Forum Moderators: open
Should I test for competion using exact match or not? I read somewhere here that high PR pages will still rank higher even if they don't have an exact match. From this I take it that I should test for competition without using quotes, but everything I'm searching on is very very competitive, hundreds of thousands of pages. Four word phrases I'm trying seem to have little or no searches being run on them.
Hope this isn't a dumb newbie question, but then I am a dumb newbie at keyword selection.
Thanks for any advice,
Paul
It is true that having an exact match doesn’t mean you are ranked above all pages that don’t have an exact match. Yes, PR is one reason for this but there are many others like anchor text, keyword density, keywords in title, etc.
As an example, assuming the same number of searches and the same relevancy, I would target a keyword with 100 exact matches and 2000 matches before targeting a keyword with 200 exact matches and 1000 matches. However, if exact matches were the same then I would take the keyword with fewer matches.
Often you see people say they rank number 1 for a phrase out of several million pages and think they are fantastic (and sometimes they are), but more often than not the search has been made using one of the generic keywords (often the word "free" and as soon as you include the word free you add several million competing sites but not related to your specific phrase. When you search for the phrase using quotes it narrows down to the site being top out of 5 or 10, which tends to make us think the keyword is not searched for enough and using Wordtracker often confirms this.
Each PPC has got different rules relating to plurals and because it's PPC they only offer exact matching, with Google you can use square brackets for exact matching [keyword combo] - only shows with those words in that order, using quotes "keyword combo" - will show for "keyword combo", but also "combo keyword" or with other words too, choose these with caution or you'll rack up a big bill in no time.
You can't really compare unpaid traffic with paid. On paid ads you can chose the ad title and description aligned with deep content linking and the position. Competitive keywords doesn't translate to lots of searches.
I am not able to follow your post. Perhaps you are talking about options that are available on some PPC services.
I believe ppg/Paul was finding search frequency with Google Adwords and Overture (perhaps he should also try Wordtracker) and then he was looking at the plain Google SERPs to analyze the competition. In this case a query with quotes can find the number of pages with an exact match of the keyword phrase. A query without quotes can find the number of pages that have each individual word somewhere within the page. For example I just conducted the following searches in Google.
“blue widgets” – 519 matches
“widgets blue” – 101 matches
blue widgets – 27300 matches
free blue widgets – 12600 matches
The query “blue widgets” is different from “widgets blue”. Also, -- blue widgets -- has fewer matches than -- free blue widgets.
My earlier post about competition analysis is unrelated to PPC search engines like Overture. Competition in Overture is represented by the bid amounts and not Web Page matches.
gmoney - yes, thats exactly what I was doing. I've been messing around with wordtracker and it looks great, but I've just been using the free trial and I'm reluctant to put too much emphasis on statistics that only relate to 3% or so of searches. Time to pay up I spose.
What was worrying me was that perhaps it wasn't worth targetting keywords with huge amounts of results for a non-exact match, but all the phrases I was trying did have lots of competition for non-exact matches. From your post I take it that analysing exact match searches is valuable too. Thats a relief. With non-exact matches I was having trouble finding any keyword phrases that looked worth going for.
That said, the one three keyword phrase I've tried targetting so far has got to me to position 6 out of 26,800 with a non-exact match search. If I'd tested this phrase out first as I've been doing, with that number of results for a non-exact match I would have assumed it wasn't worth going for.
I'm not doing any PPC stuff yet, just targetting google to make sure I can make this stuff work before I go any further.
Thanks again for clearing things up for me.
I'm using the paid version of wordtracker now. I wonder if someone could shed any light on the following for me:
I've found a 4 word phrase which is relevant to an area of the site I work on.
If I run it through wordtracker (just getting results for google) with an exact match, I get this:
KEI: 180
count: 30
competing: 5
without quotes, I get this:
KEI: 0.002
count: 30
competing: 588000
None of the top pages in google's results have an exact title or <h1> match for the phrase for the non-exact match.
just in terms of PR, the top few results for the non-exact match have similar or lower PR than me (3 or 4).I know now from what gmoney posted that PR isn't the only factor to take into account, but its one of the main ones, yes?
So given this, is it true that I would stand a fair chance of getting on the first page for that phrase with a non-exact match despite the amount of competition and abysmal KEI?
Three of the words in the phrase are pretty common, not as bad as say 'free' or 'products' but pretty common.
In this months update, I've had some success with getting in the top 10 against what looks to me like insurmountable odds - 7th out of 344,000 for a 3 word phrase, and the best one: top out of 2,240,000 for another 3 word phrase. Other ones I havn't done as well on, but ok - 2nd or 3rd page.
This seems to have been acheived mainly through exact matches for the phrases in the title, header tag and in the content.
I wasn't doing any keyword research for these pages, with the result that I now find that no-one seems to do searches for these phrases at all, so what looks like great results is actually pretty useless. I guess you could say I was mainly trying out the theory to see if it worked. So now my job is to do the same thing for next month with keywords that people actually search on.
Given the above results, is it possible to give any indication based on the above numbers that this phrase is worth going for? Or are there too many other things to take into account to say for sure?
Thanks for your patience. I seem to be having a bit of trouble getting my head round this. I'm much more used to technical aspects than any kind of marketing, including SEO.
“I've found a 4 word phrase which is relevant to an area of the site I work on.
. . .
competing: 5 (exact matches)
. . .
None of the top pages in google's results have an exact title or <h1> match for the phrase for the non-exact match.
. . .
just in terms of PR, the top few results for the non-exact match have similar or lower PR than me” – ppg
I’d say you have an excellent chance of getting on the first page for that keyword.
”Three of the words in the phrase are pretty common, not as bad as say 'free' or 'products' but pretty common.
. . .
competing: 588000 (word matches)” – ppg
Minor setbacks, but you have shown that it doesn’t take much for pages to rise above these setbacks.
-----------------------------------------
I am always happy to find keywords with similar credentials to the one you talked about and I find it worth the effort to incorporate them in my site. Sure a count of 30 isn’t much to write home about but a number of keywords like this one can generate some noticeable traffic.
By the way, is there a way to get Wordtracker to give you the 588000 number or did you have to hand pick it out of the Google SERP? I usually don’t incorporate the “word matches” in my keyword analysis but if it were readily available then I might include it.
If you like I'll let you know how I get on with that phrase next month.
regarding the figure of 588000 for a non-exact match search on google - in wordtracker I used the 'keyword universe' section of the menu to do my research. When I'd got the words in my 'basket' i just clicked on the 'competing' graphic, and it gave me those results. Theres a two option drop down for selecting whether to do the test for competing sites with or without quotes, thats where I got the figure from. It matched exactly with google's results.
Is this what you meant?
Oh, I thought of an example related to low count numbers like 30. Anyway there is a keyword that I am only on the second page of Google. The keyword has a count of 9 (it was 6 a couple of months ago) but yet I received about 5 to 10 referrals for it each month from Google alone. This is much more that I would expect and it is kind of a freak example, but things like these do inspire me to give low count keywords a chance (as long as there isn’t much competition).
I’d be curious to hear how your keyword works out for you next month. I’m guessing that you will be on the first page. I’d also be curious to hear how the referrals pan out.
