Forum Moderators: Robert Charlton & goodroi
I can understand penalties for jamming in keywords in Long-Urls-that-makes-sites-look-like-spam but otherwise im gonna do what I have been doing for years!
where have you been reading this? a forum/blog? or trustworthy source?
".. whether the occurency is in a title, in a URL, in the body, sidebar .. "
See Matt Cutts view here,
[google.com...]
Yes, just read the patent about -950 reranking.".. whether the occurency is in a title, in a URL, in the body, sidebar .. "
I'm not sure if you're trying to say a keyword in the URL is bad or not, but the patent referred to is actually about spam detection based on the actual frequency of a phrase and related phrases and the ability to predict a 'standardized' frequency of the phrase and related phrases contained in a document.
In case I'm not saying exactly what I mean in English above: They're looking for the presence of a phrase and related phrases in a document within a 'threshold' of what's considered to be a 'normal' occurrence rate for the phrase and related phrases.
I keep laughing about this, because the only way I can see Google actually penalizing a site (or page) solely based on the presence of a keyword in the URL is like this:
Google Exec: How can we maintain our lead in searches conducted over the other search engines?
Google Engineer: We can penalize sites if they use the keyword in the URL.
Google Exec: What will that do?
Google Engineer: Well, if a person searches for a keyword or 'key phrase' and we've penalized all pages using that keyword or phrase in the URL because it seems to indicate they have what the person was searching for the person will have to search again because they won't easily be able to find a page that's actually about the phrase or keyword... Then people will always conduct more searches when they visit our search engine!
Nearly ever page from php.net and here on webmasterworld.com (just to mention a couple) would be penalized if the occurrence of a keyword or phrase in the URL invoked a penalty...
+ it makes sense right?
I think its excessive usage of "keyword based anchor text" used for internal linking which could raise flags e.g if your site is about widgets and your home page has 30 links to internal pages and all those 30 links have anchors like blue widgets, green widgets, gray widgets, black white widgets etc, it might lead to an over optimization penalty.
I think its excessive usage of "keyword based anchor text" used for internal linking which could raise flags e.g if your site is about widgets and your home page has 30 links to internal pages and all those 30 links have anchors like blue widgets, green widgets, gray widgets, black white widgets etc, it might lead to an over optimization penalty.
Big Red Widget
Small Blue Widget
Long Hairy Widget
It sounds like that would incur the same penalty you're talking about, but it's really necessary to have "Widget" in each of those. Big Red? Long Hairy?
If its internal linking, i have seen sites doing perfectly fine with excessive usuage of similar keyword for internal linking, such as "Big Red Widgets" "Small Blue Widgets" "Long Hairy Widgets" where Widgets is the keyword, to me i think its over optimization, a little bit of screaming, specially when all of your internal links have keywords "Widgets" as their anchor text. You can probably use "Big Red" "Long Hairy" for your anchor text, but on particular landing page, you can set meta titles which truly indicates the nature of your page, which is Big Red Widgets, you can also use it in H1 tag (not that they are too much of value anymore) and somewhere in text as well. Finally, get few links to that particular page targeting your keyword. Someone correct me if i'm wrong.
[edited by: zehrila at 2:42 am (utc) on Dec. 3, 2009]
Wherever you read such nonsense, stop reading there.
Also, let's make sure we are all talking about the same thing here:
URL - the complete address for the page
Domain Name - the part of the URL up to the first single forward slash
File Path -the part of the URL that follows the first single forward slash
I say this because MANY people are pretty loose about which term they use, especially switching the first two around.
I've been reading about Google applying penalties for having keywords in the URL.
Wherever you read such nonsense, stop reading there.
[webmasterworld.com...]
tedster says:
I'd say the point is don't use the same keyword in the anchor text for many different internal links. Some repetition is certainly natural, but too much can hurt you.
I've been reading about Google applying penalties for having keywords in the URL.
Wherever you read such nonsense, stop reading there.
[webmasterworld.com...]
SEOPTI says:
You should worry about keyword rich internal navigation and keyword rich URLs. This is the perfect mix for a -950 re-ranking.
I think this is one of those fairly easy questions to answer just by observation and it will probably be found either the post of SEOPTI is being completely misunderstood, taken out of the context it was intended for, or is incorrect. In any case it's a single post and can easily be evaluated by simply searching and reviewing what the sites you see are doing.
And it's just FUD. Using keywords in URLs is friendly to users, search engines, and yourself. This forum is called webmasterworld.com/google/ not webmasterworld.com/aysts/ for the benefit of everyone. There is no Google downside to keywords in the URL, and plenty of positives for Google and humans. And more to the point, the most well-constructed, well-ranked websites do it this way.
Shhhhhh... Don't go telling everyone how easy the question is to answer. Make them at least search a bit. K? Thanks! :) And don't even mention the bread crumbs, or the links you click to get into the forums, alright? I mean seriously, every forum about Google has the word Google in the link. That's a penalty right?
So /widgets/blue-widgets/widget-service.html seems to me to be a helpful structure and in my experience works very well on Google.
Keyword stuffing of URL, anchors, nav and content is a different matter IMO. I don't think there's an OO penalty but there is a keyword stuffing (stupidity) penalty. You can stupidly highlight your stupidity by stuffing keywords in all of the key areas of your site, then when you stuff the content of a file boom you are stuffed!
Cheers
Sid
PS Some folks still think that Florida was about an over optimisation penalty. IMO this was a myth propagated by people who were not bright enough to realise what was really going on.
Think about what you would call folders & files on your computer and that's probably a good idea for URLs. That's all the Internet really is, so it stands to reason you should name things accordingly.
IMO this doesn't make sense:
widgets-blue/fuzzy-blue-widgets/super-fuzzy-blue-widgets-for-sale.html
Because IMO that's not anything most people would call the file or directories on their computer...
widgets/blue/fuzzy/super-fuzzy/for-sale.html
AND
widgets/blue/super-fuzzy/for-sale.html
Make quite a bit more sense to me, even though the word fuzzy is repeated in one.
Personally, I might shorten your URL a bit depending on the exact situation:
/widgets/blue/service.html
Of course I might leave it very similar to what you had, depending on the exact situation. ;)
I was able to nail it down to the keywords in the URL after months of testing, believe it or not.
;-)
Also, how many URLs are using 'essentially the same' pattern both as a count and percentage of the site?
There seem to be quite a few sites in the index with a keyword or two in the URL(s) that are not penalized...
Of course I don't believe it, especially since I'm on a page with keywords in its URL and it isn't penalized.
But now anyone else reading this with make up their own minds.
I was able to nail it down to the keywords in the URL after months of testing, believe it or not.
The phrase-based spam detection patent [appft1.uspto.gov] mentions these factors, among others:
grammatical or format markers, for example by being in boldface, or underline, or as anchor text in a hyperlink, or in quotation marks... whether the occurrence is a title, bold, a heading, in a URL, in the body, in a sidebar, in a footer, in an advertisement, capitalized, or in some other type of HTML markup.
Based on that patent, and my experience with -950 sites that people brought to me - the threshold is triggered by a combination effect. That is, measuring single factors in isolation builds an assumption into the test that is no longer the way that the ranking algorithm is constructed.
This means working with a single factor can produce correlations that appear to be "the cause", but the full cause is actually having the combined score cross a threshold. This is further complicated by the fact that the threshold is regularly recalculated - so sometimes your site crosses the threshold and sometimes the threshold is what moved, not your site. And as a side note, I expect that Caffeine will allow more frequent threshold calculations.
Certainly some factors in the combined factor score are weighted more heavily than others. I'd say the heaviest weights probably do go to keywords in the anchor text, URL, and title.
Sid, given all those moving parts, did your testing take at least some of them into account?
It was SEOPTI who said that, I was being ironic with my ;-) while thinking exactly what steveb said.
I have a hypothesis that goes back to Florida and all that talk about semantics (which I fully accept) and that is this:
If you have a page/site that is keyword stuffed in all of those areas mentioned in the patent you point to AND the page/site is not semantically rich around that topic then you will get a 950 BUT it is much harder to go through the threshold and get the penalty if your page is semantically rich on that topic.
Cheers
Sid
it is much harder to go through the threshold and get the penalty if your page is semantically rich on that topic.
Yes! -And that effect is a result of the phrase-based indexing methodology, too. If the semantic richness (co-occurring phrases) goes way too low or way too high, that's when it kicks in. The "way too high" is something that can nail mix-and-match scraper pages.
When those patents were first published, it seemed clear that copy writing should be much more natural than many SEOs were advocating. Even before that time I noticed that co-occurring phrases could boost some rankings. Those phrases also add a powerful new dimension to the long tail traffic.
[edited by: tedster at 7:13 pm (utc) on Dec. 5, 2009]
