Building on the support tools around Google Search, the company has improved Google Trends with new topic reports. The new topic reports aim to be more precise by taking into account the differences between similar search terms. Google’s own example demonstrates how analogous search terms can skew results on Google Trends. Looking up “rice” on Google Trends could mean either measuring search interest in Rice University or the rice you eat. Searching for a popular term like “Gwyneth Paltrow” also does not factor in the misspellings people might be using to search for the actress. The improved topic reports seek to fill these shortcomings.
Now, if you type “rice” (as in Rice University) to measure the trend vis-à-vis Harvard University, Google will automatically predict Rice University, and you can thus make a fairer comparison instead of warping the result with anything that’s analogous to “rice”.
The new Trends has also logically stepped up to handle misspellings which possibly get left out of search data, though representing genuine search interest.
The Google blog says,
So, when you measure interest in “Gwyneth Paltrow (actress)” our algorithms count many other searches that mean the same thing “Gweneth Paltrow,” “Gwen Paltro,” etc. As our systems improve, we may even count searches like “Lead actress in Iron Man.”
Topic reports are a beta feature. Google is building up the data sets, and presently it has data for 700,000 unique topics, with the ability to measure the search interest data worldwide or in Brazil, France, Germany, India, Italy, UK and the US. If you use Google Trends to compare search trends, this precision is surely worth a good comment.