|
View previous topic
::
View next topic
|
| Author |
Message |
Phil Tanny
Joined:
22 Aug 2003
Posts:
1324
Location:
Gainesville Florida USA
|
Posted: Fri Nov 19, 2004 7:54 pm
Post subject: overture search counts?
|
 |
|
Hi all,
Can you offer opinions on the quality of data Overture provides when it gives you the number of searches for a keyword?
http://inventory.overture.com/d/searchinventory/suggestion/
Do the Overture numbers give a reasonable comparison between one term and another?
In other words, will both Overture and Wordtracker show that Search Phrase A is about twice as popular as Search Phrase B, whatever the numbers each service might use?
I'm programming a keyword research tool and need to be able to write code that can find the number of searches for a keyword.
The number doesn't have to be exactly accurate, but the numbers should fairly accurately describe the relationship between one keyword and another.
Thanks in advance for any feedback.
_________________
Free Forum And Ezine Hosting
http://Engage-Engine.com |
|
| Back to top |
   |
 |
esteneker
Joined:
28 Sep 2003
Posts:
258
Location:
The Netherlands
|
Posted: Fri Nov 19, 2004 8:57 pm
Post subject:
|
|
|
Phil,
I think the consensus is that the figures are inflated, especially because plurals are also shown in the single terms count.
Relatively, however, I'd guess that your right.
Hope this helps.
Erwin.
_________________
Customer Service Point
Family Threats |
|
| Back to top |
|
|
Charlie
Joined:
22 Aug 2003
Posts:
3305
Location:
UK
|
Posted: Sat Nov 20, 2004 12:24 pm
Post subject:
|
|
|
| esteneker wrote: | | I think the consensus is that the figures are inflated, especially because plurals are also shown in the single terms count. |
There are a few reasons. Planet Ocean did a pretty scathing article on this a few months ago. According to them, Overture counts terms searched with their partners' sites more than once.
One thing's for sure though, the numbers are certainly way high, from my experience.
| Quote: | | Relatively, however, I'd guess that your right. |
So would I, but I still wouldn't use Overture's numbers as a basis if I could, Phil.
You can stress "relative" till the cows come home, but judging by the number of times people ask questions about SBI Manager (based on Overture), many people will still read "absolute".
Just a thought,
Charlie.
_________________
"Before I speak, I have something important to say."
- Groucho Marx |
|
| Back to top |
|
|
Phil Tanny
Joined:
22 Aug 2003
Posts:
1324
Location:
Gainesville Florida USA
|
Posted: Sat Nov 20, 2004 12:55 pm
Post subject:
|
|
|
Thanks for the input guys, much appreciated
I tried to do a manual test of counts for the same keyword at both WT and Overture. It's slow going, it's rather difficult to get a count for a specific word via WT free trial.
As discussed, the counts at Overture are much higher.
It's way to early to draw any conclusions as I've only tested a dozen or so words. But in that tiny test I couldn't see any relative match up between the WT and Overture numbers. The numbers seem to have no relation to each other at all. Again, this is based on a very small test.
The only appeal of Overtures numbers is that they are easy to grab. I'd be happy to use other "number of searches" data if I knew how to get it.
It looks like WT somehow gets millions of dogpile queries and then counts them in house.
BTW, the point of all this is I'm trying to code up a keyword finder/analyzer that is better than WT. "Better" to me is a much simpler interface to more useful data.
Getting the keywords is pretty easy. Building an easier to understand interface is REALLY easy. 
And Sean Burn's formula for analyzing keywords seems much better than KEI to me. Sean's formula analyzes what matters, the top 10 listings, not millions of pages like KEI.
In order to run Sean's calculation on the keywords I need to grab the needed data from somewhere, and so far the "number of searches" part seems the most challenging.
Should you know of other sources of "number of searches" data many thanks in advance for a pointer.
_________________
Free Forum And Ezine Hosting
http://Engage-Engine.com |
|
| Back to top |
|
|
Debs
Joined:
16 Aug 2003
Posts:
4296
Location:
NY
|
Posted: Sat Nov 20, 2004 4:42 pm
Post subject:
|
|
|
You can't compare the free wordtracker tool and overture ... the free tool is based on MSN searches only ... Overture searches cover a lot of sites where their ads appear. I would think you would have to get the paid version of WT, and add up the searches made on the overture partner sites to even come close, and that's even if the partner sites are all listed in WT.
Debs
_________________
Learn how to turn keyphrases into quality, well-targeted articles your visitors and SE's will love with Gary Antosh's new ebook "Web Content Made Easy!" |
|
| Back to top |
|
|
AllanGardyne
Site Admin
Joined:
02 Jul 2003
Posts:
6326
Location:
by the beach, Australia
|
Posted: Sun Nov 21, 2004 5:12 am
Post subject:
|
|
|
Phil, I hope this doesn't offend you, but you seem to be just playing at this. I recommend you get hold of the Planet Ocean article and subscribe to Wordtracker. You can subscribe to Wordtracker very cheaply for one day, long enough to do a LOT of tests.
Some tools, such as Keywords Analyzer, allow users to paste Wordtracker results into them. That's a very useful option for people who want to do serious keyword research. You could consider offering your customers that option.
_________________
Allan Gardyne
AssociatePrograms.com - You're here. Explore it! |
|
| Back to top |
|
|
Charlie
Joined:
22 Aug 2003
Posts:
3305
Location:
UK
|
Posted: Sun Nov 21, 2004 10:41 am
Post subject:
|
|
|
| AllanGardyne wrote: | | Some tools, such as Keywords Analyzer, allow users to paste Wordtracker results into them. That's a very useful option for people who want to do serious keyword research. You could consider offering your customers that option. |
But wasn't one of the original reasons for this tool to overcome the "heavy" interface at WT?
Cheers,
Charlie.
_________________
"Before I speak, I have something important to say."
- Groucho Marx |
|
| Back to top |
|
|
Phil Tanny
Joined:
22 Aug 2003
Posts:
1324
Location:
Gainesville Florida USA
|
Posted: Mon Nov 22, 2004 3:30 pm
Post subject:
|
|
|
Hi guys,
Allan, your frankness is always welcome here. No filters for me please! Reading between lines strains my literal minded brain.
The Planet Ocean article? They did a study of Overture and/or WT search numbers? I do see a way now to test Overture numbers against WT, more in followup post. But it's not a trivial job, so I was fishing for the group wisdom as a place to start.
Please remember that I have minor ability with Perl and a passion for interface design, but everybody here has done more keyword research than me, thus my questions. I need your help seeing the job through the eyes of someone who has done a bunch of keyword research.
I'm playing hard here, and making progress. Here's an update.
My emerging code grabs keywords and counts from Overtures keyword suggestion tool. You can click on terms you like and do new searches based on that term. Duplicates are removed etc. You keep doing this until you've got the list you want.
Then the software goes and gets the information needed to analyze via Sean's formula. Lowest PR of top 10 results, number of listings with keyword in title. You have to estimate how much PR you can give your pages, and enter that number.
End result is a list of your keywords, ranked by Sean's formula.
Sean's formula seems preferable to KEI to me, but this is a subject worthy of debate.
Yes, if you don't like using Overture for the keyword data, it's easy to offer a form for entering a WT report. Copy/paste WT report in to form and hit submit. So, for now, we'll resolve the Overture vs. WT issue by letting users decide.
The interesting part (for me) is putting the above in to a simple interface that explains itself. Ideally, any newbie should be able to see what to do without need of a manual or a lot of experience. This is what my teacher rants are all about.
Once I get a working rough demo up I'll PM you guys and you might perhaps kick the tires and tell me if I'm on the right track.
Thanks for the comments!
_________________
Free Forum And Ezine Hosting
http://Engage-Engine.com |
|
| Back to top |
|
|
Charlie
Joined:
22 Aug 2003
Posts:
3305
Location:
UK
|
Posted: Tue Nov 23, 2004 3:22 pm
Post subject:
|
|
|
| Phil Tanny wrote: | | The Planet Ocean article? They did a study of Overture and/or WT search numbers? |
It's a few months ago (at least) and it wasn't very kind to Overture.
I'd love to see a tool that uses Sean B's formula...
| Quote: | | Once I get a working rough demo up I'll PM you guys and you might perhaps kick the tires and tell me if I'm on the right track. |
Feel free. I always keep my big boots handy. 
_________________
"Before I speak, I have something important to say."
- Groucho Marx |
|
| Back to top |
|
|
Phil Tanny
Joined:
22 Aug 2003
Posts:
1324
Location:
Gainesville Florida USA
|
Posted: Tue Nov 23, 2004 4:30 pm
Post subject:
|
|
|
Here's a test I've got on my to do list. Does it make
sense to you?
The point of the Overture, WT, SBI or other search
total numbers is to give us an idea of the relative
popularity of different search terms.
Isn't it the ratio between one number and another that
is relevant? If someone reports 1000 searches for
"madonna" and 100 searches for "dog food" then we can
hopefully assume that madonna is about 10 times as
popular a subject as dog food.
So in that case the compartive ratio is 10:1.
If WT reports 100 searches for madonna, 10 searches for
dog food, that's the same ratio. We make the same
decisions when designing our site.
So the test would be to import thousands of terms from
Overture and WT, and then compare lots of ratios, one
service vs. the other.
If both services report roughly the same ratios for many
terms, then we might assume that they are both
measuring the same reality, each in their own way.
If the ratios bear no resemblance to each other, then
we might guess that one, or both services, have stinky
numbers, but we wouldn't know who the culprit(s) are.
Does this sound like valid testing logic to you?
Is anybody aware of tests like this? Or are we mostly
trusting the numbers because big companies are
giving them to us?
Thanks for your thoughts.
_________________
Free Forum And Ezine Hosting
http://Engage-Engine.com |
|
| Back to top |
|
|
Phil Tanny
Joined:
22 Aug 2003
Posts:
1324
Location:
Gainesville Florida USA
|
Posted: Tue Nov 23, 2004 4:32 pm
Post subject:
|
|
|
Side bar question:
Anybody aware of sites that mirror google results, like this site does:
http://prsearch.net
Thanks!
_________________
Free Forum And Ezine Hosting
http://Engage-Engine.com |
|
| Back to top |
|
|
AllanGardyne
Site Admin
Joined:
02 Jul 2003
Posts:
6326
Location:
by the beach, Australia
|
Posted: Wed Nov 24, 2004 10:32 pm
Post subject:
|
|
|
| Phil Tanny wrote: | So the test would be to import thousands of terms from
Overture and WT, and then compare lots of ratios, one
service vs. the other. |
I'd be very interested in seeing the results of such research. If the Planet Ocean article is correct, you'll see wide discrepancies.
_________________
Allan Gardyne
AssociatePrograms.com - You're here. Explore it! |
|
| Back to top |
|
|
Phil Tanny
Joined:
22 Aug 2003
Posts:
1324
Location:
Gainesville Florida USA
|
Posted: Wed Nov 24, 2004 10:49 pm
Post subject:
|
|
|
| AllanGardyne wrote: | | I'd be very interested in seeing the results of such research. If the Planet Ocean article is correct, you'll see wide discrepancies. |
Which would tell us what?
It seems wide discrepancies would tell us that one or both data sources perhaps can't be trusted. I don't know how to go farther than that unless I can test a larger number of sources.
Where does SBI get it's keyword/searches data? Other sources?
Is Planet Ocean worth the time?
I read Sean and Dr. Andy. Is Planet Ocean SEO overkill, or useful info worth the time investment?
_________________
Free Forum And Ezine Hosting
http://Engage-Engine.com |
|
| Back to top |
|
|
AllanGardyne
Site Admin
Joined:
02 Jul 2003
Posts:
6326
Location:
by the beach, Australia
|
Posted: Thu Nov 25, 2004 12:57 am
Post subject:
|
|
|
| Phil Tanny wrote: | | It seems wide discrepancies would tell us that one or both data sources perhaps can't be trusted. |
But what if there were only very minor discrepancies? That would tells us something very useful.
| Quote: | | Is Planet Ocean worth the time? |
Only if you use the info. I must admit I'm often not good at doing that. I've been a paying subscriber for several years. I like the way it keeps us up to date with an ever-changing scene, with monthly updates and articles. I regard it as an essential part of doing business on the Internet.
If it's important in the SEO industry, Planet Ocean will discuss it.
| Quote: | | Is Planet Ocean SEO overkill, or useful info worth the time investment? |
There's a heck of a lot of info, so I suppose it depends on your personality and working style. Sean Burns' book is much easier to read quickly and absorb.
_________________
Allan Gardyne
AssociatePrograms.com - You're here. Explore it! |
|
| Back to top |
|
|
Charlie
Joined:
22 Aug 2003
Posts:
3305
Location:
UK
|
Posted: Thu Nov 25, 2004 11:42 am
Post subject:
|
|
|
| AllanGardyne wrote: | | Quote: | | Is Planet Ocean SEO overkill, or useful info worth the time investment? |
There's a heck of a lot of info, so I suppose it depends on your personality and working style. Sean Burns' book is much easier to read quickly and absorb. |
I'd say Sean Burns' book is more of a "philosophy in action" rather than PO's "science" - not heavy, just analytical in approach.
I have a computing degree (unused and out of date), and I think like a programmer in many ways. Perhaps this side of you, Phil, will get on well with PO, too.
I mentioned PO here, also, if you're interested...
http://www.associateprograms.com/discus/sutra35750#35750
All the best,
Charlie.
_________________
"Before I speak, I have something important to say."
- Groucho Marx |
|
| Back to top |
|
|
| View previous topic :: View next topic |
|
FAQ
Search
Memberlist
Usergroups
Register
Profile
Private Messages
Log in
