Dealing with Local Search

If local search is the way of the future, how do you track local traffic?

[ciaoenrico]

I am trying to determine search volume for a client who is locally based. Unfortunately, all of the traffic predicting tools that help me decide where searchers are looking, and what for, are nationally generated, and so are useless for forecasting.

Wouldn't it be neat if there was a tool for doing just such a thing? I'm thinking it would key off of the IP address of the computer doing the search.

Does such a tool exist?

If someone ultimately builds such a thing based on this post, and I get a copy, I'll be your friend. :)

[earlpearl]

Using #*$! or overture, you will often find search traffic numbers for the regional or local version of the phrase. This list is way down and typically a fraction of the generic term.

For instance, for the #1 phrase for my industry, dependant on month (it is seasonal) I might see overture estimate anywhere from 10-20,000 searches/month. For my region which comprises portions of 2 states and 1 city I might see something like 50-200 visits that month for any one of the major jurisdictions. So the predicted total might be in the vicinity of 300 visits (per overture).

More specifically, and in order to capture traffic I looked at the most recent search traffic for my site over a 7 day period.

About 750 search visits. Now my site is well optimized for national terms with a lot of SE 1st pages for the main phrase and secondary terms.

But this is the difficult part in optimizing or predicting traffic.

Of the 750+ searches, 281 were for a variety of generic terms including the #1 phrase. We had 64 visits on the #1 phrase in a week. The rest of the visits for industry phrases totaled more than 3 times the number 1 phrase.

On the local side we had about 260 industry phrases with a state, city, town, or regional connection.

You know what is so hard to predict. The highest number of searches for a local phrase during the week was 9. We had over 220 different phrases with industry terms and a relevant geographical reference.

Predictions are terribly hard with 220 different phrases that incidentally, are the most likely to convert. Especially when those 220 phrases cover only 260 different terms. Not a strong basis on which to establish predictions.

Rather than predict the volume of traffic I would emphasize the importance of the long tail effect to ensure that your client's site be visable for many variations on the #1 industry phrase or #1 industry phrase with a geo description.

Good luck,

Dave