I bring in organic traffic through Google in first position for two different keywords, the only difference between the two being one is a plural of the first. The Search Volume tool shows that the plural brings in approximately twice as much traffic as the singular. Only after reviewing my site stat logs does it show quite the reverse. In fact the singular brings in 4 times as much traffic as the plural. How much trust do you put into the adwords keyword tool when figuring your advertsing budget?
As with all analytics, it's tough to trust any single source. I find at times that Google's tool underestimates traffic volume especially for niche keywords. I have also seen WordTracker be incorrect on a large scale. So I really don't trust any tools and try to use multiple sources or I just rely on a degree of magnitude and let reality tell me the details. Do you find Yahoo's new keyword tool more accurate since it provides specific historical searches per keyword?
I think, there is a tool, is better than nothing. At least, help us to check the trend and other related matter. Even not accuracy but okay for me .
What might be interesting is a tool that compares the keyword results for 3 of the most popular keyword tools. A tool created by Google seems like it would be the most accurate, because they probably have the number 1 data source to work with. If they put it in a format like wordtracker that would be great. I suspect they aren't using their search records db very well with this tool.
I think putting your trust in the tool's accuracy would be wrong. However, it is definitely a good way to study trends and get an idea of what to go for. No tool is perfect, but they're all somewhat helpful.
A tool is better than nothing. At least it serve as a guide. I guess it would be useful when Search volume is used together with the Advertiser Competition to serve as a guage to bidding cheap and most searched keywords.
I've never had an accurate estimation of traffic (except once, when it said there was no traffic, and there wasn't!), and the cost per click is generally pretty hopeless too. Personally, I don't use the tool at all unless I'm looking at a completely new market - you can hazard a reasonable guess at cost per click by guestimating the value of a conversion and the conversion rate (though this is more likely to be an upper bound than an accurate estimate), and the best way of finding out the traffic is to run the campaign for a week or two.
I also feel the tool can be given a chance as the data is not accurate (as none of the data can be accurate for online search). The search volume works for me sometimes, even that trend which shows the months when that particular keyword was searched higher is also Coooool.