Google today launched a new service called Google Domestic Trends.  It provides a view of search traffic across several key categories of the economy and allows investors to compare these trends to the various market indexes. I was excited about this because I always thought online data could be useful in predicting real-world events. A million people start searching on vampires so you know something like Twilight is going to be a big hit. As more and more intentions come online via Twitter, Google, and social networking sites, there must be a way to gain insight into markets and economic trends.

So is the Google data useful?

Take a look at the chart below. It compares Google search traffic for clothing versus actual clothing sales. In describing the chart, Google writes that:

“From the graph below you will see that actual clothing store sales (source) and retail sales search volume on Google have been decreasing over time. There are seasonal spikes in actual retail sales, but over time there has been a pronounced decrease.”

I think this is a poor example for Google to use. I only visually inspected the chart (without any glasses) but I don’t see the trend.  To me, actual clothing sales were relatively flat between 2005-2007 while search traffic fell precipitously. Looking at the Google data would make someone believe that clothing sales by almost 50% between 2005-2009 while actual sales seem to have fallen by much less.

On to another example.

Google allows you to compare these trends to different market indexes – S&P 500, Nasdaq, Dow, etc. Take a look at the Real Estate index. I compared the year-over-year change in this index to the S&P 500. The resulting chart shows that real estate began a sustained decline in late 2007 approximately the same time that the S&P 500 peaked. But then real estate searches bounced back. You’d have to be a pretty savvy investor to tie that drop in real estate searches to a future plunge in the markets.

A lot more analysis needs to be done before a final conclusion can be drawn, but here’s what I think. Power users will find the information useful as another data-set in any model they build. I’m sure that if the data is correlated and crunched, there will be many useful insights. But to the casual investor, it’s unclear how much additional information Google Insights provide.

If you have any other thoughts on using Google data, let me know. Post your thoughts below.