Any Wordtracker Experts Here?

I need some verification that my assumptions about Wordtracker are correct.

[Christyl Stevens]

I hope that some of you who really understand how to best use Wordtracker can help to clarify and/or verify my assumptions.

Here is what Wordtracker reports.

used cars houston

used cars houston 148 140
houston used cars 142 134
used cars in houston 58 55
used cars houston texas 44 42
used cars houston tx 13 12
buy used cars houston 6 6
used cars in houston tx 6 6
houston used cars and trucks 5 5
used cars texas houston 5 5
houston texas used cars 4 4
houston used cars prices 4 4
used cars + houston 4 4
used cars + toyota + houston 4 4
used cars and houston 4 4
used cars in houston texas 4 4

used cars bakersfield

used cars in bakersfield 4 4
bakersfield used cars 3 3

Question #1

Is it safe to assume that since Bakersfield does not show the same type of keyword phrases as Houston does, that it does not mean that people are not using those EXACT keyword phrases to search in their own local area with the only difference being the location, as in this case, Bakersfield instead of Houston?

Question #2

Then, just because "buy used cars houston" shows 6 searches and "buy used cars bakersfield" shows 0 searches, does not mean that people aren't REALLY searching for that phrase, right?

I mean, Wordtracker is just an ESTIMATE of the Metasearch engines DogPile and MetaCrawler, right?

So that means that Wordtracker is only a sample of less than 5% of all the searches performed using all the search engines, right?

So that means there REALLY could be searches for "buy used cars bakersfield" on the other search engines that Wordtracker doesn't collect statistics for, isn't that correct?

Therefore, if I have a valid understanding of how Wordtracker collects its data and if I have a valid assumption of the searching habits of people (no matter their location) then would I be safe to optimize a page for the term "buy used cars bakersfield" and the to also EXPECT that folks in that part of the country will use that exact keyword phrase to query Google?

TIA

[martinibuster]

bakersfield is a relatively backwoods kind of town. I drive through there on my way to or back from Arizona.

I doubt too many folks are looking for used cars over the internet when all they have to do is open the curtains look out their windows, so to speak.

Many people assume the search engine knows where they're at, so they search for "used cars" rather than "used cars bakersfield." If you're business is in Bakersfield go ahead and optimize for it. It'll be easy to do. I wouldn't expect much, though.

[Macro]

I would dump those results as being not far from useless.

If you don't use KEI or advanced wordtracker tools then don't pay the wordtracker subscription. You can get the frequency of searches FREE from overture tools. In fact, I get exactly the same results in Overture as I get in Wordtracker.

And they are both equally useless for frequencies under approx 200 because they are all guesses. Yes, GUESSES. An extrapolation tool is used so when Wordtracker says KW1 was searched for 178 times and KW2 was searched for 90 times it is quite possible that they were both searched for just 2-3 times and Wordtracker is guessing to come up with 178 and 90.

[DylanW]

bakersfield is a relatively backwoods kind of town

That said, have you tried checking Wordtracker for "used cars" in larger cities near Bakersfield? Assuming they're within reasonable distance, people might travel to do their car shopping (of course, that may apply to new cars more than used cars, but it's worth a shot).

[Robert Charlton]

That said, have you tried checking Wordtracker for "used cars" in larger cities near Bakersfield?

Not sure whether Christyl is actually targeting Bakersfield, or that's just a hypothetical example. As martinibuster suggest, Bakersfield is fairly far from everywhere. I'd guess Fresno and Los Angeles are each 100 miles or so away and that's pretty much it for larger cities nearby.

For such a small local area, I absolutely would not trust Wordtracker data. Their sample size is way too small.

Whenever I target any small local area, I assume the search patterns are going to be roughly parallel to those of larger areas... the difference being that some areas might include state ID or regional descriptors.

Eg, the San Francisco area often has Bay Area tacked on. A Bakersfield search, if localized at all, might say Central Valley or whatever they call it. And some towns, liked Portland, Maine, might include the state name (or initials) to distinguish it from Portland, Oregon.

If you are targeting an area as tiny as Bakersfield, chances are you don't have to worry about getting the phrase variants and word order exactly right, as long as you include all the words. The searches are probably so uncompetitive that any word order would suffice. And, with what Google's doing right now, fuzzy targeting might be better than exact target.

The thing to remember about local searches is that there is no one pattern. Some searchers may include state name... some may use the abbreviation. Generally, though, if you don't have at least one instance of whatever it is on the page, you're not going to come up for the search, so for local targeting you have to target the page well for your core phrase, and then pick up all the peripheral targets with your location terms.

If the areas and terms are competitive enough, like big city hotels, you're going to have to build special pages (or even sites) for each location.

Many people assume the search engine knows where they're at, so they search for "used cars" rather than "used cars bakersfield."

True, but chances are you're not going to be able to rank on "used cars" by itself, so you'd better hope that your searchers use your location name too.