I’ve been having some fun playing around with Google Trends, which is a simple, graph-based interface showing how queries at Google compare to each other. It compares the following:
- relative popularity; no actual counts.
- searches by city
- searches by region
- searches by languages
- and most interesting of all, the graph plots out events that may have influenced positive or negative spikes
Check out 49ers vs. Raiders vs. Cowboys. In this example you’ll see that it actually catches multiple meanings of Cowboys (I was thinking of the NFL team, but this also catches an NCAA basketball team.)
Google finance vs. Yahoo finance clearly shows the spike in May when Google launched a Finance site. The spike quickly drops after the initial launch.
Here’s a fun one, Jack Daniels vs. Jim Beam vs. Maker’s Mark. [As a sidenote I guess normalizing is minimal, b/c when I entered it as Jack Daniels vs. Jim Beam vs. Makers Mark I got a higher Marker's Mark count.] Jack Daniels spikes a few times with no apparent explanation, while Jim Beam had some news, such as Starbucks selling a coffee liqueur drink, that seemed intriguing but resulted in very little change in volume.