homepage Welcome to WebmasterWorld Guest from 54.205.254.108
register, free tools, login, search, pro membership, help, library, announcements, recent posts, open posts,
Become a Pro Member
Home / Forums Index / Google / Google News Archive
Forum Library, Charter, Moderator: open

Google News Archive Forum

This 294 message thread spans 10 pages: < < 294 ( 1 2 3 4 5 6 7 8 [9] 10 > >     
WWW2 & 3
different results.
needinfo




msg:209450
 10:10 am on Mar 6, 2004 (gmt 0)

WWW2 & 3 are showing big changes in the SERPs to WWW as of this morning in the UK.

Anybody else noticing anything,

 

canuck




msg:209690
 8:25 pm on Mar 11, 2004 (gmt 0)

One big reason Amazon is where it is in the SERPS is due to their links from the Affiliate programs that they have.

On several searches I performed this morning and saw Amazon in the top-10 I did a back-link check on those pages, guess what 25-100 Affiliates for each page, almost no natural links.

My beef is that this is not a natural vote for Amazon since there is a monetary reward for the Affiliates to place these links. If G is so keen to discount purchased links that some here do & other artificial linking schemes, then why wouldn't G also want to discount Affiliate links?

SyntheticUpper




msg:209691
 9:39 pm on Mar 11, 2004 (gmt 0)

Funny thing, do an allinurl:amazon and you get

...wait for it...

181 million results!

O.K., there's also a little river called Amazon, and those strange ladies who did a reverse Pammy Anderson. But even allinurl:amazon.com comes up with 12.5 million pages - and then they've got all those regional URLs which this figure excludes.

zomega42




msg:209692
 10:45 pm on Mar 11, 2004 (gmt 0)

Here is some wild speculation about what G means by "constant updates."

Why tweak the algo several times a month? How are they deciding which algo works best after they tweak it? Probably not just by surfing thousands of test terms looking for spamminess. If they had some way to evaluate "how good is the algo" just by looking at the results, then they would not bother testing different algos on a live server. They would use their testing centers, find the best algo, and stick it online.

So here's my guess. G is no longer trying to guess the best algo, they are planning to resort to an automatically evolving algo, or natural selection of sorts. Ie, they try different parameters (maybe even randomly) in their algos, and only the strong algos survive. G tracks us when we click on their search results, and the longer we stay on a page before returning to look through more results, the stronger the algo. This is why G needs to test various algos on the public, because it is the only way they can get enough data to determine the strength of an algo.

The key behind natural selection is that it needs random variation. So G will keep tweaking the algo, and gradually, the most relevant sites are guaranteed to rise to the top under every search term, where "most relevant" is determined by how long people stay on the site after clicking.

We give ourselves too much credit thinking G would put in the huge effort to tweak the algo every week just to stay ahead of SEO. One tweak per month or even 2-3 months would be enough for that. They must have deeper reasons, this is just one idea.

It might be crazy but the right answer is probably just as crazy. And in the end G would be the crazy one not to use all the data they get from their click tracking script.

Any other ideas on why they would bother tweaking so often?

steveb




msg:209693
 12:13 am on Mar 12, 2004 (gmt 0)

"It didn't use to be that way."

But it has been completely expected. Last year Google gradually got much better at crawling long urls. A DIRECT result is Amazon doing better. It's an authoritative site with high quality content and millions of pages.

I think some people can't see the obvious future. There are only ten sites, sometimes only five domains, in the top ten. Megadomains nearly inevitably will take those listings a decade from now. Imagine if Amazon had the capability to seo all those pages individually... make great anchor text links, make thematic link webs, make the urls more friendly like site.com/keyword/book-title/ and so on. How could anybody compete against that? We aren't at that point yet, but Amazon titles it pages well, and has simply tons of anchor text links, so it is going to do good in the serps, and it should from a seo perspective.

==

Improved crawling of long urls is like instant "fresh" ranking of new domains... Google thinks these abilities are good, and they are not. They lead to weaker search results.

SyntheticUpper




msg:209694
 12:36 am on Mar 12, 2004 (gmt 0)

What do you mean "high quality content"? Absurd. They're essentially a bookshop - but what efforts have they made to explain / describe their content? Sorry Steve - but I can't see your point at all.

idoc




msg:209695
 1:00 am on Mar 12, 2004 (gmt 0)

"They're essentially a bookshop"

exactly... and I would venture that apart from the category pages and the index it is all completey auto-generated from a database without any human intervention at all. If I did that everyone here would call it spam... of course I don't have the adwords budget they do and don't have thousands of affiliates linking to me. I had thought that millions of affiliate type duplicate pages was one of the things google would want to filter to start with. I'm with the some of you that don't get it. Why amaz* and not barn*... they are essentially the same.... except that barn* doesn't do the ads.

[edited by: idoc at 1:06 am (utc) on Mar. 12, 2004]

IITian




msg:209696
 1:00 am on Mar 12, 2004 (gmt 0)

Imagine if Amazon had the capability to seo all those pages individually... make great anchor text links, make thematic link webs, make the urls more friendly like site.com/keyword/book-title/ and so on. How could anybody compete against that? We aren't at that point yet, but Amazon titles it pages well, and has simply tons of anchor text links, so it is going to do good in the serps, and it should from a seo perspective.

Does this mean it can be ahead of NASA's site just because it has a few books and a few book reviews related to NASA?

Let's say I type in "buy plane ticket tokyo." If Amazon had 1000 books on buying, out of which 100 are on buying plane tickets and 10000 books are on Tokyo, 1000000 books on Japan, 100000000000 books on Earth, should it come ahead of a poor discount ticket seller who is selling tickets only to Tokyo and has only 5 unSEOed pages on his site?

Stefan




msg:209697
 1:04 am on Mar 12, 2004 (gmt 0)

They're essentially a bookshop

Yep. They really should just show up for searches of "online bookstores" etc. The rest is spam.

They're like one of those big-box stores that roll into a town and send every local shop into bankruptcy. It's the Walmart of the internet. They sit there in all their fatness on top of their big heap of absconded business and serps. Babylon system.

(Sorry about the foregoing rant... I hate the thing... ;-)

newwebster




msg:209698
 1:24 am on Mar 12, 2004 (gmt 0)

www2 and www3 seemed to have now changed. Anyone else notice this?

steveb




msg:209699
 1:58 am on Mar 12, 2004 (gmt 0)

So widget sellers think their widget selling sites should show up before a large widget selling site? Interesting worldview.

Calling Amazon a bookshop is a bit of idiocy, if only because they sell cameras, dvds, clothes, tons of products. They also have reviews and everything else widget sellers should have.

If someone types in a query, a listing for a book devoted to that query is a fine result. It likely isn't the best result, or a top ten result, but if you type in travel planning or rollerskating or whatever, getting as a result an excellent book on the topic is successful search engineering.

Amazon is a powerhouse website in terms of seo, and whining about that fact won't change it. All the things that search engines value in determinging a quality website are there with Amazon.

IITian




msg:209700
 2:05 am on Mar 12, 2004 (gmt 0)

Calling Amazon a bookshop is a bit of idiocy, if only because they sell cameras, dvds, clothes, tons of products. They also have reviews and everything else widget sellers should have.

I am doing a search for a few popular car models (without the year) and Amazon shows up in top ten with their repair book on very old models (90-98) in many of them. As it has been pointed out, when people are interested in buying a book for repairing car, they will go straight to Amazon or some other bookseller.

Of course, there is a place for that book in the serps. Let's say, reserve the 5th page of the serps for Amazon, epinion, dealtime, ebay etc and let them fight it out there. Leave the real serps for real sites.

steveb




msg:209701
 2:06 am on Mar 12, 2004 (gmt 0)

The stuff that was on 216.239.57.* is now showing for me on www, but is no longer on 216.239.57.* while the junk that has been on www for the past couple weeks is now on www2 and www3.

Stefan




msg:209702
 2:18 am on Mar 12, 2004 (gmt 0)

All the things that search engines value in determinging a quality website are there with Amazon.

True. What if that changed though? I can imagine a new SE that filters out all the directory/affiliate/Amazon stuff and just leaves the source/niche sites. It could be very cool.

To drift back towards the topic: I'm just watching our most popular 15 or so serps. They've been wandering slightly up and down, no great changes at all lately. We're getting freshtags constantly, though, on many pages, and I had 2 pages that went up on Mar 8 in the serps on Mar 9 with freshtags. I was impressed. Whatever G is doing right now, I like it.

steveb




msg:209703
 2:22 am on Mar 12, 2004 (gmt 0)

Hmmm

off topic blog page

PR4

with one backlink

moves to #6 (along with seven other very weird results in top 10) for a hyper competitive terms because...

one of the blog contributers signature name is the query

Now THAT is weird. No anchor text, no PR, no theme relevance, nothing in the page title... hmmmmmm

Just this one specific term seems to have gone completely haywire.

Putting my black helicopter hat on... Google had a data failure when they tried to move the 64 data over. This explains those weird hyper fluctations we saw briefly, also explains the return of anchor text results, the lack of PR update, and the relative stability of a weak batch of results for more than two weeks.

They were about to hit a homerun and tore a hamstring.

[edited by: steveb at 3:14 am (utc) on Mar. 12, 2004]

steveb




msg:209704
 3:11 am on Mar 12, 2004 (gmt 0)

The duplicate site bug is back. Same url appearing more than once in the results.

====

Looks a lot better to me: [66.102.9.99...]

[edited by: steveb at 3:36 am (utc) on Mar. 12, 2004]

trimmer80




msg:209705
 3:29 am on Mar 12, 2004 (gmt 0)

omg
google has not pulled a hamstring. Its ripped a leg off. I just did a search for a competitive term and number 1 is a site with pr3 and only has the term in the title tag and about 4 other words on that page.

Quite comical

Spine




msg:209706
 4:28 am on Mar 12, 2004 (gmt 0)

Agree that the SERPs on [66.102.9.99...] look cleaner than what I've seen on the other datacenters since Sunday or so.

ILuvSrchEngines




msg:209707
 4:42 am on Mar 12, 2004 (gmt 0)

Thew new Google filter is a total joke. Worthless results. I expect Google will be out of business in about a year.

My logs are already showing significant increase in Yahoo based searches.

The Google officers better have their golden parachutes packed because they are going down.

Krapulator




msg:209708
 4:46 am on Mar 12, 2004 (gmt 0)

You don't really luv search engines do you?

idoc




msg:209709
 4:49 am on Mar 12, 2004 (gmt 0)

I've seen some pages without any text except a one line kw link phrase... all the rest of the page gif's with every image the same paragraph of spammy alt text at #1 for a $12 overture term. Also many plain text pages do rank as granpops said... only you don't usually see them with active scripting enabled... just the redirects or iframes which are nothing like the ranking page. Then the affiliate spam sites of which amaz* is king. They do have alot of products listed at the site... most of it is not their area...books, but they use the traffic generated from their site to take orders for other companies stuff... which makes them no greater than the other "internet affiliate" "order taking" "spam generating" sites to me and no, I don't think they should necessarily rank higher than the actual other companies whose stuff they are selling.

Besting the seo's has taken priority over the end product at G... though G will say the two are one and the same... I would concur they could become the same, but they aren't the same now.

ILuvSrchEngines




msg:209710
 5:05 am on Mar 12, 2004 (gmt 0)

I do luv search engines, just not search engines that don't work any more. It amazes me that so many here are wasting so much time trying to figure out something that can not be figured out.

What we have witnessed is the death of Google. We just need to get through the common stages of grieving.

1. Denial, shock and Isolation
2. Anger
3. Bargaining
4. Guilt
5. Depression
6. Acceptance
7. Hope

I went through all the common phases already. Believe me, there is hope for those still in pain over Google’s death but not without acceptance. I am now past the acceptance phase and now I hope they go bankrupt. :-)

caveman




msg:209711
 5:24 am on Mar 12, 2004 (gmt 0)


Putting my black helicopter hat on... Google had a data failure when they tried to move the 64 data over. This explains those weird hyper fluctations we saw briefly, also explains the return of anchor text results, the lack of PR update, and the relative stability of a weak batch of results for more than two weeks.

They were about to hit a homerun and tore a hamstring.

Don't know if you're kidding, or guessing in a sorta serious way, but that sure would explain a lot of the quirky SERP's and the odd way that they've evolved since shortly after those nice, sharp early Brandy SERP's vanished.

Either that, or they intentionally trashed what everyone else applauded a few weeks back...and they actually think these SERP's are...good? :-¦

Sure wish I was one to buy books on the Web more often. Then I'd *love* these SERP's. ;-)

IITian




msg:209712
 6:00 am on Mar 12, 2004 (gmt 0)

zomega42
So here's my guess. G is no longer trying to guess the best algo, they are planning to resort to an automatically evolving algo, or natural selection of sorts. Ie, they try different parameters (maybe even randomly) in their algos, and only the strong algos survive. G tracks us when we click on their search results, and the longer we stay on a page before returning to look through more results, the stronger the algo. This is why G needs to test various algos on the public, because it is the only way they can get enough data to determine the strength of an algo.

I too thought similarly except that I thought the only factor that counts is the clicks. If enough people ignored Amazon it would be relegated to lower rankings.

Anyway, I have one more widely speculative theory. Google hired a few hotshot Ph.D.s from MIT and Stanford, all of them with very big egos. They couldn't come to an agreement on a common algo and therefore each of them has been given free hand in devising their own and the percentage of time it will be used for searches will be proportional to number of stock options granted to them.

steveb




msg:209713
 7:48 am on Mar 12, 2004 (gmt 0)

"Believe me..."

No thanks.

==

"or guessing in a sorta serious way"

Yes.

steveb




msg:209714
 8:38 am on Mar 12, 2004 (gmt 0)

GG, since you are hovering around tonight... is there a way to send you feedback about this not-an-update thing going on? I just have one serp that is completely strange. Everything else makes basically consistent sense, but the results for one term are bizarrely weird.

ILuvSrchEngines




msg:209715
 1:54 pm on Mar 12, 2004 (gmt 0)

I don't think you will be seeing GG around much any more. People are not buying what he is selling, or maybe he is busy getting his resume up to snuff.

SyntheticUpper




msg:209716
 2:22 pm on Mar 12, 2004 (gmt 0)

Regarding "Amazon is just a bookshop" - I'm pleased my post querying its relevancy generated some debate. You see, Am#zon has the illusion of relevancy - but the relevancy is illusory in the sense that whilst the content of *the books* is without question, you have to buy them. The actual information content of the site is questionable. There's actually very little, and what is there is written by the customers themselves.

I have a lot of respect for Am#zon - I spent thousands of pounds with them in order to blag my way through my last degree :) - they were able to obtain specialist books and deliver them very rapidly.

But, they shouldn't be at the top of the serps in virtually every subject area - they're a bookshop.

<note: hash inserted in Am#zon inserted in case this looks like an advert for them - they don't need any marketing help as far as I can see :)

<note2: they don't rank this highly on other SEs do they - I wonder why ;)

Jon_King




msg:209717
 2:31 pm on Mar 12, 2004 (gmt 0)

>>People are not buying what he is selling, or maybe he is busy getting his resume up to snuff.

I hope he stickies me with that resume :) :)

Dayo_UK




msg:209718
 2:32 pm on Mar 12, 2004 (gmt 0)

>>>>>The actual information content of the site is questionable. There's actually very little, and what is there is written by the customers themselves.

Nothing wrong with having customer/user reviews and this is often very good and relevant content (to the books/product). Afterall where would forums (WW) be without user reviews/comments ;)

ILuvSrchEngines




msg:209719
 2:41 pm on Mar 12, 2004 (gmt 0)

Here is a real-world example that maybe people can understand. Let's say your elderly mom just called you and says she needs to see a doctor today or tomorrow because of problems breathing. You discuss with her that an ambulance is not needed but a doctor visit ASAP is needed.

Now, your mom lives in a different state that you. She does not know where to go. Your task is to ONLY use a search engine to find an ‘internal medicine’ doctor for her in 'Dallas Texas'.

After Florida update, what search engine would you use to help your poor sick mother?

1. Search engine #1 that does not have a ‘Florida’ filter
2. Search engine #2 that brings you back some great results for ‘internal medicine’ books from Amazon.com

Let's think about this for a moment. Hmmmm

Dayo_UK




msg:209720
 2:47 pm on Mar 12, 2004 (gmt 0)

Just tried that search and it look OK - if not excellent to me for the search terms I used.

This 294 message thread spans 10 pages: < < 294 ( 1 2 3 4 5 6 7 8 [9] 10 > >
Global Options:
 top home search open messages active posts  
 

Home / Forums Index / Google / Google News Archive
rss feed

All trademarks and copyrights held by respective owners. Member comments are owned by the poster.
Home ¦ Free Tools ¦ Terms of Service ¦ Privacy Policy ¦ Report Problem ¦ About ¦ Library ¦ Newsletter
WebmasterWorld is a Developer Shed Community owned by Jim Boykin.
© Webmaster World 1996-2014 all rights reserved