Google Updates and SERP Changes - April 2010

[wmxpert]

< continued from [webmasterworld.com...] >

We are in the same situation, since March 15, 2010 our traffic is continuously declining with the average of 6% every week. In current week it gets worst, approx. 15% decline!

Overall, 30% to 35% traffic has been reduced compare to peak week of January 2010.

"site:" is also showing huge drop in index pages, though it only provide the avg. not actual.

Don't know what's going on, Is it really the effect of Caffeine roll-out?

[edited by: tedster at 9:46 pm (utc) on Apr 1, 2010]

[macas]

OK , I'm fed up with this mad SERP changes , when this will be over ?Anyone ?
Watching my indexed pages bouncing and disappearing in every minute don't make me happy right now , it's driving me crazy !

I saw some GI ( Google Images) changes, I got new indexed pictures on GI search,a lots of them, actually. And here too I have crazy bouncing and disappearing ...

[TheMadScientist]

when this will be over ?Anyone ?

After the holidays...

I'm not sure which holidays and I've heard speculation they're the ones in 2015, but personally, I'm not holding my breath.

[gouri]

One of the characteristics of Caffeine I believe was supposed to be a smaller number of competitive sites showing up for a keyword search than what we currently see on Google.com.

This week, the number of competitive sites for keyword searches has increased on the Caffeine datacenter and on Google.com.

Could this mean that Google tried to rollout Caffeine and it did not work? Are what we seeing now sort of the middle phase of Caffeine and are the SERP going to stay this way or revert back to what we were seeing 2-3 months ago (on both the Caffeine datacenter and on Google.com) if infact the rollout did not go as planned?

[MrHard]

I don't get it, serps change daily so I don't see the point of a big serp change thread.

[walkman]

I don't get it, serps change daily so I don't see the point of a big serp change thread.

There are changes, real changes and then there's caffeine ;)

[brinked]

is it just me, or did google add more spacing between each serp listing?

[CainIV]

when this will be over ?Anyone ?

When will what be over? What you see is the new SERPs.

:)

[seolearner]

It is thursday and the google traffic is slowing down...and this usually continues until the weekend...I notice this pattern for the last few weeks (from Feb or march). Anyone else experiencing this?

I also have the same experience. My site is not so big but I always get less traffic and less sales in weekend. My site is up in caffeine and I think G* switch off caffeine in the weekend.

[Shaddows]

My site is up in caffeine and I think G* switch off caffeine in the weekend

Or, its the well documented time-variable SERPs, but more pronounced now than it was for the last 2 years (at least) that its been around.

[Kelowna]

My site is up in caffeine and I think G* switch off caffeine in the weekend

I don't know if it is for the weekend or not but I stopped seeing caffeine on all the USA proxies I check yesterday afternoon till now. They did let them run right through the last weekend though so they still can come back any time. I hope so...

[gouri]

Can someone please tell me what it means when someone says that they say are seeing filtered Caffeine results on the Caffeine datacenter and also when they are seeing unfiltered results?

How are you supposed to tell the difference between the two?

[Kelowna]

How are you supposed to tell the difference between the two?

I have been keeping a close eye on both caffeine and regular sets of results and do not see any type of special filtered results. Both sets have the flux going on so possibly people are seeing a slightly different set of results because of this.

I can see the caffeine results using a proxy in the east and then see just slightly different results sometimes using a proxy service somewhere else... this is just normal flux as bot sets are being constantly updated.

[curioustoddler]

We had stopped working on one site last year(May) and it gradually dropped in rankings( very slowly) but now it is appearing back on top !

[maximus12]

Kelowna,

I am seeing the same thing. I keep a close eye on caffeine and was seeing it all last week in the east coast, now I cannot seem to find it anymore? Can anyone confirm they are seeing caffeine on any PROXY in the US?

Maximus

[trakkerguy]

We had stopped working on one site last year(May) and it gradually dropped in rankings( very slowly) but now it is appearing back on top!

I also have noticed increased power of older domains/urls in caffeine. Can't tell if it's due to age of backlinks, or age of the urls. Any thoughts?

I've been revisiting and working on some older, semi-forgotten pages, and am testing to see if new pages on my very old, trusted domains have newfound strength.

[1script]

I also have noticed increased power of older domains/urls in caffeine. Can't tell if it's due to age of backlinks, or age of the urls. Any thoughts?

Most of my sites suffered a huge Google traffic loss one month ago. But the only two sites that have not been affected are both older than 8 years old.
Additionally, looking at the pages that Googlebot comes to visit, I see them trying to read pages that do not exist since 2008 and have since been 301-ed to new URLs (on the same site)

[crobb305]

I am observing .edu pages ranking higher with increasing prevalence in the serps. Unfortunately, spammy commercial pages are being placed on .edu domains and are ranking very well. I won't discuss the niche, but you can imagine it is a lucrative one. I have been emailing the Deans at some of these universities to grab their attention, and in some cases I have been successful in getting the pages removed. But Google should realize that .edu can be just as easily manipulated for commercial purposes as any other TDL, and .edu pages shouldn't be held in such high regard (in serps or for "link juice")

[JoeSinkwitz]

Same with .gov crobb

[internetheaven]

I am observing .edu pages ranking higher with increasing prevalence in the serps.

It may have nothing to do with the tld, as people mentioned before, old domains with old backlinks are treated very well. There aren't many better examples of old sites with old links than Uni sites!

[crobb305]

But Google should realize that .edu can be just as easily manipulated for commercial purposes as any other TDL, and .edu pages shouldn't be held in such high regard

Sorry, I said "TDL" and I meant "TLD", just wanted to clarify for newcomers, TLD = Top-level domain. This is just the perfectionist in me. I am sure it's trivial, but I wanted to state explicitly so someone isn't wondering what the heck a TDL is. :)

[tedster]

Google absolutely knows the .edu pages can be manipulated. They also know that in recent times edus have been hacked, in mass quantities, for parasite hosting. Students and faculty here and there have also been paid off to host splogs that get folded into link pyramids and so on.

Matt Cutts has repeatedly said that the Google algo does not look at the TLD as a factor. Google's current metrics for trust, authority, relevance, etc make that approach look very "90s".

[indyank]

Additionally, looking at the pages that Googlebot comes to visit, I see them trying to read pages that do not exist since 2008 and have since been 301-ed to new URLs (on the same site)

This is exactly what i too notice.Googlebot is looking for some URLs that have been 301ed long time back.They seem to be playing around with too many datasets, some of which are too older.

[curioustoddler]

I may sound funny and stupid but i have a strong feeling that google is behaving like someone who was out of the world for the last eight months or more and now trying to put together those eight months from different pieces.
If this is really true, did this also happen during bigdaddy?

[tedster]

Googlebot is looking for some URLs that have been 301ed long time back

That kind of checking does go on periodically - otherwise a website could change what they do with the URL and it still would be accounted for in the historical way instead of being current.

If it is happening more than usual right now, then it does sound like some kind of a "catch up" happening to bring a data-set more into present time.

[curioustoddler]

Age of backlinks and TLD has always been a factor. But now i think google is not confident of data they have for the last 8 months or so. They have it in different datasets. They have always kept things with a timestamp, whether it is a TLD, a page or a backlink. But this time this is all messed up. So they are showing more confidence in what they had 8 months ago.
It may continue while they sort this out by holidays!

[tedster]

To add some detail to the idea of different data sets:

We've developed an interesting trick that speeds up the first step: instead of storing the entire index on one very powerful computer, Google uses hundreds of computers to do the job. Because the task is divided among many machines, the answer can be found much faster.

To illustrate, let's suppose an index for a book was 30 pages long. If one person had to search for several pieces of information in the index, it would take at least several seconds for each search. But what if you gave each page of the index to a different person? Thirty people could search their portions of the index much more quickly than one person could search the entire index alone. Similarly, Google splits its data between many machines to find matching documents faster.

[google.com...]
One query can use in the area of 500 different servers - and there are thousands of servers that hold various versions of that data. So yes, you can easily be seeing the results of different data sets a lot of time.

And with regard to the time stamp, yes, the data from every url is stored at various stages of its evolution and each one has a time stamp. Not all data-sets contain every time-stamped version of a given URL.

Watching Google move data around is a pretty intensive sport - it can either confuse things or clarify things for you, depending on how accurately you picture what they're doing.

[aristotle]

Has Google ever said how long they keep copies of old versions of pages? Are they permanently deleted after a certain time period, or are they kept indefinitely like the Wayback Machine?
Does anyone know?
Thank you

[tedster]

AFAIK, Google does not "permanently" delete any crawl data - although it may eventually end up only in a backup somewhere. They certainly seem to keep old URLs indefinitely, and once in a while, googlebot comes looking for URLs that were gone years ago.

[curioustoddler]

THey may keep old URLs but how many copies of the same URLs they keep ? Suppose one topnews.php keeps changing after a regular internal, then how many copies of this same page will remain with them? And how do they use this information about different versiions of the same page.

[walkman]

I see older serps now. maybe just less caffeine DCs live?