Forum Moderators: open
Dave.
Does that means that anchor text and (or) *newly adopted* anchor text filters have not been factored in for the new version?
Otherwise looks like anchor text has been severely devalued or maybe even penalized?!
What a room for knocking down competitors! Just got to all FFAs you can find and put your competitor's title in the anchor text! Easy - peasy :(((
Are those black hat or white hat tactics, Omnipotent Google Gods?
[edited by: WebmasterFisherman at 1:25 pm (utc) on Nov. 18, 2003]
About industry specific links:
In the real world you do not get a lot of industry specific links linking to your site because your competition just will not do this. Likewise you will not link to another person selling widgets. So a legitimate link would more than likely be from a person in different industry. Changes to favour the industry specific links hurts the merchants. I wonder if this is being done to encourage ad word sales?
[edited by: DocElder at 1:41 pm (utc) on Nov. 18, 2003]
Besides a pure eyes-of-God algorithm, the very best algorithm would look highly on quality sites in one niche that look favorably on quality sites in that same niche. When Ted Williams said Tony Gwynn had the best baseball hitting technique he had ever seen, that meant more than a dozen sports talk radio philosophers saying the same thing.
It will always be likely that widget sellers will not very often link much to other widget sellers, but then, those that have the confidence to do it, and those that get the links will very likely be high quality sites.
It would be a good thing for an algorithm to appreciate that.
1) Message board spam
2) 10 keywords spammed at bottom of page all interlinked to 1 page sites.(different domains) Theres like not even a paragraph of content on any f the sites.
3)Same Guy different domain
4)Same Guy diffrent domain
5)Same Guy different Domain
6) Same Guy different domain
7) Message board spam
8) Good Site
9) Good Site
10)Same Guy different domain
Sticky me if you want the search term
Dan
1) This algorithm is a change to improve business for Google.
2) Google revenues comes from advertising sales.
3) Any changes that encourage more advertising are good for Google.
Industry specific links will also help sites that are not merchants like the .gov or .edu sites.
Let's be real and get out of the fairy tale:
If you sold Blue widgets you would not link to any site that sold more Blue widgets than you for less money. This is a fact.
Googles search directory is just an advertising business. I think we forget this sometimes.
[edited by: DocElder at 2:02 pm (utc) on Nov. 18, 2003]
What you describe is what has been a winning combination for several months, and some of that is still in the serps, but algorithm changes to value quality/peer linking and downgrading reams of anchor text links from bogus family "sites" is a seriously positive thing. I'm not saying at this point that is what occurred, but certainly this could possibly be something that was at least attempted.
Hopefully soon the adult and commercial sites will sort themselves out equally well and you can all have a big sigh of relief. (-:
This does not make good long term business sense to me. The only reason Google has the traffic that it has is because of the quality results(until Sat) that it produced. If this change sicks, then I think over a period of time people will use less and less of Google to find what they are looking for. Adwords are the dessert to the main course meal. If the meal sucks then the dessert will not follow.
Just my opionion
Anchor Text:
It appears *at this point in time* that Google used to give more credit than I thought to internal backlinks and anchor text pointing back to the homepage. Now it appears that they give almost none. Reasons-to-believe this include: index pages with high rankings in allinanchor, allinurl but with no corresponding showings in the SERP's. In their effort to curb spam, they turned the dial waaaaaay too far...
Early guess: he/she in here who supposed that leaving the "-" thing untouched was a programming ommission (or decision?) at the Plex hit the nail on the head. Problem is, paired words with and w/o dashes often have different meanings or connotations (e.g., "go-go girl" vs. "go, go girl").
If G applies the same filter to the keyphrases connected by dashes that it applied to the keyphrases not connected by dashes, there will be new collateral damage in a smaller set of very innocent pages...taking the situation from bad to worse. How about just turning the dial back? :-)
I have been #1 for nearly a year (except in may) for a very competitive search term and managed to get #1, #2, #3 positions only 2 months ago.
Now I am looking down on page 2 or page 3. It's totally bizarre to me, the guys above me seem to have gotten there by sheer luck.
Having said all this I am not yet worried! The whole serps look a complete mess, my site seems to be on page 2 and page 3 for the same search term and even when I type in the domain name of my site without the 'www' or '.com'
the site will appear, but it will appear lastly after all my inner pages. Anyone else seeing this?
My only hope is that www-mc seems to be showing results which are more in my favour although I cannot help think that this is old data from at least two weeks ago.
What is the significance of the google labs viewer? The results on there are always different and look quite tastey from where i'm looking.
Finally, that kw1-kw2 thing does nothing for me, if anything it puts me even further down the serps.
I think we just need to wait this one out.
This algorithm is a change to improve business for Google. Their money comes from advertising. Any changes that encourage more advertising are good for them.
Has no one considered the possibility that Google may be trying to improve the quality of its search results?
For what it's worth, I haven't noticed any significant changes for the keywords/keyphrases or the competitors and comparable sites that I monitor. (Note that I'm not just talking about my own site.)
The main difference I've seen is that more directory pages from the ODP and Yahoo are showing up in the top 20 or 30 search results. I'm not sure that's a good thing, but it's no worse (and probably better) than finding boilerplate affiliate pages in those same positions.
I suspect that a lot of the people who are complaining about the latest update are those who have relied too heavily on easily detected SEO techniques. Techniques that aren't inherently suspicious or even "spammy" in themselves, such as optimizing heavily for anchor text, may cause problems when used in combination. For example, it wouldn't be that hard for Google to detect a combination of keyword-optimized anchor text AND a keyword-optimized title AND hyphenated domains with the anchor-text keywords AND high keyword density in the body text--and to give less weight to each of those factors if they exceeded a certain threshold on the same page. This wouldn't be a penalty per se; it would merely be a correction or weighting shift to compensate for what Google regarded as artificial (and excessive) SEO.
The only reason Google has the traffic that it has is because of the quality results(until Sat) that it produced.
As far as the user can see, it is still returning extremely high quality results.
TJ
<added: in fact, as far as I can see it is still returning extremely high quality results>
[edited by: trillianjedi at 2:27 pm (utc) on Nov. 18, 2003]
It looks to me like a points scoring system. When you get too many points you get demoted for your keywords. So far as I can make out you score points for:
keywords in title
keywords in headings
keywords in normal body text
keywords in outgoing links
keywords in the majority of inbound anchor text (especially links internal to your site)
to me it seems if you have a couple of these it doesn't seem to cause a problem. But if you score in most of them you get the royal boot.
What's important now is - is our "over-optimized" websites doomed for life or is there something we can do to correct the situation (other than throw away the domain and the company name, since in my cas this is the same etc)...
Nightmare i tell ya
Has no one considered the possibility that Google may be trying to improve the quality of its search results?
Of course, in fact I'm sure that's what they're *trying* to do (I don't think it's some big conspiracy to sell more AdWords), but I think something quite big has gone wrong for seemingly a number of sites.
And if they are trying to correct excessive SEO (ie the title, text and anchor text are all the same - not spamming, but optimized) it's not working very well. If someone is searching for (as I mentioned before) 'Widgettown FooCorp Resort and Spa' and Google doesn't show that site in the top 100 listings then it's clear something is wrong.
GoogleGuy mentioned that there'd only be a "small" amount of data added today sometime... I don't think a small amount is going to help much.
[edited by: canuck at 2:49 pm (utc) on Nov. 18, 2003]
Is anyone else seeing this? Has their been some sort of shift in inktomi & AV algos or are people flocking away from google?
OK So let's assume there is a penality for "over-optimization" with anchor text links - I think this theory makes sense. What's important now is - is our "over-optimized" websites doomed for life or is there something we can do to correct the situation (other than throw away the domain and the company name, since in my cas this is the same etc)...
I don't think it's as simple as Google's saying, "Hmmm...Widgets.com has links to "widgets" and "widgets" in the title of its home page. More likely, it's a combination of many things that, when taken together, create an artificial pattern.
Instead of trying to second-guess what that pattern might be (or how it's detected), the safest approach may be to simply do what comes naturally. In other words, try to think like an editor, not like a bleeding-edge SEO. Stick to the basics: useful content for the reader, descriptive page titles and headlines and links, etc. Resist the temptation to over-optimize, and you're less likely to run up a "spam score" that can result in a correction.
In relation to Google, you might ask yourself if you're trying to help Google or trick Google. A title that includes the word "Widgets" is good spider food. A title that says "Widgets Blue Widgets Red Widgets Widgeting and Widgetizers from Widgets Inc." is more questionable, because it's obviously artificial.
Interesting theory but I can see sites that are highly optimized (as you describe) doing very well. On the other hand, I have a site that has almost zero SEO and none of its incoming links have been solicited. The site can no longer be found when searching for KW1 KW2 but can be found when you search KW1-KW2. The answer to this mystery lies in the hyphen. Does anybody have any theories on how/why the inclusin of a hyphen displays the missing sites?
Here's hoping!
That theory definitely doesn't work for educational sites... in the searches I just did, most of the top-ten sites (including mine) have the topic in the title, heading, links, and text body. Google may devalue those factors beyond a certain point but it certainly isn't penalizing for them; otherwise these educational searches would not look so good and relevant.
Of course, educational sites tend not to view each other as 'competitors' exactly, so we're all linking to at least some of the others. This may stabilize those sites somewhat against spam or spam-busting techniques, maybe?
Anyway, I find it hard to believe that Google could be penalizing anyone for getting too many links or having keywords in the title or anything like that. If so these educational sites would be dropping like flies, and instead, the educational searches look better than ever. My personal guess is that some of these missing commercial sites must be getting caught in an automated spam filter of some sort; hopefully the innocent ones are being released as fast as possible upon review. Or maybe some factors that favor commercial sites just haven't been factored in yet. I'm just idly hypothesizing, really. But I *don't* think it could be possible that links, titles, or headings could be getting penalties given the search results I'm seeing on educational topics! Pick a history topic of the sort that kids might be studying in high school and type it in yourself to see. (-:
Europeforvisitors I have a real problem with this theory of yours.
How can Google realistically lessen the relevancy weight of a site which happens to be linked to it using a logical description of itself?
How can Google realistically lessen the relevancy weight of a site which has a domain name which describes itself?
How can Google realistically lessen the relevancy weight of a site which has a TITLE which describes itself?
How can Google realistically lessen the relevancy weight of a site which mentions things about itself in the body of its own text?
I realise you say "if they reach a certain threshold" but I personally think if there was such a threshold then it would need to be set very very high.
Imaging a site selling all coloured widgets. It's homepage want to mention each colour of widgets individually and naturally link of to pages which are colour specific. This site could have the majority of its inbound links with the search term "coloured widgets" in them. It could logically have the phrase "coloured widgets" in the TITLE and meta description. Also in the main body of the text it could mention the phrase **** coloured widgets (where **** = colour) many times in conjunction with each different colour of widget.
As a result of this, this site could have many inbound links with the phrase "coloured widgets" in them. It would have the same phrase in the TITLE and Description and it would also mention the same phrase many many times in the body of the text and logically in the internal text links - ALL PERFECTLY LEGITAMETLY.
So if you had to draw a threshold line I would say it would have to be mighty high.
[edited by: needinfo at 3:18 pm (utc) on Nov. 18, 2003]
Whats the URL of data center 8?
[edited by: prodigyfx at 3:45 pm (utc) on Nov. 18, 2003]
You want to see the future of SEO and SEM?
One clear example is EFV's site.
- jeff esq -
As time marches on without significant change in the SERPs, I am left to conclude that I did indeed over-optimize and we're seeing the final result (notwithstanding some minor changes).
The $64K question now is when will Google recognize my page changes and put me back into the game? If this is indeed a penalty, is it a "hard" or "soft" penalty, with the latter meaning that the issue will correct itself after Google fresh crawls and indexes my new pages? Or, <gulp> will the penalty stick for a while?
but is there no added relevancy at all?..you have to check backlinks and how relevant the backlinks are for the -keyword...id have thought many of them will be or point to urls with -keyword simply by the nature of on topic reciprocals
When is this nightmare going to be over? :)
Thanks
David
>>Whats the URL of data center 8?
The URL of data center8 is [www-cw.google.com...]
It contains my site SERPS before changes.
What google is trying to do is eliminate SEO. Scrap your lame black hat/what hat talk. It doesnt matter to google. Fundamentaly, if you are manipulating the SERPS, they dont want it, period. How can you make money off of PPC when people can manipulate the results? Get over yourself, google doesnt care.
So, what they did was study organic high ranked pages vs. SEO pages, find the difference between them, and then penalize sites that are doing SEO.
On to the meat. What google did is add a new algo for every page in the index that states:
1. Look for high keyword phrase density in H1, title, in-site links, text, etc. This identifies keyword phrases to watch out for in step 2.
2. Look at inbound link anchor text. Find pages with a high percentage of inbound link anchor text made up exclusively of keyword phrase identified in step 1.
3. Add a massive penalty for pages that are doing things seen in steps 1 and 2.
How do I know this? Type allinanchor:keyword1 keyword2, check results compared to keyword1 keyword2. Lots of SEO sites now appear with the allinanchor results. Also, check keyword1-keyword2 vs. keyword1 keyword2. You can see a huge difference here. Ive seen it on all the sites I run. Keywords on inbound link anchor text have been killed.
Final piece of the new algo, that applies to sites, not pages.
1. Use results from previous step 1 and 2 to identify SEO pages.
2. If these pages are the main index page of a site AND there are little or no deep links into the site with the keyword phrase, apply another penalty.
You will see lots of SEO site index pages dropped. Organic sites dont exibhit this behavior.
Finally, what can you do:
1. Inbound link anchor text must never contain more than 20% - 40% of your targeted keyword phrase.
2. There should never be more than 50% of your inbound links pointing at your index page for you domain.
Happy Hunting!
Am I the only one that is noticing an updated toolbar PR, like from today or yesterday?
No. Looks to me updated across the board. We've gone up a notch on a couple, others remained static.
PR is irrelevant though in terms of the major address of filters in this update, so not surprising that everyone is discussing other things.
TJ
Dangdiggetynabbitt - you try to play by the rules, and they zap you for playing by the rules? My site is squeaky. During the whole Dom/Esm shakeup I was so freaked out that I scrubbed every speck of even possibly vaguely questionable stuff off the site.
Well, I guess I'll wait until the official "all clear" before making any changes. This totally sux.
For all of my websites I have used the same "over optimized" techniques such as: heavy inbound links with the same keywords, heavy keyword density on the meta-title and in the main text, heavy interior linked pages using the same keywords, etc!
So, my conclusion is that new link data has not been factored in and when it does many sites will return. My overly optimized sites that have been ranked for over 8 months have not been touched.
There are exceptions to this, but on 20 of my sites that I track, "overly optimized" sites seem to be working just fine it seems to more a timing thing with the set of data Google is using. Thanks!
Any opinions on this?
Among other things, this group is great for venting :)
Now we just need a couch to lie on:)
Thanks
David
I am seeing similar results ... My client's site is no where in the SERPS and 9 of the 10 data centers confirm this.
Any idea when there will be another Google update?
It seems Google is using 3 month old data and has not factored in PageRank and filters for SPAM and irrelevant content. Suggestions, thoughts..?