Why are these "data refreshes" being called "updates"?.
I don't see any algo change within Google search.

You're absolutely correct. I just edited the thread's title to say "changes" instead of "updates".

Well mainly because GG said it would be an "update".

Now, if you ask me (and you didn't), I will again espouse the concept that BD "fixed" the primary causes for radical algo-based "updates" in the first place (lack of data storage).

soooooo......

you will see "refreshes" every month.
and from reading between the lines, GG estimates these "refreshes" to become more weekly-type events.

To be honest, this might be the last time you hear the word "update" officially from a Google Employee's mouth (err. post)

What if you run a unique content site on a CMS with fixed descriptions across the site? Am I doomed?

While this isn't optimal, I don't think that this alone would cause problems. There would also need to be other factors that need to be addressed.

I am seeing big changes with one of my sites. Some search terms that used to return an inner page are returning the homepage. This looks like a bad update for us so far.

I'm seeing the pagerank on my inner pages slowly climbing on an increasing number of datacentres from 1, erm, all the way to 2. Better than a poke in the eye. (The poke in the eye was a few months ago, when they dropped from 3 to 2 to 1,) I'm even seeing 4 or 5 google search referrals a day, for the first time since I launched the site back in Feb. Heady days indeed!

What I am seeing is that we had thousands of unindexed pages with cache dates dating back to 1995. Now they are completely gone but the related subpages (eg. technical details / specifications) are still there but supplemental.

In fact if I am not completely wrong: we have app. 200'000 pages and 198'000 are supplemental, missing or dead.

Thank you Google...

btw:

After some checking I tend to say that is one of the worst updates that ever striked me. And I have seen some ugly ones before...

Thanks

itloc

So, what's the difference between sites that are experiencing major changes (for better or for worse) and those that are seeing few if any changes for their traffic and keywords?

I'm one of the "What changes?" crew, and my site's characteristics are:

- "Content site"

- About 5,000 pages, all of them HTML "flat files" with plain-English filenames

- Editorially diverse (One "umbrella topic," with several major subtopics and lots of minor subtopics that fit under the umbrella theme)

- No SEO other than what's recommended in the Google Webmaster guidelines (i.e., "spider food" such as descriptive titles and headlines)

- A decent number (not a massive number) of inbound links, including some from authority sites such as magazines, academic libraries, etc., but very few reciprocal links

- Five years at its current domain in October

Over the years, I've seen very few changes during Google updates except for a 33% boost in Google referrals immediately after the Allegra update and, later, a two-month drought in spring, 2005 (apparently because of the www vs non-www duplicate-content issue) that was resolved after I implemented 301 redirects in my .htaccess file.

How many other members have sites that fit this pattern? I know there are some of you who haven't experienced dramatic ups and downs. What's the secret to stability: Being "organic"? Taking an old-school KISS approach to Web design and content management, rather than being on the leading edge? Having a good balance of inbound links, rather than unnaturally large numbers of reciprocal or purchased links? Would anyone care to hazard a guess?

[edited by: europeforvisitors at 3:01 pm (utc) on Sep. 19, 2006]

Since i'm not doing any major changes to my site the last two weeks regardless of these ups and downs... i feel that i'm monitoring whatever is happening with Google.

It started before sep 15th in my opinion, i have been seeing my BLs updated on and off every half a day or so from around the 12th. Before this, that test datacenter was the only one with this information. Now it's on Google, everywhere. The data is NOT presented the same though. It's similar but it's not the same, perhaps it's just filtered in a different way. The inbound links it shows are the same.

Traffic didn't go up. Nor did it go down, there's just no place for it to go any lower, it's still the 'PR0 for no good reason' effect . It's a brand new photo site btw, for those who didn't read my whining before. Was a decided PR0 from the day it was uploaded... and wasn't re-calculated for more than three months now... or perhaps it WAS and still is evil in Google's eyes. Whatever, that feeling's mutual now.

But the point is... PR wasn't affected with this change.
BLs and some results were.
How about your sites... were all aspects affected or just BLs and position?

The OTHER point... regarding the keyword in the anchor text link idea. Although you mentioned site navigational links, and this might just be a coincidence. Like google itself is just a bunch of coincidences now.

I used to type the domain without the extension into google ( and yahoo and msn ) to see how it's doing. The domain was chosen because this word was non-existant on the web before anyway. Easy to look up any feedback about the site ( we like feedback, especially if we didn't ask for it. )

Many links we exchanged or got from kind people who are of course all managing sites either about travel, photos or travel-photos :)... have the text of our domain name ( with or without the www or the .com ) in the link that point to us... for that's the site name.

This i noticed today, don't know if it's to stay this way:

I found my homepage, i mean the index page gone from the results when entering this word... heh... yeah, i was wondering how it could get any better. No change in the traffic though it's just frightening in general.

The help page does come up... with another result, the about page, which has been merged into it and GONE for a month now btw, but somehow is still in the cache, even though i have all the tags to tell Google that i don't want it to be there.

By the way it's No.7 on the SERP... No.1 is webmasterworld.com ... :)
( Then a photo directory, 2 results, then flickr, 2 results... then our site )

If i type in the entire domain name with the www...
Our site is second in the results... but the 1st result is an RSS feed from another site we posted some photos to. Oohkay...

So i typed in the domain with the extension .com and without the www... nope still 3rd only. A photo-site-directory comes up with an indirectly scripted link to us.

Now...

I know that if i type it in as text it will be treated as such, and sites with higher PR, higher authority that feature our site's name as TEXT will come up first.

But the point is... that the site was no.1 for all three phrases until yesterday! Three months straight this WAS the keyword we ranked no.1 with... sad but... funny at the same time, the domain that's a non-existant word we dominated with :) And we ranked no.1 with the home page as the result. Now that page is gone from the results if i type in the keyword, and is pushed off if i type in the domain in any other way than the complete URL ( you know... with the http:// :P ).

The only thing that has been changing the last two weeks is that there are more and more links pointing to the home page with the keyword ( the domain and site name ) in the anchor text. The site is the same, cache is from the 11th too.

Might be a coincidence though. I WANT it to be a conincidence.

...gah... i just tried another thing.
This is a single word, remember? I mean the domain name.
If i put that single word in quotes the site comes up first, ahead of webmasterworld.com :X

wow... i didn't know i get different results if i put a single word in quotes... ( okay i tried this trick now with other words... but nope... the number of pages are different ( not 1.170.000 but 1.160.000 ... ) but the SERPs are an exact match. )

We're special! Yay. :/

Interesting engine... this google thing.

I don't know. I have ten sites of my own that I watch, plus another 12 or so that I keep a really close eye on for clients, and I just went through everything this morning, and if anything, everyone's UP in traffic since Friday. Most of them are business sites that slow down on the weekends, so it's really hard to say for sure yet, but they all had a pretty good Monday - much better than usual. Even the one site that got hit hard in March is doing well with this one.

None of these sites has many links, not even the ones that get 40k visits per day. They tend to be catalog driven, and there's no reason (and really no place) for them to linked much of anywhere - the only thing that seems to exist for such sites is spammy B2B directories.

All of them have text links (see other forum post).

My three new sites still are only showing the home pages, but all the other domains range in age from 1994 to 2003, so they've all been around awhile.

tedster wrote: "So, if your site is about widgets, then I think having the actual text "widgets" in too many of your navigation links could cause a problem."

That's exactly what I was referring to. My store page navigation has category titles such as "Pink Widgets," and too many of the links under that category were something like "Square Pink Widgets," "Round Pink Widgets," etc -- "pink" and "widgets" being superfluous, so I probably was guilty of keyword spamming.

tedster also wrote: "For another thing, Google recognizes the page template and to a degree can "see" the navigation as contrasted to the body copy. The repeating parts of the template have a distinctive footprint, in other words. If you start varying that anchor text from page to page, it just "might" look like you're trying to play a game with Google."

I'm not varying anchor text for the same navigation links on different pages, but I do have different navigation systems for different parts of the store (different templates for each area), so you can navigate each part of the store easily. But that would be the norm for a store, I guess.

I also recently have started adding more than one "Back" and "Next" navigation system to each page. One set of "back" and "next" links takes you through the whole store, and another is just for that particular section of the store, or yet another for a subsection you might be interested in.

I hope this couldn't be causing some kind of SEO problem, since I think it really helps with store navigation.

I'm also wondering -- is it just possible to have too many navigation links per page? I have a group of more generalized links at the top of each page to help move you through different sections of the store, while the ones in the sidebar are mainly more specific ones for that category of the store. In addition though, at the top of many pages I also may add a few links for more narrow areas of interest -- such as "curly pink widgets for the home," "curly pink widgets for the garden", "curly pink widgets for your car," etc.

I'll also have these same links in the sidebar navigation. It sounds like a lot, but none of this was meant for keyword spamming purposes, just ease of navigation -- so you can get from here to there easily without having to scroll up or down.

Of course, now I'm wondering if I need to remove some of these links.

Just thought I'd share my experience...

My pages dropped from middle of page one to somewhere between pages six and ten as a result of the June 27th update. I never did figure out why but I tidied up the code here and there and resolved a few minor issues.

Lo and behold on August 17th they returned, better than ever, to the top of page one, where they've stayed ever since. Until September 15th that is.

So now they are still on page one, just loitering around the bottom. Weird thing is that traffic to these pages from Google is up by 15% since positions slipped. I've had my best days ever. So for me, this update has been bittersweet. I can cope with that..just about.

Since i'm not doing any major changes to my site the last two weeks regardless of these ups and downs... i feel that i'm monitoring whatever is happening with Google.

It started before sep 15th in my opinion, i have been seeing my BLs updated on and off every half a day or so from around the 12th. Before this, that test datacenter was the only one with this information. Now it's on Google, everywhere. The data is NOT presented the same though. It's similar but it's not the same, perhaps it's just filtered in a different way. The inbound links it shows are the same.

There was a major shakeup on Sept 15th in Google's SERPs. This was one of several that happened this summer. It is like one really massive rolling update with several phases. Thus far the major update dates for this summer have been June 27th, July 27th, Aug 17th and Sept 15th. Different sites have been hit by each update. One hallmark of all of these updates has been that when a site gets hit it totally disapears from the top of Google SERPs for all key phrases. On subsequent updates affected sites may see a partial or complete recovery.

My site for instance was not affected by the June 27th update, but it totally disappeared on the July 27th update. On the Aug 17th update it recovered for a few of its most core search phrases and then it completely recovered on Sept. 15th.

It is like Google is doing a major overhaul of its dataset one section at a time. The question is how much longer are these "rolling blackouts" going to take place?

EFV wrote: How many other members have sites that fit this pattern? I know there are some of you who haven't experienced dramatic ups and downs.

I have four sites that fit this pattern, i.e.

Single human-readable URL for every page

First on-line between 1998 to 2006

Mostly old fashion table based layout, combined with CSS

No reciprocal links

Semantic page structure, i.e. <h1> for the title, <h2> and <h3> for subsections

Unique and on-topic title per page

Unique and on-topic meta description per page

number of content pages per site between 60 and 1000

Strict non-www to www 301 redirect

These sites have a total of 4 (four) supplemental results in the SERPs, caused by a programming error by myself.

These sites float slowely to the top of the SERPs. They never jumped up or down during a Google update. The only movement during a Google update is when sites that previously artificially climbed to the top of the SERPs are removed, and my sites move up to the empty places.

Even the in June 2006 launched site (only 5 pages, no javascript, no ads, no images) is #1 in the SERPs for a competitive search term with only 4 backlinks and a toolbar PR3, outperforming PR5 pages of established sites.

I do however have other sites generated by rapidly available free CMS systems:

PHP query string type of URL or "Search-engine friendly" URL

First on-line between 2004-2006

Unique, but semi-identical titels due to auto generation by CMS system

Unique, but-semi identical meta descriptions due to auto generation

Strict non-www to www redirect

Result:
Many supplemental results (80% of URLs in site: search) through unavoidable duplicate content problems (without extensive rewriting of the CMS system). Very vulnerable for Google changes and in general poor rankings.

My site also conforms pretty closely to the criteria listed above.

I was just reading over the Google Webmaster guidelines, and I'm now concerned about this statement:

"Avoid "doorway" pages created just for search engines, or other "cookie cutter" approaches such as affiliate programs with little or no original content.

If your site participates in an affiliate program, make sure that your site adds value. Provide unique and relevant content that gives users a reason to visit your site first."

This worries me. I don't have doorway pages, and the Amazon store section of my site wasn't created just for search engines, but -- it really doesn't have any original content! What it does have is a particular theme, with items chosen to fit that theme.

The rest of my website, of course, is all original content, but my store is now much larger than the rest of my site.

Should I be concerned about this?

miki99, I have a site with something similar. My affiliate store is all under one directory so I just added that directory to the disallow section in robots.txt.

tflight, I hadn't thought of that. The trouble is though that a majority of the traffic to my site lately has been due to the store itself -- people doing searches and being brought to my store by the results. I'm really wondering if I should maybe cut my store loose, at least make it a separate subdomain or something.

Before the 15th, I was finding that pages I added to the store really seemed to HELP my site's SEO -- especially when I added pages of items closely related to my site's art niche.

I was just over at the official Amazon forum, and there seems to be a consensus there that Google has it in for affiliate sites.

I should add (if I haven't already) that my Amazon store is now much larger than the rest of my site.

Now I know G is losing touch with reality.

I just did a site:domain.com on a bunch of data centers

ON 64.233.167.104 It's got all my pages supplemental, bad enough, but the really strange thing is that there is a url in the results that is not on my site, not related to my site, not even in my language.

Anyone seen a stray url in their site:domain.com results before?

Alzheimers is setting in at the Google farm

[edited by: rden17 at 8:07 pm (utc) on Sep. 19, 2006]

Yep, happened to me too. I got a site url in Russian or something.

Yup same here. Over the top and messed up site: searches for my site. The first listing is a Czech site in that language. My index page is not listed. Every other page is sup.

Same with me, Chines and Russian pages. The worst yhing is that wit the :site command my home page is listed on page 5 with supplemental results

Yep, they've really blown it this time. I have lost all confidence that they'll ever get it right again.

I see 30 results that have nothing to do with my site or its topic, all in languages that I cannot identify. Then starting in the 31st position I see supplemental results from my site. A day or so ago I had to go 600+ results in to find the beginning of the non-supplemental results for my site. I imagine that it's now 600+ plus 30 to find them.

I've fixed all that I am able to fix on my side of the fence. I can do no more.

weird: I just chekc site:domain.com, and for the first time in a year, ALL my pages are indexed, however, about half are supplementals. All pages are very interlinked with each other, so I see no reason for the supps, but will try and see if a homepage link fixes that.

>> What if you run a unique content site on a CMS with fixed descriptions across the site? Am I doomed? <<

You aren't doomed, but you could do a lot better by having a unique title and a unique meta description on every page of the site. Matt Cutts has already mentioned this stuff very recently [threadwatch.org].

>> I also recently have started adding more than one "Back" and "Next" navigation system to each page. One set of "back" and "next" links takes you through the whole store, and another is just for that particular section of the store, or yet another for a subsection you might be interested in. <<

Check the URL that is in those links. If the URL has different parameters, or parameters in a different order to those in any other links that can get to that page, then you are creaating duplicate content issues by the bucketload.

See the stuff I wrote about "next" and "previous" links in vBulletin [webmasterworld.com] and other forums, and how those cause massive problems with Google indexing.

EFV wrote: How many other members have sites that fit this pattern?

My main site ( 1998) fits that pattern even down to the 2005 hit and the same resultant fix.

Have a few supplementary pages -duplicate conent but "user friendly" so they stay.

No real SEO , no major inbounds.

Currently steady Google traffic and steady repeat vistors.

No www.root.com/page.htm internal linking.All relative.

Steady page additions with good clean content,all page design changes are done in stages and don't seem to have a problem ( cross fingers here ) merging affiliate content.

Newer site major problems Jan/Feb 2006 -now seems to have settled down.

Conclusion: - write naturally - add useful content - use H1 and H2 as they are designed to be used - don't force feed the search engines!

OMG! Thanks g1smd. I don't even really understand how that qualifies as duplicate content. I'm off to read your posts.

europeforvisitors - thanks for that information. I have thinking lately - what is it that I can do to make sure this kind of stuff doesn't happen to my site in the future.

The main things I see different about my site and europeforvisitors is that I have way fewer pages and links and my site has only been around since early 2003.

I did some reciprocol links years ago, but nothing recent. I started a blog some time ago and have been getting links through that, but apparently not enough.

I did fine through big daddy until now.

Any supplementals that I had at one time have all been gone - except when I checked today - now I have tons of supplemental pages! What did I do to cause those to happen? The only thing I noticed when doing a site: search was that in many of my directories I didn't put in a index page and google is now listing the default server page that lists all the links (don't know what those pages are called, but I think people know what I mean). I guess I need to get rid of those real soon.

g1smd, have to admit I'm not even sure what url parameters are -- over my head, I'm afraid. I'm using simple HTML links, I don't have a forum or anything. Maybe I should just delete all my BACK and NEXT links?