Google offers a whole suite of applications besides their famous search engine. Although Google Trends is still "in the lab", this fun tool can be useful to show what people are searching for and when they do it.

1. Gold or Silver?

Anyone who sells hand-made jewelry will know that silver sells better. I put it to the test by searching for "gold jewelry, silver jewelry" worldwide for the last four years. Lo and behold, the first graph (click for a larger image) shows that silver (red)does interest people more than gold (blue) and that the highest search volume is the last quarter of each year, shown by the spikes. Spikes are sometimes labelled. For instance the blue "A" was related to an article by Goodhousekeeping entitled : "Gold Jewely : You may not be getting what you think". Interest in the subject was no doubt due to the Christmas season.


2. Necklaces, Bracelets or Earrings?

Next, I tried to see what interested people more - necklaces, bracelets or earrings (Graph 2). People searched for necklaces (blue) more than bracelets (red) and earrings (yellow) for the last quarter of 2005 and 2006. I always thought earrings sold better as people could afford those more often than necklaces and bracelets. This may still be so. All the Google data shows is greater search interest in the more expensive necklaces and bracelets, presumably for gifts.

However, the last quarter of 2004 showed an anomaly - there was a large red "hill" for bracelets due to Lance Armstrong's Livestrong bracelets sparking the charity bracelet fad back then.

3. Can Google Trends work for different regions and years?

Google Trends allows the user to select countries and sub-regions such as cities for any given year since they started the service. To test this, I limited the search for necklaces, bracelets and earrings to just Canada and the last graph really shows up the data gaps in the past when Google Trends did not update regularly or maybe not as widely. Google only very recently started updating daily. Perhaps with time, and if there is sufficient search volumes, localised data would be more meaningful.

What Google trends show are just broad search patterns, fun to do to get a general idea but as they say themselves, don't base your PhD dissertation (or entire marketing plan) on the results.