| 11:53 pm on Apr 10, 2003 (gmt 0)|
Ultimately, competition dictates the answer to your question.
1% more than competition - 1 PR higher than competition... so the game goes.
Common knowledge is 5-20% density.
But in the words of DeeJay..."if it ain't broke don't fix it"
| 1:23 am on Apr 11, 2003 (gmt 0)|
20%? you must be high
| 1:24 am on Apr 11, 2003 (gmt 0)|
anyone have a serious answer?
| 1:29 am on Apr 11, 2003 (gmt 0)|
Section 'F' : 5 - 20 % density
| 1:40 am on Apr 11, 2003 (gmt 0)|
I've seen the 5-20% density suggestion a lot. My question is 5-20% based on what? I've been using the Search engine world one and I uncheck the meta items. Is this what others are doing when they say 5-20%?
| 1:44 am on Apr 11, 2003 (gmt 0)|
You looking for enemies? Or do you generally spout comments without reflecting?
Brett says 5-20%, others say 2-7%, etc. etc. It all depends who you are reading and what reference point they are using.
What does the competition look like? How many keyword phrases are you going after per page? How much content is there per page. What is the page about (multipurpose) etc. etc.
I have e-commerce pages that are over 20%, I have content sites that are 5%.
5-20% in the SEO world is accurate given the variance of site types and competition, among other easily discernable factors.
[edited by: Chicago at 1:50 am (utc) on April 11, 2003]
| 1:50 am on Apr 11, 2003 (gmt 0)|
Chicago puhleese, now i'm you enemy, ok whatever, bring it on.
20% density means every 5th word, gimme a break.
| 2:04 am on Apr 11, 2003 (gmt 0)|
You have to clarify what parts of the page, or which analyzer you're using to come up with the figure. One analyzer's 25% is another analyzer's 7%.
Analyses that count alt and title tags will always give you a higher density than ones that don't. The sim spider works with visible and alt text but not title tags. One analyzer you'll find with Google gives a good breakdown of density in different parts of the page-- keyworddensity.com
** be sure and check the "raw data output"
If this counts as a URL, mods please delete. I figured tool links were OK.
[edited by: BGumble at 2:06 am (utc) on April 11, 2003]
| 2:05 am on Apr 11, 2003 (gmt 0)|
? Notepad... Chicago gave you a serious answer, and a legitimate one at that. Just a suggestion.. you'll get a lot more mileage out of the folks around here with a bit of courtesy. *smile*
The problem is, as Chicago stated, that you generally can't separate keyword density from other factors in a range of serps well enough to get a definitive answer. And yes, I have had pages at number 1 for a two word phrase with as much as 31% density... I have also had them at the same position with 6%.
If you seriously want to study keyword density then you can do it quite easily for yourself. The trick is that it takes time – you have to build, get em spidered and study the update results…. And you can only do it on your own site. You can’t control enough factors on your competitors sites for their results to mean anything to you.
The best environment is one with probably a sales type of site setup with a fairly rigid structure and minimum variability between pages… say on a widget sales site if you wanted to study the phrase “large widgets”, something like:
Say large-widgets.html links to the following pages:
The blue, red and yellow widgets pages are set so that they have the same template – navigation, page layout, heading markups, etc… the only substitution is the colour of the widgets (to avoid duplication penalties and so you still have a friendly site for visitors!). The page size should be kept roughly the same, but the body text rewritten in each case to include a little more or less usage of the phrase, so that you end up with three (or more) pages that are the same (as much as is practical) in all respects except keyword density, say:
If after the update a search for “large widgets” brings blue up as the first result for YOUR SITE (we’re ignoring all competitors here because we can’t control the other factors on their pages), then you can fairly say that something less than <7.5% density is preferred. Red = 7.5 – 12.5% preferred, and Yellow first indicates a preferred density of >12.5%.
It doesn’t matter whether your page comes up first in the serps or 137th for the purposes of this test… the important factor is which is the first of YOUR pages to show up.
The catch is that this test only tells you what the preferred density is for the current index.. this month only.
On the other hand, you have built yourself a nice little insurance policy for the following months in that you now have your key phrase targeted at a range of different densities to allow for shifts in the algo.
| 2:05 am on Apr 11, 2003 (gmt 0)|
NotePad, actually I prefer not to get personal.
Have you ever optimized an e-commerce site (page)- wherein your content is product driven?
Have you ever optimized a page that has less than 100 words?
Also - have you ever optimized a highly competitive keyword phrase? It is unbelievable what begins to happen to content when a search term is highly competitive.
Many around here will tell you to write for your users and the rest will take care of itself, but for those that do not have that luxury, high keyword density is extremely important and effective and 20% is not odd if a site is structured to handle that. Most of the time that is a site that is dynamically generated, often commerce related. Certainly not an article.
Lastly, don't take the high end of the example and draw sweeping generalizations. That is the right side of the spectrum and certainly not the norm.
It is all relative, notepad. That is my point.
| 2:26 am on Apr 11, 2003 (gmt 0)|
My home page has a kw density of 23% for one of my main terms. This is a content site but the density for this term is high because it is included in the site's title, is the main keyword, and is the name of the area. I'm not worried about this being too high but I am concerned about the two word phrase (full name of location) that occurs 43% of the time. Is that too high or does it not matter with multiple word phrases.
| 2:28 am on Apr 11, 2003 (gmt 0)|
I'm sorry Chicago, it wasn't my intent to make anyone mad, i was out of line, please accept my appology.
On my message board things get wild and wooly, i guess i'm used to that...again, please accept my appology :)
| 2:32 am on Apr 11, 2003 (gmt 0)|
lol the keyword density for "the" on my site is only 5%
| 2:34 am on Apr 11, 2003 (gmt 0)|
>Have you ever optimized a page that has less than 100 words?
I think here is the flaw in NotePad's logic. I presume that he is thinking in terms of pages that are designed to have human readable sentences. However, consider the possibility of a page that basically is a menu with links to other pages. Example:
Page title: About Widgets
Which is just basically a page of links to other pages. Such as:
What are widgets?
Widgets in the 1600s
Widgets in the 1700s
Widgets in the 1800s
Widgets in the 1900s
Widgets in the 2000s
Uses for widgets
Different types of widgets
Where to buy widgets:
It isn't hard with a page like the above to get huge keyword density, and still have a page that doesn't look ridiculous to humans. And, knowing how to create pages like the above is part of the art of search engine optimization.
| 2:44 am on Apr 11, 2003 (gmt 0)|
Browser visable text on index page 4% key words (actually, just about this - don't want to give absolute figure openly)
Html text on index page - same
But the real ranking page is your second page. So, if your index page is www.domain.com/, then you want www.doamin.com/blue-widgets.html to have
Browser visable text 7%
Html text on page 4.5%
(same, these are near enough, don't want to give the exact % I found as it took me a long time to write the script that did this).
I actually ran a backward calc script on different sites from different categories, and it was amazing. They all had the exact same percentages (well, around 60% did).
You can try it yourself. Just be careful trying to compare like for like though. Look for new sites appearing, not old ones. Factor in the backward links they have (and don't just accept the ones that Google declares - just search for the domain without the link: and you will find the PR3 and lower sites that link to them as well - many of these, if they only link to the target site, will have a huge influence, epsecially if they are relevant sites).
Also, I have seen some sites in competitive areas that have flash front ends, and perhaps a single <!-- key word --> statement, nothing else. I guess Google have nothing else to go on but that.
| 3:26 am on Apr 11, 2003 (gmt 0)|
Perhaps I'm out on a limb on this, but I don't fret about kw density anymore. IMO, density has been abused and consequently devalued in the algo. Yes, some top scorers in the serps are density freaks, but they are there not because of density, but for other algo-related reasons despite their KW density overkill.
I concentrate on how the keywords are used. My m.o. (apart from seeding the page/site with the actual words) is to:
1>Verify that the page isn't accidentally optimized for another phrase (watch ALL the repetitions, not just the targeted phrase).
2>Watch the plurals
3-? >Other considerations.
Remember, there are around 100 aspects to the Google algo. IMO, KW density is a thing of the past and isn't that important anymore.
Feel free to criticize.
| 6:00 pm on Apr 11, 2003 (gmt 0)|
Ah, at last... Martini,
I recently discovered the expresso martini- what a treat!
Martini writes "Remember, there are around 100 aspects to the Google algo. IMO, KW density is a thing of the past and isn't that important anymore."
Of the 100 aspects you address, the overwhelming majority of them are a function of words - keywords. When a page/site is spidered all it has to look at are words. Words make search engines go round. The spider finds words in all different places- in metas, in headers, in bodies, in anchor text, on ones site, in url's, on other's sites, etc. The key is to carefully spread your keywords throughout all the areas that a spider reads. Consistency and comprehensiveness in the aforemention areas is critical. The idea of density is a comfounding one, that should not be read into too much. Density is simply a measure of the # of keyword occurances relative to OTHER words as well as a competitive measure. In and of itself, density is not a tactic or aspect of the G algo. One should strive for comprehensivness and consistency in areas a spider reads, not density.
| 6:36 pm on Apr 11, 2003 (gmt 0)|
I never did a page with a kw density higher than ~20%. However, within the last six months i observed some competitors climbing higher and higher with quite high kw density - up to 50%.
Now i did a redesign of a site and first time i built a density of up to 35% - the site is fresh indexed and gets good "fresh featured" listings - but not (yet) for the 35% keywords... however, it's too early to see any difference to lower density.
IMHO there's no general rule. It's the mix of various factors that counts and i don't expect to win *only* because of the kw density.
| 12:53 am on Apr 12, 2003 (gmt 0)|
I know other factors (links to you, and internal links) are VERY important. Also as to where you put your key words (a little here, a little there, a little, why, everywhere!). But I went to great lengths to isolate those sites that had achieved high rankings based on keyword density. Yes, they of course followed the "a little there rule", but it was "WOW". I kid you not, I picked different areas, and they all came back with the same density.
The trick for me was to look at the other factors and avoid sites which had pole position due to links over all else, and pick the sites that were doing well based on density only. It is not easy, but it is, especially now with the new results. Pick the ones that come into the top pages on Google yet have little inbound links. You will see if you take as long as I did :-)
| 10:15 pm on Apr 12, 2003 (gmt 0)|
Brett, If you read this could you let us know which tool you used and if you included meta tags in your suggested 5-20%? I have wondered about this for some time.
One thing I learned by trying this and that is that it's important to have the keyword in title tags. Also a short title is better. Someone mentioned this on this forum so I shortened my title to 3 words including the key word and it gave me a good boost in that word in the serps. I wonder if a one word title with the keyword being that title would be even better. But that would make my title kind of meaningless and I believe people do look for interesting titles as well.
Also having the word in your H1 tags is helpful (usually that would just repeat your title). I am just trying adding the word in my H2 phrase as well.
I can really see that trying to get too many keywords on a page is not productive. Different keywords emphasized on different pages is the key.
| 12:41 am on Apr 13, 2003 (gmt 0)|
I have MANY pages with near 100% kw density.
thwe whole site is basically the interface to a database (directory).
So some pages may only contain a name (the kewords) in title headign and body, and perhaps an address and/or tel number. Or it could be a listign for categories, each of which is a kw in anchor text...
So basically these pages are pure keywords, even though they look perfactly natural to human visitors and perfectly fullfill their purpose.
It's content vs commerce vs information sites really...
| 2:16 am on Apr 13, 2003 (gmt 0)|
18.35% - #1 all across the board, not just Google. You decide. ;)
| 2:37 am on Apr 13, 2003 (gmt 0)|
Marcia, could you give a more detailed breakdown, maybe like the pertinent info from Webjectives' analyzer (keyword.htm)?
Here is my new one awaiting May Google update:
There are 374 words in the body.
Frequency in the body = 35
Weighted Keyword Density in the body= 9.36%
There are 1 words in the Heading tags.
Frequency in the Heading tags = 1
Weighted Keyword Density in the Heading tag= 100%
There are 26 words in the alt tags.
Frequency in the alt tags = 15
Weighted Keyword Density in the alt tag= 57.69%
A HREF Tag Link Text:
There are 131 word(s) in the href tags.
Frequency in the HREF tags = 14
Weighted Keyword Density in the HREF tag= 10.69%
A HREF Tag URL:
Frequency in the href tags URL = 2
HTML Comment Tag:
There are 9 word(s) in the comment tags.
Frequency in the comment tags = 3
Weighted Keyword Density in the comment tag= 33.33%
There are 412 word(s) in the entire page
Frequency for the page = 81
Weighted Keyword Density for the page= 19.66%
| 2:43 am on Apr 13, 2003 (gmt 0)|
Marcia, i checked out your site and did several searches in you genre and didnt see your site at all..don't get me wrong, i'm just tring to figure out what keywords you are talking about :dunno
| 3:18 am on Apr 13, 2003 (gmt 0)|
>18.35% - #1 all across the board, not just Google. You decide. ;)
Depends what this single keyword is. Give me a word that occurs infrequently on the WWW, and is also one that nobody else out there is competing much for, and dominating the SERPs would be easy. Hmm...getting #1 for "Pectenodoris" should be a snap. ;)
| 4:46 pm on Apr 13, 2003 (gmt 0)|
Can anyone say, "Analysis Paralysis"?
Let me see...
keyword density for "analysis"...20%
keyword density for "paralysis"...20%
keyword density for "analysis paralysis"...hmmm, not sure how to figure that one out...is that 20% or 40% or something else.
Tomorrow, I think I'll put a team on that one!
| 6:27 pm on Apr 13, 2003 (gmt 0)|
As some of you allready may have noticed, i'm quite frustrated with the improvement of keyword stuffed pages over "normal" pages in the new index. Today i did a lot of analyzes and came to the conclusion that depending on various other factors like crosslinking, internal link and site navigation structure, a high keyword density expecially keywords and phrases repeated in alt texts of image links can boost a site. Even if the alt texted images only link internally. The more you use the better the results. I don't mean 5 images with 100 keywords alt text each but 25 images all using one single keyword or max three word phrase linking to the same internal page.
I still don't want to jump on that train but i'm asking myself: if i don't cheat with content cloaking or other dodgy stuff and the people exactly find at my site what they where searching for at google, is it a low or a high risk to increase density, link anchors and alt texts and play the same game as the current (or returning) winners?
(Sure my site is high quality and worth to be #1 within my theme ... ;)
| 7:09 pm on Apr 13, 2003 (gmt 0)|
> frustrated with the improvement of keyword stuffed pages over "normal" pages in the new index.
Yes, agreed. But regarding alt tags specifically, it was, as we all know, more or less "SEO 101" several years ago that people should use keywords in their alt tags. The SE's counted them, and essentially those who didn't use them put their sites at a disadvantage as far as the SE's go. I'm not sure I see this as spamming. (This is more akin to pro athletes arguing calls even when they know they are wrong...it's a part of the game, and very transparent.)
Plus, the SE's can, and should, deflate the importance of alt tags to a great extent via their algo's. But to not use the alt tags seems to me to be shooting oneself in the foot.
At least in our camp, we make a huge distinction between keywords in alt tags (OK), versus hidden text, cloaked pages, and other dastardly tricks (NOT OK).
When it comes to what we can control on own sites, doesn't it mainly come down to a few simple rules:
--use cross linking and anchor text appropriately;
--use titles, meta tags, alt tags that reflect your content;
--don't do anything that you would have trouble explaining to a SE review board!
| 7:40 pm on Apr 13, 2003 (gmt 0)|
|But to not use the alt tags seems to me to be shooting oneself in the foot. |
With great respect for your opinions, I have to disagree: that is a gross overstatement of the importance of alt tags. In fact, if there is any weight given, it is of such insignificance I rarely pay attention to it anymore. (My client's sites do exceptionally well in the top five for all important phrases).
There is a certain keyword tracking web site (I won't mention the company's name) whose alt tags are stuffed harder than a holiday turkey. An seo job analogous to slathering oneself with musk oil to attract the opposite sex: Desperate and Counterproductive.
The upshot is that it doesn't work. Makes no difference to their rankings. Even worse, their sloppy effort has ruined the surfing experience of any sight impaired visitors to their web site.
The key thing to remember is that it is counterproductive to try to "trick" the search engines (comment tag stuffing being among the most futile and useless tricks among them...).
The correct way, in regard to keywords, is to structure your doc properly, and in a manner that is helpful to the sight impaired surfer.
| This 37 message thread spans 2 pages: 37 (  2 ) > > |