|Thanks for the info. Does that mean no more data refreshes until the new year? |
IMO, data refreshes will continue to occure once a month or so.
However, you may wish to read this part of what Matt Cutts wrote today:
|I know that webmasters are especially sensitive to quality/webspam/ranking changes in Q4 because of the holiday season. If weíve got something that evaluates well and that we think will improve quality, we canít just pause for 1/4th of the year, but if anything big launches Iíll try to be available to answer questions and help get a handle on any changes. (Right now Iím not expecting radical changes in webspam ranking, but I know better than to make a promise.) |
It also reverts to an older one, and then sticks there, if you suddenly link to a bad neighbourhood from that page. I wrote about that effect several times before.
Can you explain bad neighbourhoods
You'll know one when you see one.
Link farms, free-for-all pages, fake directories, etc.
System: The following message was spliced on to this thread from: http://www.webmasterworld.com/google/3117217.htm [webmasterworld.com] by tedster - 3:21 pm on Oct. 11, 2006 (EDT -4)
Following the discussion on the trust rank.
It looks like old infrastructure datacenters for example 64.233.183.* are not indexing for new websites anymore and didn't updated the PR change a few weeks ago.
Had some major improvements on PR last update, these are not seen in these datacentres. Google news indexing from these datacenters are not showing newly added sites to the google news website etc.
Why is google not synchronising this infrastructure anymore? Or am i totally wrong?
This happened 6 months ago, and maybe at other times too.
Datacentres that had the new infrastructure and results on them were being updated regularly, but the last few datacentres that had data that was going to be phased out were updated less regularly - until one day their data was completly swapped to the new stuff. I guess, why waste resources on updating something that you know is going to be scrapped in just a few days time?
But do a PR check on all the datacenters and see for yourself. Anyway for some countries these datacentres hold the main datasets (NL/BE).
So you can understand that not updating web and google news for new websites is not really a nice thing. And probably also more countries datasets have this problem.
Matt Cutts said earlier today that several datacentres would be running on old stuff maybe for a month or two. In the bigger picture it isn't a major issue. Might even give some people a few clues as to what works best when you have two sets of results to compare...
Matt Cutts says he first noticed the shrinking site: results on the 5th. That's several days after it was first posted in here.
>> - For a brief while last week, site: only returned three results from a host. Someone mentioned it to me by email, but the first web report I saw was by DaveN on Friday (thereís your link, Dave). Fixed/working by the end that day, I think. It was related to a binary/executable that was going out, but a different binary than the one mentioned above. <<
|Matt Cutts says he first noticed the shrinking site: results on the 5th. That's several days after it was first posted in here. |
Yes. I was surprised to read that. And that indicates again that Matt hasn't been well informed :-)
Maybe he need to visit WebmasterWorld Forum 30 at least once each morning and each evening. His info for sure will be up-to-date, always ;-)
Two things in life are unpredictable:
For few days ago, I thought that The Mother of All DCs likes my keyphrases. Today I see that the lady has changed her mind.
Oh well..... women and DCs (:(
So that's why you go "eh?" every time you see the results?
That DC is gfe-eh...
Penalized pages are back to being pinned at the bottom of site: results, which is both helpful and annoying to see how stupid google can be.
Matt's forecast just makes me shake my head:
"It brought smarter Googlebot crawling, including tricks like full gzip support and a crawl caching proxy that means less bandwidth usage for site owners."
Seriously, the new googlebot crawling is pathetic, and nearly infinitely stupider, if only because maybe 1% of webmasters care about "less bandwidth" rather than "index my pages you inept search engine".
Sadly, Google really still seems awfully clueless about their key problems... which doesn't even mention their idiotic "index every blog and freehost page we can find" priority.
Forecast = continuing problems with no fix in sight this year
steveb, I had googlebot on my site consumming over 30 Gb of bandwidth about 3 years ago... My allowances were small so it did cost considerably higher.. My point is that ok it is small for most but I did contact google about it..
|Forecast = continuing problems with no fix in sight this year |
The world of search engines and the way they crawl/index the web has changed forever once SERPs PPC was introduced to the world. The BigDady of all slippery slops.
Forecast 2006-2007 = Polish and refine your PPC skills. You are going to need them if you want to survive this crap.
Reminder -- the topic is datacenter watch.
|Reminder -- the topic is datacenter watch. |
And on that note, I am just not seeing anything worth discussing.
Is anyone seeing any true movement?
Hmm, cache date is the 6th. Last piece of content seen in the cache is from the 5th.
The page ranks for the topic added on the 9th and the words of that topic are seen in the snippet.
Look at datacenter 64.233.183.* , i'm seeing the results described untop here. Old infrastructure is giving totaly different results for certain keywords then the new infrastructure.
I think that's the data we're being served in the UK.
Doesn't look too hot to me.
I thought the results from updated PR DCs was looking good the other day. I don't seem to be hitting these at all anymore.
I see today that site:domain.com counts are down 20% to 40% across the board.
The pages were previously normally listed, and now just do not appear in the SERPs at all.
It is NOT due to a change in Supplemental, or "phantom supplemental" URL reporting. Those are still there in the searches that I looked at.
OK. I have now worked out some of what that is all about.
You know how I keep on saying:
>> "Supplemental Results cannot harm you if they are for URLs that are 404, or are for URL that are redirects?" <<
Well scrap that. Now they can. And very badly.
Disclaimer: This is based on looking at only three sites that have lost 20 to 40% of their pages from the index in the last 48 hours (site:domain.com). The effect occured in all datacentre at the same time. This effect might be temporary. This might all be co-incidence. Yada. yada. yada.
You'll recall how I wrote just a few weeks ago about a site that was perfectly indexed.
The site had 160 pages and all were listed as www. Site has all the usual redirects and fixes in place for a long time. There are also 20 pages that recently changed their URL, when they were moved to a different folder.
So as of last week, there were 160 normal results in the index for a site search, and a further 20 Supplemental Results for the 404 pages, when uisng &filter=0 search. Omitting the &filter=0 parameter made about 20 normal pages hide behind the "click for omitted results link".
Suddenly, all of the NEW URLs for stuff that has moved are NOT indexed at all. So, the site:domain.com search lists 120 pages that have not moved, and 20 Supplemental Results that are for the 404 pages. The 20 new URLs for the stuff that moved are NOT shown in the site search any more.
Moving pages to new URLs has always been a bad idea. If Google is now going to stick with old URLs, and delist the new ones, then it has become a very very bad idea to move pages, unless there is a 301 redirect from the old to the new to capture the traffic.
On this site, the old URLs now serve a custom 404 page with basic site navigation within. That is going to get changed tout suite to individual 301 redirects for the 15 moved pages.
Again, this might only be a temporary effect in the SERPs, but changing the old URLs from returning a 404 to instead return a 301 will at least get visitors who see the old URLs flowing directly to the new URLs right now.
There are another 20 URLs affected on that site, and I need to look and see what happened to them, and why too. So, it isn't conclusive that it is just moving pages about that is the problem.
and what if we talk about thousands of URLs being changed to new ones?
I don't think that is the case here... at least I hope... site: results have been different every day for the past 2 weeks.
One day 70 pages the other day 140 pages, the third 40 pages, the fourth 19000 pages... today 960 pages... I am possibly looking through different data centres each time, but these types of differences don't make sense.
Only the first 20 or so results seem stabilised and proper...
I'd wait another week or until things are stabilised before jumping into conclusions... Unless you know something that we don't... ;)
I see very different SERPS on at least these two datacenters:
On these domains if you type a keyword they list almost only artikles about the keyword. All big sites (shops mostly) are a goner.
All other datacenters show the "old" SERPS.
Is google trying this way to inprove the clicks on the ads?
[126.96.36.199...] - 11.300.000 results
[188.8.131.52...] - 11.300.000 results
[184.108.40.206...] - 13.400.000 results
[edited by: Gerwin7 at 1:32 am (utc) on Oct. 15, 2006]
I see some older supplemental results (for URLs that have been 301 or 404 for a long time) cleaned up on that datacentre, but no other major changes as yet (gfe-nf).
"I see very different SERPS on at least these two datacenters:
I see the same thing on all 64.233.183.x range.
Wonder what it is?
All I can say is that G normally test some pretty crazy stuff during
"People watching US Football" hours.
So i wouldn't put too much stock in anything you see for the next few hours.
When the results return 25,270,000,000, instead of a more manageable 27,000,000 for a keyword, then I know the SERPS prolly aren't too reliable. ;)
|"I see very different SERPS on at least these two datacenters: |
I see the same thing on all 64.233.183.x range.
Wonder what it is?
The only thing I can say at the moment is some of my pages which lost ranking for several months ago on competitive keyphrases are back (not as good as they were, but back in fair positions) on that set of DCs. Maybe its a testing DCs set.
does anyone have anything to add about these dc's? I am in the top 10 for my keywords across all dc's. I am > 100 on these with the homepage not even showing up :( God I hope this is just a testing dc or something --- or I am ruined.
I think it can almost only be a testset. For a local search in NL a, pretty new 2 word keyword phrase google is ginving me 9.270.000 hits. If i do the same search on these dc's i'm getting only 520.000 hits.
With old PR data, not so frequently updated datasets etc. I think and surely hope this is just old infrastructure. The only problem there is, is that a lot of people connect to these dc's and get also some strange results, as a lot of not so relevant sites appear on top.
What do you think of DC 220.127.116.11 Reseller? Are those old results?
Using my testing keywords within the sector I watch, I see 18.104.22.168 display results which are similar to few other DCs. However, what use to be site #1 isn't there anymore, not only on 22.214.171.124 but on several other DCs too.
Call it minor re-shuffling, maybe ;-)
| This 175 message thread spans 6 pages: < < 175 ( 1 2 3  5 6 ) > > |