Enough with Ridiculous KW Tool Results

[Ryan Gabriel]

I really hate to create another thread on this theme, but with my most recent search, I feel compelled to get some opinions.

I did a keyword search at SEO Book in a less than mainstream niche- over 100 longtail results came back and they all had insane search numbers listed.

I'm talking about a 7 word keyword about a rather obscure subject getting 4000+ daily searches. Number 100 on the list gets 400+. All keywords above #24 get over a thousand searches daily. Probably 75% of these domains are currently available, as well.

There is absolutely no way this can be even remotely accurate. A search for the number one keyword brings back a Squidoo lens that looks to get very little traffic. It's been voted on by visitors a total of twice. There's a poll on the lens with *one* vote.

What could possibly be going on here? How can such a fundamental aspect of internet marketing be so unreliable?

[jakesellers]

There's basically two ways to estimate search volume, one involves voodoo math which will tend to be inaccurate, the other involves browser toolbars, spyware or other mechanisms which generally extrapolate from a statistically insignificant and unreasonably biased sample of searches. Personally all I care about is Google searches and the Adwords keyword tool is right(ish) more often than it's wrong.

[Ryan Gabriel]

Originally Posted by jakesellers

Personally all I care about is Google searches and the Adwords keyword tool is right(ish) more often than it's wrong.

It's my understanding, however (and please do correct me if I'm off here), that the adwords tool is largely innacurate for SEO/organic ranking because it factors in the search partner network and domain parking.

No?

Is there a way to determine traffic of individual webpages the same way alexa does for domains?

[Steve Faber]

I know that the majority of those stats are not very accurate, because I've had No 1 Google rankings for many different keywords, and very rarely did my traffic come close to what the number 1 position on Google was supposed to get according to the chart on W tracker.

[jakesellers]

Originally Posted by Ryan Gabriel

It's my understanding, however (and please do correct me if I'm off here), that the adwords tool is largely innacurate for SEO/organic ranking because it factors in the search partner network and domain parking.

Gotta keep the apples and oranges separate. AKT is an estimate of how many searches will generate Adwords impressions for that search. In most cases it's a fair enough estimate to work with.

In terms of "how many of those searches result in a click on #1," factor 1 is Adwords advertisers - ads sap organic clicks. My favorite example is 10 advertisers with 5% CTR - half the organic traffic's gone. Factor 2 is search intent, especially commercial. If you rank up a site that's "about" something when people are really looking for a storefront, it won't get much traffic. Wikipedia might come up first for things like "toner cartridges" for example but I think everybody pretty much knows what they are.

If you determine there's worthwhile search volume there's pretty much no way to determine if you can crack the top 10 unless you look at the top 10, and sometimes the top 100. I worry about sites that are hard to shake, like Wikipedia, really old sites, too many results - I like to see a top 100 of stuff nobody would click on for that search. In general worrying overmuch about pagerank, backlinks, keyword density or what have you isn't useful. I think "can I create a site that belongs at #1 for this search and will still make money."

My approach is probably a little different from some peoples' but practice lets you take an intuitive approach.