Phrase vs keyword

[PunkJazz]

I am sure this has been covered but I haven't found the answer in search. I want to optimize a page for 2 specific keywords/phrases - "widget" and "widget collecting". If I build a page with say 15% keyword density for each phrase will google pick on the widget in the phrase "widget collecting" and see it as 30% density and penalize me for keyword spamming or treat it as a unique phrase and be ok with it?

[futureX]

obviously to optimise the phrase 'widget collecting' to 15% would make your 'widget' density to 15% also :P

[martinibuster]

I am sure this has been covered but...

This has been going on for the last four days: In the Google News forum (which is where you posted your question).

If you have a question, scroll down through the forum you are posting in. You'd be surprised at what you'll find.

[webmasterworld.com ]

[PunkJazz]

[webmasterworld.com...] is a discussion about optimal density and does not address my question. Someone does raise my question in that thread but it is ignored, no one replies.

Can anyone give a straight answer to this question? The "obvious" answer above makes no sense, given how google treats single keywords vs. phrases in returning rearch results, which is why I am wondering about this.

[heini]

Punkjazz, I think futurex has the answer here.
In a sentence containing the phrase widget collecting you have covered widget as well as widget collecting.
Your page would be found under widget as well as under widget collecting. The density would be the same for both phrases.
Also it's important to note that keyword density doesn't mean much in Google's algo. Much mroe importance has the anchor text on links, internally and externally, leading to that page. Also much more important is what you use in titles, and bolded.

[deejay]

Hi Punkjazz

First, have you used Brett's keyword density analyzer at www.searchengineworld.com? I highly recommend it.

Now to your question.. I think I can see where you're coming from.

(let's ignore stopwords for the moment.. that's a whole 'nother kettle of fish)

If you have a page of 100 words, and insert the phrase 'widget collecting' five times, you'll obviously have 5% density for 'widget'... 5 occurrences out of 100 words.

'widget collecting' x 5 = 10 words though, so you are concerned you will have a density of 10%... 10 words out of of 100.

In short: I don't know. Logic says to me that if the search engines do consider that to be 10% density, then they have to have varying levels of 'ideal' density according to how many words in the phrase.

It seems much easier to do what Brett has done with the keyword density analyser, which would allow them to set one value as ideal, regardless of phrase length. I've just run a page through, and the results look something like this:

Recurrent single words that aren't stopwords: 57
Occurrences of those words: 251
widget = 34 occurrences = 34/251 = 13.55%

Recurrent two word phrases that aren't stopwords: 25
Occurrences of these phrases: 80
widget shop = 8 occurrences = 8/80 = 10%

etc...

See how that works?

1 Ignore stopwords.
2 Count recurrent phrases
3 Calculate occurences of phrase against total phrase occurences.

The rub is that we don't know how the SE's do it, and we're not going to know because they're not going to tell.

To complicate matters, you're probably not going to find two density analysers that do it quite the same way either.

I suggest you find a density analyser that you like/are comfortable with, and stick with it. At least then you are getting consistent results and from there you will start to get a feeling for what's working.