C:\Documents and Settings\FEIYA.FEIYA-CI5XO0GKI\amaya\2\New.html

http://forums.searchenginewatch.com/showthread.php?p=111740#post111740

RobBothan:

ok with density is it best to have a list of keywords for your widgetwebsite like:

widgets, small widgets, big widgets, blue widgets etc

or...

widgets, small, big, blue, etc

mainly for use in meta tags and titles

beu:

Keyword density is not important. What is important is using keyword termsfor which users search. After you determine those terms, then you shouldinclude those terms in the TITLE, meta data and H "tags".

RobBothan:

i am aware of the differing levels of importance of those parts of a page- however what i mean is whether google et al. have recomendations on keywordrepetition/density with regards to the words in titles/tags -

eg for title pages - "blue big widget - big widgets - the widgetcompany"

meta tags - "big widgets, blue widgets, the widget company"

mcanerin:

Most professional SEO's hate "keyword density".

Yes, of course you have to use the term often enough on your page so thatthe search engine will know that the page is about the subject. But there isno ideal keyword density (KWD).

First off, pages are compared to each other based on the search term. Sothat that means that the KWD for a page is actually based on the term. It's alot easier to use the search term "Javascript" on a page more often than"Floccinaucinihilipilification". There are some terms that look strange whenused more than once or twice on a page, whereas you can use others a coupledozen times without too much of a problem. So the "ideal" KWD for "RealEstate" is based on a comparison to other sites that contain that term, notsome magical specific number.

Second, it's the goal of a search engine to provide the best results tovisitors, not to reward webmasters points, extra lives and bananas for"playing the game" well. This means that there is no specific score or magickey to aim for. In the old days, if you used a keyword one more time on apage than your nearest competitor, then you would "win" and be placed abovethem in rankings.

The result was people typing in keywords on a page dozens of times - everytime their competitor typed in another keyword, they would type in 2, and soforth. This creates stupid, useless pages and thankfully the search engineshave been putting a stop to it over the years.

This is one reason why links are so important - they help deal with theKWD arms race by giving search engines another method for judging pages.

The search engines do pay attention to keywords on pages. But not just tohow many times they are used. Also, they look at where they are used, whatsupporting words are present, what the surrounding content is about, and soforth.

My suggestion would be to use the keywords as often as makes sense on thepage for the topic, no more or less.

the search engines don't give that kind of advice. They try to rewardpeople who make webpages naturally and for visitors, not people who usetemplates and rules intended only for achieving rankings. Having said that,there are some generally accepted good practices.

Since it's KW specific, the best way would be to test it for your niche.In general, if you think you are using them too often, you probably are.Titles, for example, should be short and focused, not long strings ofkeywords. It's not necessary to mention a keyword more than once in a keywordmetatag (it's not actually necessary to even have a keyword metatag).

RobBothan:

Again I dislike the keyword density approach but was starting to worryabout a page I had that in the natural copy ended up sounding slightlyrepetitive, and wanted to check that i wouldnt be penalised for this, i'mglad my approach of standard joe to copy works! thanks for the explanationand i'll keep on learnin!

rob

borgie:

I am not sure if I am following what everyone is saying about KWD. Havingthe keyword phrase you want to win for that page in the title is veryimportant. Using that phrase on the page is also very important. I usemultiple tools to measure the number of times a keyword and keyword phraseare used on the page. There are tools that measure single keyword density(http://www.webconfs.com/) on a page and phrases (Long Tail Keyword Cloud )....

You have to tell a search engine what your page is about and you do thisby putting it in your title and in the body text.....which is measuredthrough Keyword Density. I have read that the sweet spot for keyword densityis between 4%-7%....this way you are not spamming the search enignes but aretellin gthe engines what the page is about...I guess I am not sure why peopleare opposed to KWD? I believe this is one of the major on page factors thatneeds to be execute successfully.

mcanerin:

You've been reading old articles, borgie - it's a common issue. You arenot wrong, but you are not right, either. Let me break it down:

1. Just so everyone is on the same page, KWD does not refer to havingkeywords on your page. It refers to a certain percentage of a keywordcompared to the content of the page itself. If you have a 100 words and 3 ofthem are your keyword, then you have a KWD of 3% for that keyword. Naturally,you can have many different words on a page, and each would have it's ownKWD.

2. THERE IS NO SWEET SPOT. There are a lot of articles out there claimingthat there is a sweet spot somewhere around 3-7% To my great shame, I wroteone of them, back a few years ago when I was dumber..er, less educated than Iam now. Even back then, I noted that it depends on the keyword. KWD ofupwards of 15% is common in real estate, for example. For some scientificterms, more than a couple of instances on a page is considered too much.

3. KWD today is far more complicated than this. Think about it. If itreally worked like this, then everyone would have a KWD of 3-7% and then whatwould the search engines do? Further, if I take an article about rabbitbreeding and then stuff in the term "real estate" so it reaches a certainKWD, does it suddenly become an article about real estate instead of rabbitbreeding? Of course not. That's why search engines don't use, and have neverused, KWD.

4. KWD is a concept made up by SEO's, not search (Information Retrieval)scientists. IR scientists (Google has almost 80 of them) use Term VectorAnalysis. There are many types of TVA, but they all boil down to trying tounderstand what a document is about. Technically KWD is a form of TVA, butit's the most primitive form. It's certainly not used by any modern searchengine. Latent Semantic Indexing (LSI) is another type, though I'm notpersonally aware of any search engine that uses full LSI. It works way betterthan KWD, though!

5. One interesting thing about KWD is that, up to a certain point, it canhelp a site. It's going about it the wrong way, but in general addingrelevant keywords to your site can often help your rankings. But this issomewhat like stopping your car by driving it into the nearest wall orturning off the engine, instead of using the brakes - it works, kind of, butthat doesn't mean you are doing it right. It would be far better to ignoresilly guidelines and just put as many instances of a keyword on a page as isreasonable. This will sometimes be very few, and sometimes will be a lot.Guess what? It will work, even if it's more or less than 3-7%.

6. One final thought, and the spike through the heart of KWD. KWD issingle keyword-centric. It teaches people that they need to stuff more orless of a specific keyword on a page. But TVA (which is what the searchengines use) is about measuring the *relationships* between words on a page.That means that in order to optimize for "real estate", you may need tomention the term "city" nearby it. Or "agent", or other terms. In general,people focused on KWD of one term end up lowering the expected KWD for thesupporting terms, making their page less focused and more like spam. Inshort, it's a good way to lose the forest because you are concentrating ononly one tree. I have on many occasions lowered the KWD of the main keywordon a page while adding in more natural text, which raised the rankingssignificantly.

Yeah, you need keywords. But you need the right amount in the right place,with the right number and types of supporting words. KWD doesn't do that, andcan't do that. Natural writing with a focus on a keyword can, by it's verynature. Search engines find those supporting keywords by indexing hugenumbers of documents that are results for the original keyword. They are thenable to extract which additional words are used and where, as an average.Since most of those documents were written naturally, the data is based onnatural usage. Once again, KWD doesn't do that.

Forget KWD. Forget SEO by-the-numbers. It's about content and focus foryour pages, and trust for your site. It always has been.

borgie:

1. Totally agree, it is the density of that keyword or keyword phrases onthat page.

2. Is that the exact sweet spot? No is it a safe guideline to optimizeby...yes. Is 40% keywords to much? Yes! Is .5% keywords to little...Yes!...just was giving a safe recommendation to live within...of course everyindustry is different but just because it is common doesnt mean it is right...correct?

3. Of course there are many other factors that go into ranking a web site.On page factors are only facet of the game....but it is one you can control.Any the example of rabbit breeding and real estate is about as off as my .5%- 40% KWD ....there are certain things you would assume that the topic youare writing about is about those keywords you are writing. ...this goes backto the many different elements that go into ranking a site.

4. Exactly, they all boil down to what a document is about. Using KWD as atool to help you to make sure you are somewhat on target...is it the answer?NO...not at all...but it given you a high level overview of what the articlemay be about. Of course none of us have tools or the resources that Googledoes to know the exact factors......or if you do ...do you mind telling me?;-)

5. I liked your example in this one...it made me lol! I totallyagree...dont keyword stuff but dont forget to use the keyword phrase orphrases when you can in your content.

6. One of the tools I posted earlier does 2 and 3 (long tail) KWD phraseswhich addresses the "supporting words" and helps you focus on those long tailkeywords. Exactly what you are saying...using the keywords together is veryimportant.

When you say its about content and focus for your pages....what do youmean?

maybe you dont agree with me but all i am saying is KWD is just a tool(elementary as it may be...but yet to find better ones...please post if youknow of better tools...paid or free) to use to look from a high level at whatyour page is about....

Thanks for taking the time to write up a good post!!

bue:

Content is about providing unique and/or valuable information to users.When you do that, folks link to your site and your site gains importance inthe "eyes" of search engines.

Focus is providing enough related information so users as well as searchengines can determine that your football site is about NFL football and notFIFA football.

The problem with these tools is what I consider their ability to determine"context"!

Take two pages, one with a higher keyword density and one with a lowerkeyword density. The page with the higher keyword density should rank higherthan the page with the lower keyword density, right?

Wrong, the page with the lower keyword density has the "keyword term"being measured, not only in the URL, the TITLE and meta description but alsoin the first H1 on the page, bolded in text content as well as in an ALTlower on the page.

Point, keyword density tools only see the density of keywords on a pagethey don't see how those keywords are used on the page.

Borgie:

I was under the assumption that people were putting the keyword in thetitle, h1 and in the first couple of sentences....i guess no assumption issafe tho...since that is basic blocking and tackling.

mcanerin:

All other things being equal (and they never are), the following scenariosare also true:

1. A page with the keywords in the top half will outrank one with keywordsin the bottom half.

2. A page with term vectors near the keyword will outrank a page with termvectors further away.

Neither of these are identified by KWD.

Here is the other issue: lets say you can measure the KWD of a keyword ona page. Big deal. You still have no idea what the appropriate keyword densityshould be. Oh, sure, you can grab the top 10, 30 or 50 results and do a KWDcheck on them, as well, but that really isn't going to help much for a coupleof reasons. First, the top ten are usually there for reasons other than KWD(another reason KWD is not a part of a professionals toolkit - we like to bein the top 10). Second the effect of linking and anchor text is so strongthat the rest of the results are suspect, as well.

Measuring the KWD of a site that got to the top 10 because of it's anchortext is like deciding on the winner of a car race based on it's paint job.It's entirely possible for every single site in the top 30 to have "bad" KWDbut have great anchor text and link love. The system is set up that way.

So now you are using bad data to arrive at a fake percentage to use auseless tool to tell you how many keywords to artificially stuff into adocument.

But hey, because there are actual numbers and percentages involved, itmust be accurate, right?

See what I'm getting it? In computers, there is a concept of "artificialprecision", which is closely associated with "GIGO", or Garbage In - GarbageOut". A computer will happily take a couple of guesses, crunch them with badlogic processes placed there by a programmer, and give back a result to asmany decimal places as you want. The fact that there are decimals andpercentages and numbers spewing out of it does not mean any of them are inanyway accurate.

The last time I had to explain this was to someone who had developed theirown KWD tool and was trying to promote it in the forums. They didn't get it,either.

The problem is that search engines tend to use fuzzy logic andprobabilistic systems across a huge raw dataset, whereas tools like KWDanalyzers try to come up with a precise average taken from an alreadyprocessed set of results. Most KWD analyzers don't even bother to give youthe range of results ("the top 10 had a KWD of between 2-13%") which would bemore useful and quickly educate users just how inaccurate they are.

Maybe that result at the upper end was closer to the "best" KWD and theothers were all there due to link weight. In this case, the KWD tool istelling you to go much lower than you actually should be going because theaverage is lower.

A search engine is not a big calculator. It does not give you 10 pointsfor a title, subtract 2 points for not having a metatag, then add 15 pointsfor link weights and come up with a number. For some reason, a lot of peoplethink this is how it works.

It actually processes things in several steps - grabbing a large relatedset first and doing an initial sort, then grabbing a subsection of that anddoing another sort, and so forth. At each stage, it's looking for somethingdifferent - removing spam, looking for sites that are related to each other,identifying "types" of sites, etc. They try to preprocess as much as they can(the Supplemental Index is an obvious example) but at the end of the day thedata is processed several ways for several reasons before it finally getsdisplayed. The final result has little to do with KWD. KWD gets you throughthe first sorting stage (relevance) only.

Finally, let me say this - I used to use a KWD tool when I first started out as a full time SEO. Now I don't. When I first started out, I was making maybe 10K a year. Now I'm making 250K+. Compared to most professional SEO's , I'm severely underpaid (I've been told several times). Do you really think that I would have dumped KWD tools if I thought they worked? Would I have taken a chance on losing my income over it? At the end of the day, "Money talks, bull**** walks".

KWD tools don't get results for hard sites. They only get results for sites with SEO's who don't understand search engines and need a crutch. They can be a good initial training tool for teaching people to think about keywords. But they won't take your site to the next level, and they are certainly not the mark of a professional. It's like putting training wheels on your racing bike and thinking it's helping you. It will help - at first, when you fall over a lot and have a hard time remembering to steer and pedal at the same time. But then they just hold you back and limit your options. </p>