Forum Moderators: open
Anybody else noticing anything,
---------------------
When G does what it's doing right now to the SERP's, it bugs us for two reasons (basically the same reason twice):
1 -- On our two lone e-commerce sites (used to be one), conversions drop significantly, as they have been doing since last Friday when G started showing variations of these SERP's we're seeing now.
2 -- Our many info oriented sites show much higher abandon rates off of the landing pages. In other words, people are finding us for searches that don't very well match page content, so after they get a look at the page, they split.
Happens when G fiddles with adding too much of that semantic sauce, I believe, turning back to the direction of Florida/Austin...
I'm also seeing lots more 4 word key-phrases in my logs again. Things seem just like Florida/Austin in many ways, I want my brandy back.
If they ever update their cache to more recent info that's when I expect to see some sort of change, hopefully for the better!
They crawl and crawl and crawl but never seem to update the cached versions of the pages.
At one point I noticed a few cached pages had changed but then they reverted back to the old cache. Usually this happens when they are about to update them all but this time it just bombed out.
Alexa use Google for their search and put Adsense on every page - Alexa are owned by Amazon - Google don't want to upset Amazon because of the large amounts of cash they are making from Alexa - therefore there will be no Amazon Filter.
For all those who are praising 'Saint Google' tell me that there is another reason why the perfect, non-corporate-greed Google would keep billions of Amazon listings clogging up the results?
Um, because it's a massive domain... with topical content pages on endless subjects... with tens of thousands of anchor text and inbound links from other thematic pages?
Two pages from Amazon likely should rank in the top 100 for almost anything. IMO, seldom should they be in the top ten, but any search engine algorithm must rank Amazon pages fairly well for a whole lot of things.
But in the top 5 - on many searches?
Nyet - this is an embarassment - and an algo failure.
I've said it before: we search the Internet to obtain (almost) instant information - not to be given the option to buy a book on the subject. If I want a book, I go to a bookshop -or very often am#zon.co.uk.
Their domination of the Google serps isn't doing Google, or the web searcher, any favours at all. It just leads to frustration.
It's very odd - like an irresistable force driving them towards "authority" over relevance. It's interesting to observe which will be deemed "authority" when you own a large amount of sites or which will make it through "the cut". The net effect for us has been around zero through this all but I can't help but find the troubling direction G has chosen disturbing.
It's very odd - like an irresistable force driving them towards "authority" over relevance
Nicely put - I think Google should open an academic bookstore - this is clearly what they really desire.
This will leave the rest of us to surf the web.
nope just leads them onto other engines MSN traffic building nicely thank you :)
Then we can all move on: stop our desperate tinkering with our perfectly clean sites, and see traffic from elsewhere.
So go ahead Google - throw us another wobbly algo, I'm keen to see you finished now.
[edited by: SyntheticUpper at 1:49 pm (utc) on Mar. 11, 2004]
surprise surprise what is happening now
steve
steveb, not sure I'm with you on this one. Well actually I am sure; I'm not. ;-)
The frequent littering of G SERP's with Amz listings represents a choice on G's part. It didn't use to be that way. It was not even that way with those early Brandy SERP's that many of us thought we so sharp.
Sure, there's logic to their appearance, for the reasons you note. But there's little use to them, since 99.9% of surfers know about Amz already. These listings clutter the current SERP's and push down results that in all likelihood are more useful for the vast majority of searchers, who are not using G to look for bookseller pages. I'd feel no different if Walmart pages started appearing all over the SERP's. (Don't laugh folks.)
It would not be difficult for G to de-emphasize these Amz pages, and similar ones.
I must agree with steve b on this one--it will be tough not to have supersites somewhat clutter the serps.
i feel that it is difficult to infer what google wants be looking at one corporate site success and link it to a higher motive.
despite their difficulties the last few months, google has shown me that are willing to take risks--wise or not-- to attempt to stay the most relevant search engine.
all speculation though--of course
... my reason? Oh, I don't have one .... tell you what I'll go look one up on Google ....
.... couldn't find one .... I did however buy a book on on 'Disagreeing' from Amazzon ... ;)
By them cranking 1 little mathematical change it could do what everyone is seeing. It doesn't take a lot of mathematics to change the entire board.
BTW the reason I posted is that I have looked carefully for all this variability you are all seeing and I see nothing. Very stable results for the past few days on the datacenters I checked and for www and www2/3. I'm working from a Canadian base. Could that have something to do with all this?
Well, Y has announced that they are working with Library of Congress to be given access to their database.
So with G, you have Amazon listings scattered in the SERPs. With Y, you'll soon have .gov and library books littering the SERPs. What next, ebay auction listings (oh yeah, G already has ebay store listings)
I agree with it in principal but I don't think it can apply to search engines. Google HAS to keep switching the algorithm frequently even if it means bad results for a while. If they keep switching between different algorithms and constatly change the ranking factors then spammers will have to keep changing their spammy techniques.
I think Google is playing with webmasters rather than search results. If you've noticed, the number of 'What can I do to get ranked well in Google?' queries have dropped considerably since Florida. All we have now is speculation on what works because Google changes so frequently.
I expect that this will continue for another couple of months before Google will settle down and by that time, hopefully spammers will have turned their attention to Yahoo.
Does this 'conspiracy' hold weight do you think?
G must have been given access to the Amazon database to crawl and be listed in the SERPs.
Most likely the case. Especially when you consider the complexity of the dynamic URLs that Amazon generates.
Um, because it's a massive domain... with topical content pages on endless subjects... with tens of thousands of anchor text and inbound links from other thematic pages?
Err, ever check the backlinks for many of Amazon's interior pages? Even the ones close to index page? Give it a try. (hint: no backlinks = no anchor text = no other thematic pages). Amazon is just a massive domain Google has partnered with.
seasalt
Over the past couple of weeks I have seen stable, but different sets of SERPs that seem to reflect one that weights large sites with good internal link structure more highly (with good on topic pages, with Amazon near the top), and another that emphasizes authority sites (smaller, but with more incoming links from on topic sites, with Amazon down in the 10 or 20 position).It might be the case that Google sticks with 2 or 3 different, but stable algos, but alternate which results they post (like each data center having a specific algo -- or something).
I see the same thing, I have been speculating to myself that Google is working with algos in paralell and the SERP fluctuations we're seeing is from tweaking during the process of merging those results.
Maybe trying to strike the proper balance between commerce and info? In any event, I've been making good use of that 'Help us improve' link.
I'm in the UK (for server reference) and after 7 weeks Google has finally freshened the data on my main page. (This is odd as it is normally updated every couple of days).
Stranger still is that although fresh tags seem to be appearing next to results again (9th and 10th of March) they are not appearing next to many sites that always showed fresh dates. Even most of Google's buddies at stanf#rd.edu's pages are not showing fresh tags.
The thing is though, there are now more fresh tags showing that over the past month so what do you guys think has been going on? I've already stated my theory of a new spider, anyone concur?
However, both sets of serps look all right to me. They're clearly emphasizing different things, but I don't see a lot of spam in either one. It's just weird that it keeps fluctuating like that. Maybe they're going to blend the two together eventually, or maybe they've just decided it gives even exposure to all of the top-ranking websites to alternate them. (-:
And as for Amazon, I think it would be impossible for Google to get rid of those results without manually penalizing the amazon.com domain. It's a popular domain with thousands of non-duplicate pages each of which has unique content that changes every time a new joe-schmoe writes a review. I don't see how they could get those pages out of the serps, other than banning them (which wouldn't be appropriate as they're doing nothing wrong) or hugely overemphasizing backlink anchor text (which would replace them with lower-content spam).
Hopefully once they finish tuning this semantic stuff, the Amazon pages will naturally start showing up only for searches that are very relevant to them (the book title, the author's name, a phrase that's actually repeated on that page, etc.) That would be just fine as far as I'm concerned.
Amazon has the most incredible backlink network through its own affiliate programme, you just can't track it because it includes an affiliate i.d. which isn't present in the page you see on the site.
Everyone else's affiliate scheme points to a tradedoubler type organisation, who gets the benefit of the links, whilst Amazon, who pretty much invented the concept built their own and accidentally enjoyed the benefit of all that Google juice.
And when if comes to long urls, you will notice that although they are long, they are also free from the dreaded query '?'. Again an accident of home grown tech, in my opinion but has served them extremely well.
Amazon are there through sheer merit by every measure you care to through at it. IMHO anyway. You're just all jealous ;-)
Steve
And when if comes to long urls, you will notice that although they are long, they are also free from the dreaded query '?'
It's still dynamic. Just put a '?' in place of the last '/' and you get the same page. Anyway, Google can crawl pages with the '?' just fine.
I don't buy this Amazon/Google conspiracy.
Just because we're paranoid, it doesn't mean someone isn't out to get us. :)
Most of Amazon's result in the new Yahoo! are Yahoo!/Inktomi trusted feed results. Why wouldn't a behemoth like Amazon have something similar in Google - even if it is just a private datafeed? If I were Amazon, I would definitely approach Google with this; and if I were Google, I definitely would work something out.
Do I think Amazon would be paying directly for something like this? No. But there are other ways to compensate for this - such as a higher cut for Google on Amazon based AdWords ads.
Google can make more off of one Amazon than 10's of 1000's of little guys.
You're just all jealous ;-)
Not jealous. Just making some observations and assumptions.
seasalt
Does google spend its day asking how to scr*w SEO? No.
Maybe this is part of the reason why so many recent flucuations are taking place/
