"Real" Pagerank (as opposed to a toolbar guess) has been given to dynamic pages for the first time. Dynamic pages tend to be updated more frequently. That's why I suspect that there's been a major shuffle in some competitive areas. Those areas - like <edit>widgets</edit>, to use the above example - tend to have a lot of dynamic sites for all the different information. Pages "not yet ranked by google" always have shown up in SERPS, but there would have to be an almost exact term match and little competion for it to show up near the top. Now that there are millions of pages with a real page rank that didn't have one before (even a PR1 is better than not ranked, right?) the SERPS are going to be majorly different in those areas.
[edited by: NFFC at 11:27 am (utc) on Sep. 29, 2002]
What I found to be weird - for the first time ever this has happened to us on Google with a new Client's site. It got a full crawl beginning of September - plently of good inbound links - but not in the index... A quirk? Probably... Google is still the best SE - nothing in life is perfect! :)
To give my example:
I have had top ranking for a 'location widget' search all year. Not a huge location like NY or Barcelona.
The interesting thing is I don't know how it got 1st place anyway. It is a PR2 with no outside links. It remained #1 until the last update. Now it is #84. Ironically if I search for 'location' only I am #11. Weird!
If this is not just a settling period then it would appear to be a very specific algo change. Perhaps shifting weight from keywords to links. I honestly don't know.
Google is still the best search engine in my opinion. I never use anything else. Last week I was searching for a folk guitarist who's name I didn't know but who had a bad car accident which meant he could no longer play.
Using Google I found him quickly. I also found the lost song title I was hoping to find.
Basically, the guys at Google know exactly what they are doing and they are very good at it.
Remember the quotation "There are lies, damn lies and statistics!". Google control the statistical analysis through the algo's. One minor change and 50% of people here could be seriously affected.
Just my 2 cents.
[edited by: NFFC at 2:01 pm (utc) on Sep. 29, 2002]
"widget" is a money word, you bumped into the adwords algo.
[edited by: NFFC at 2:01 pm (utc) on Sep. 29, 2002]
I do not want to be rude and fully respect the laws on this forum. But I do not understand why my post and SlyDogs which I was quoting in my Edit in was edited.
[edited by: NFFC at 2:04 pm (utc) on Sep. 29, 2002]
I have had a look around and have a feeling that what I think started in the previous update has been given more weight in this update and that is freshness.
In my categories SERPs this month, I'm seeing what I started to see last month - a lot of "dead" links being cleared away. And I'm saying "Thank god" because For the last year or so the first two pages of SERPs have been filled with sites that stopped being updated a long time ago, just because they still had a lot of PR.
OTOH a site I hadn't updated in a while just dropped down to the third page. Ah well - can't win 'em all. Time to go update it I guess.
In my case freshness disappeared. All my "minty fresh" links that had been added - and had greap serp - from mid to late Sep were dropped in the update. I am waiting to see if the reappear.
Visit Thailand, feel free to take the matter up with any Mod or Admin via stickymail.
Any search terms will be edited out, if blue fuzzy widget etc will not do then please don't post.
I think Google is trying to give more weight to the "externalness" of incoming links -- so that unreciprocated links from outside sites will carry more weight than links which Google identifies as being internal or reciprocated.
In this index, I've been finding that my own pages which stayed stable or else improved are those where outside sites have linked to the content but I haven't linked back. Such links have always been A Good Thing, but now they're worth more than before.
That's my theory and I'm sticking to it.
(Lurker's comments) as per GG's comments on this thread:
"Remember GG's comment that some penalties were to be lifted?"
Could not this update's results be influenced by new "spamming" algos put into place this month by Google?
Seems to me that Google could never do "case-by-case" review of banned sites and if penalties are indeed lifted, this would be done across the board with new filters in place to trap those sites that were still guilty of spamming. But if one single spamming technique got a site banned - and that spamming technique was subsequently cleaned up - that might allow for a site banned for that one technique to be placed back in the index and to bounce to the top based on the OTHER 5 SPAMMY TECHNIQUES still employed in that site. Maybe...?
hmmmmmm...you may be right bcushion.
One of the super spammy sites that has slid into the top ten in my category is of particular interest to me and I am going to try and do a thorough analysis. The reason I noticed it, is it is based in a tiny country and I had been surprised by a ton of hits from what appeared to be primarily one visitor from this country in my logs a few weeks ago.
The kicker is, the search terms are (HYPOTHETICAL) "country fuzzy widget" and the country this site is now has a top 10 serp for is NOT their country of origin or even one they have any content for other than a single page with 1 paragraph of generic text.
>with new filters in place
A point worth raising is that we may be looking at multiple filters, there is no one single answer, some findings will contradict others.
Grumpus, regarding your comment:
'"Real" Pagerank (as opposed to a toolbar guess) has been given to dynamic pages for the first time. '
I've had real pagerank for dynamic pages for months. But I can't disagree with you that dynamic pages have been treated more kindly this time around.
I experienced a huge drop in the serps for my main keyword but my Google referrals went through the roof. My interior pages are getting loads more traffic, especially the dynamic ones. For numbers, say I got 1,800 referrals from G on a typical day, now I'm getting 2,200. I was sick to my stomach when I cheked the serps but feel lots better now that I've looked through the logs.
observations in my industry:
nbc, yahoo, dmoz dir, cnn, and other sites not specifically geared towards main keyword have bumped me & competitors down. lost some sites completely and many affiliates gone, definitely a tighter spam filter. unfortunately, many broken links and relic sites at the top...take the good take the bad
I am interested in your adwords theory. I saw you post somewhere that you thing there is a threshold that when there are more than 4 ads on the page an algorithm kicks in. What would the purpose of that algorithm be? It is certainly not to kick the advertisers off the SERPs page. In my category all the top sites are also advertisers.
I am also interested in the dynamic pages theory. If a site has a million possible permutations for a dynamic page, will google crawl them all and give each page pagerank? That would seem unfair because even a basic site could generate a million pages without much effort.
Google must have taken this into account.
>In this index, I've been finding that my own pages which stayed stable or else improved are those where outside sites have linked to the content but I haven't linked back. Such links have always been A Good Thing, but now they're worth more than before.
For health related terms, the algo changes have been very easy to see. Basically, DMOZ pages rule the roost. For a heavy content site that I have with a lot of DMOZ links and unreciprocated links from universities and schools, almost every page went up significantly in the rankings for it's target keyword - even for pages that I have not touched all month and pages that have no new links.
Most of the mom and pop listings for the health terms that I track are gone from the top rankings. It is sites with lots of DMOZ pages, pages from universities, national foundations, etc. that are in the top 20 now. The sites that achieved high rank by having optimized pages, multi-keyword domain names or trading links are mostly gone. On page topic specificity seems less importantance than how many links the site has from DMOZ and other quality sites. The sites made, as someone from another post decribed it, by "FP hacks", have greatly diminished.
I think the current rankings reflect a move towards Google's goal of delivering quality content to web searchers. They've just found some better ways to do this now through their algorithm. I suspect the current algo is here to stay. For health terms, the sites with bad grammar, dubious products, and as seen on TV type claims have all been given the boot and more of the serious, informational sites are now ranking higher.
>> In my category all the top sites are also advertisers. <<
You are probably focused on kw's that don't make it into major news sites, .edu's, large corporates or general interest sites. Those categories are immune (I can only guess what they are).
|from taxpod: I've had real pagerank for dynamic pages for months. But I can't disagree with you that dynamic pages have been treated more kindly this time around. |
As things get more and more stable, I'm seeing that roughly 15% (or so) of the pages that are in the Google index from my site now have a real pagerank. The rest are still a "guess" in the toolbar and come up with a "HUH?" when you put the URL into the searchbox rather than the "similar/linkto/cache" page. Maybe it's certain types of dynamic pages? And, another likely factor is that the page might need to be in the index for several months and Google must determine it to be significantly different than the other pages within the site (which could go along with SlyOldDog's question)... All I can say is that I've had pages in the DMOZ for months (some of them in there for 4 months and still not crawled) and this is the first time that any page outside of my index page had a result when I put the URL into google's search box.
My pages that are in high PR categories in DMOZ got a real PR boost this month, so the DMOZ observations ring true for me as far as that goes, although another page that isn't in the directory got a nice boost, too...
I agree there is more content dominating; however that content is not relevant to the search term.
I spent the last hour doing single keyword searches for items in my office. Simple things. Single keywords.
The results were at least 1 of the top ten had NOTHING to do with my keyword. Nothing at all. There were errors on those sites, moved domains...and just plain unknown reasons for those sites being there. In certain cases the keyword was not on the page, nor did the website show ANY backlinks. It was as though it was a random result.
I'm still leaning towards a randomizing 'feature' being toyed with.
I've run across several 'media kits' on the web that are now going to be outdated. (This site is #3 for the keywords 'widget puller') Well, not anymore. It's probably #53. And next month maybe #2...then #21. No more static media kits.
Another item I noticed is that DMOZ shopping/.../keyword websites seem to be displaced.
But, I digress. I am still leaning toward a set of rules that tend to produce some random results. And my belief is that this new algo is not going to hold. It is producing too many irrelevant results when COMPARED to previous updates.
"Widgets" are in my URL, header, outbound links and also show up several times in the main text on the index page.
Early on in the update my page was listed as number one, now half way down on page two, listged number one has over 8,000 inbound links to my 420, several other pages on my site have had the same fate!
The number one site has only got "widgets" mentioned a few times on the whole page, and only in the main text of the page.
I know what I will be concentrating on this month!
Submitting a new site 2 months ago, I had a site go straight in to #1.
My main site has moved up from #4 to #2, all in all a good update for me.
Around this time last year, my main site disappeared from the top 10 to the hundreds (no hidden kw's, no funny tricks) reappeared in January '02 - seems like Google is getting ready for the xmas season.
>>Most of the mom and pop listings for the health terms that I track are gone from the top rankings. It is sites with lots of DMOZ pages, pages from universities, national foundations, etc.<<
Right...you get the picture, those highly targeted, kw rich (and relevant) sales sites are getting blown out in favor of larger (and vaguer) results.
Now, this works great for you this month, but lets say that next month, google still isn't happy with the "pinky clicks" and they decide to dumb it down a bit more and favor only the *real* authority sites that are linking to you like the .edus, .gov's and kick you back about 20 pages.
You could still say "I think the current rankings reflect a move towards Google's goal of delivering quality content ...."
But I don't think you will.
> I agree there is more content dominating; however that content is not relevant to the search term.
Good point. I do see this in the health terms, too. I would think that it is something of a trade-off for the Google algo folks that when they give more weight on the site importance they are going to lose somewhat on term specificity matching. It will be interesting to see what they do next month. I think their trade-off is to keep the focus on site importance and decrease some of the spam, or increase page specific term matching criteria to get better relevance but also let more spam sites back in.
I am so glad to see this thread start. I tried to start something similar a few days ago, but it got moved into the google is horrible thread.
I don't have time to post more now, as I am heading to the river (salmon are running) but I wanted to add something for some of you to think about.
There appear to be two categories of changes here. One is in the crawling and indexing, the other is that there seem to be changes in the search algo. If you can, think about where the specific change you are seeing, might be located.
More analysis later.
I'll put my waders on.
> You could still say "I think the current rankings reflect a move towards Google's goal of delivering quality content ...."
> But I don't think you will.
Nah, I actually expect university and hospital sites to eventually dominate in the health rankings even more than they did with this algo. Google is staffed by PhDs, so it would be logical that they would give more weight to university and big institution type sites.
|Google is staffed by PhDs, so it would be logical that they would give more weight to university and big institution type sites. |
...Scott starts searching the classifieds for a big university in need of a 'webmaster'...
Its easy to take your eye off the ball here.
|In general, we judge a page by its links and content. |
Straight from GoogleGuy today! This is the basic principle from which all else flows. I think its probably simpler than many of us are making it out to be.
There seem to be a lot of theories here, and they don't seem to be easily proven, so this is what I propose:
Maybe Brett or someone else can make a small survey page, or perhaps we can just list the answers to some basic questions in our posts so we can get a better understanding of what happened. Questions like...
1. In general my site used to rank
a) first place
b) first page
c) second page
d) 3rd-9th page
e) 10th + page
2. In general my site now ranks
a) first place
b) first page
c) second page
d) 3rd-9th page
e) 10th + page
3. When I exchange links, I tend to get
a) inbound links (assorted)
b) inbound links (themed)
c) inbound links (assorted with anchor text)
d) inbound links (themed with anchor text)
.... more questions here (suggestions?)
Then we can all analyze the answers to this survey, and filter out people who jumped in rankings, fell, etc and see what happened to these cases. I'm sure there are more than enough people here that can give their insights.
A poll like this immediately assumes all 'sites' are equal, when they aren't. Some are commercial in nature, some are hobby/informative. Most are both. I suggest qualifying each 'site' before polling.
Any way we can keep GoogleGuy from seeing the results? ;)
Otherwise 2 months down the road we'll be having another poll!
Seriously, I don't see much mileage in a simple poll. I couldn't answer question 3 if I wanted to. The answer is "a bit of this and a bit of that".
I think we just need to see as many theories as possible and test them on our own sites. The ones that most people agree with will probably be close to the truth.