Forum Moderators: open
AdWords' keyword tool versus Wordtracker for Google SEO
[webmasterworld.com...]
Actually I encountered a difference in the results of Overture and Espotting for the keyword - "key stage 2", searched in the month of February'2004.
Was just wondering on whose results should I rely.
Can anyone tell if there's any tool which can tell the frequency of keywords searched in the last 3 months instead of 1 month, offered by Overture.
Eagerly waiting for ur reply
As many mentioned overture and **** , what does they actually show, number of unique searches as in from a unique ip, unique SE?
Also how can I find the number of searches for the word "Blue Widgets" for Google Yahoo,msn ... sepratly on per month basis.
Which is the best way to find the online market for a search word. As of now I use all the four mentioned wordtracker, google, overture and Espotting.
Can someone clear the definition for these sites and as Vaniaul asked why are they showing different search number per month, sometime by a big margin
Regards
Aji
...how can I find the number of searches for the word "Blue Widgets" for Google Yahoo,msn ... sepratly on per month basis.
You can't. It's an inexact science at best, and the engines keep their data very private.
...why are they showing different search number per month, sometime by a big margin
Read (or reread) the thread I link to in my post above.... It's all there. Here's a partial list, spelled out in more detail.
a) demographics - The engines have very different audiences. WordTracker pulls its data from some pretty lowbrow metasearch engines, and then assumes it can get the figures for Google just by adjusting for market share. It doesn't work.
b) sample size - The metasearch engines that WordTracker uses have a very small sample size. WordTracker extrapolates this small amount of data by multiplying it up to represent a much larger share of web searches. In so doing, they're also scaling up their errors.
Say their metasearch data is one-fourth of one-percent of all web searches. Say Google is fifty-percent of all web searches. In multiplying their figures by 200, which is what they'd have to do to predict Google traffic, they're also multiplying their errors by 200. A difference of 1 or 2 searches in WordTracker's data shows up as a difference of 200 or 400 searches in their predicted Google searches.
c) machine automated searches - These skew all the engines, but most significantly skew Overture. Let's say somebody checks his position 4 times a day. That's 120 times a month. If ten people do that, that's 1200.
WordTracker also occasionally shows automated searches... and these get scaled up unrealistically when they extrapolate their data.
d) MatchDriver - For the most part, Overture lumps singulars and plurals together, and also does a quirky kind of broad matching, alphabetizing the word order of some searches and lumping various searches together. You can do a site search on MatchDriver (or two words, Match Driver) to get a more detailed idea of what they do.
Note that the Google tool referred to in the Sept 2002 thread I link to is the old Google AdWords tool. The current one is much better. I find the AdWords Sandbox very to be useful, but it doesn't have search frequency numbers attached to it. Your best numbers will come from a test PPC campaign.
For traffic estimates, only Google Adwords gives you reasonable accuracy (the majority of searches are done on Google, so the sample size is very good for this purpose). You can also target by country (US, UK, etc), and it doesn't have to cost a lot to get the data. (that's where max CPC levels come in handy)
I mean Adwords , how do you use it to get the approx number of searches?
None of the tools gives real numbers. At best, you can get an approximate relative ordering of your phrases.
As noted in thread below, I use the AdWords tool in conjunction with the Overture Tool, with great skepticism about the numbers in the latter....
SEO search phrase research with new AdWords Keyword Sandbox
Works well in conjunction with Overture Tool
[webmasterworld.com...]
If you do want numbers, you have to run a PPC campaign, with exact matches, and then analyze your data, keeping in mind that AdSense clicks are not the same as organic clicks. The thread I referenced in msg #5 above was asking about relating AdSense clickthroughs to organic clickthroughs. That would be good data to have if anybody has it. I think the ratio varies depending on the quality of the organic serps.
Again, I strongly recommend searching WW using Google on this subject and doing a lot of reading.
Cheers, CaboWabo