Well I'm seeing geo-based search results when doing searches for what were national keyword phrases (i.e. widgets, as opposed to "widgets in location"). There was a way to turn this off, but now there is not.
Anyone else seeing this?
| 6:04 pm on Apr 5, 2010 (gmt 0)|
TheMadScientist your right about one thing what you said makes no sense at all. How do you know with Caffeine there isn't any algo change. I don't think you were around with the Big daddy but thousands upon thousands site tanked and yes there was an aglo change for it as well.
Won't effect your ranking please don't try misleading people by making this kind of statement.
Truth is non 0 nada of us have a clue what's changing we are all watching and trying to understand what IS changing.
|Pass the Dutchie|
| 6:10 pm on Apr 5, 2010 (gmt 0)|
Had a few hours of Caffeine earlier today and had a play. Arrived at home and bells and whistles have gone. That was the longest run of Caffeine I have seen on .de so far.
As I mentioned, SERPS had not changed while Caffeine's layout was evident. I have a feeling the SERPS are already Caffeinated or the SERPS will be Caffeinated after Caffeine's GUI is in place. Either way, in our neck of the woods, it seems SERP results remain unchanged weather Caffeine GUI is in place or not. This might explain why Saturdays PR update is showing the same results on both Caffeine and old Google results.
Yes I do think this is more than just Google returning results slightly faster from both a user pov and an SEO pov.
Will be monitoring my logs very closely.
| 6:35 pm on Apr 5, 2010 (gmt 0)|
How do you distinguish the layouts/GUIs?
I definatly have 2 different kind of SERPs which in every case seem to be an "old one" and an "updated one" with an algo-change (or at least more recent data influencing the SERPs while the "old one" seems to be stuck 2 or 3 month ago). I see them as Non-Caffeine and Caffeine (which might be discussed controversialy of course).
I identify "Caffeine" as well with some keywords where is not much flux in the SERPs and which has almost equel SERPs to the one on the test-DC we got last year. When I get this SERPs, I see the "udpated one" on other keywords, so for me this is "Caffeine" (or the resulting SERPs from the change of the infrastucture).
| 6:57 pm on Apr 5, 2010 (gmt 0)|
It is feasible that the differences of opinion on whether or not caffeine comes with an algo change may be as cangoou alluded to:
The pre-caffeine SERPs we see may be driven by a stale set of 3-month old data, and the caffeine SERPs we see may be driven by a more up-to-date set of data (how up to date, I dunno, I suspect better than 3 months old but not still not fully current).
So, it may be when "caffeine" is spotted, the SERPs seem different, and this is a legitimate observation... Even if caffeine isn't directly an algo change, it may be paired with a newer data set accounting for the SERP differences we see...
| 8:29 pm on Apr 5, 2010 (gmt 0)|
I also wouldn't use layout or GUI variants as a dependable way to differentiate Caffeine from de-Caf.
|Pass the Dutchie|
| 9:57 pm on Apr 5, 2010 (gmt 0)|
Yes, I would second that Ted. Full privacy, logged out of G w/ various proxys . GUI with same SERPs seem to be switching depending upon IP/browser.
Calling it a night.