Using Google Trends to make GOP election predictions

The following is a look at the GOP presidential primary race through the lens of search engine activity. Because the polls are not always inclusive of all candidates and they rarely show detailed state-by-state data, I thought this would be an interesting way to look toward the future state-by-state primaries. Of course, the true accuracy of this data is only known to Google, but it is still interesting to look at nonetheless. It could also, I think, perhaps lead to some strategic shifts from candidates - though, I am biased because I work at a firm dedicated to analyzing trends in search and search engine marketing.

Overall search activity for the past 30-days shows significant and growing interest in Ron Paul culminating on November 5th and 6th after the moneybomb.


The more interesting data, though, is to look at which states show the most interest in each candidate. While Ron Paul outpulls all candidates in all states, when you rank each candidates search volume by state, you start to get a very interesting picture of how the campaign might shape up. The following are the top 10 states (ranked in order) searching for each candidate according to Google Trends (early voting states highlighted blue):

Ron Paul	Rudy Giuliani	Mitt Romney	John McCain	Fred Thompson	Mike Huckabee
Arizona	District of Columbia	Utah	South Carolina	Tennessee	Arkansas
South Carolina	New York	Idaho	Arizona	Iowa	Iowa
Tennessee	New Jersey	New Hampshire	District of Columbia	Alabama	South Carolina
Iowa	Ohio	Iowa	Virginia	South Carolina	Alabama
Nevada	North Carolina	South Carolina	Michigan	District of Columbia	Kentucky
North Carolina	Pennsylvania	District of Columbia	Massachusetts	Kentucky	North Carolina
Utah	Virginia	Arizona	Illinois	Georgia	Tennessee
Alabama	Illinois	Massachusetts	Georgia	Maryland	Georgia
Oklahoma	Texas	North Carolina	North Carolina	North Carolina	Missouri
Oregon	Massachusetts	Colorado	New York	Wisconsin	District of Columbia

Based on this data, you could conclude the following:

• Mitt Romney is in the driver's seat to win the nomination with three of his top 10 states being early-voting states.
• Iowa is up for grabs between Romney, Thompson and Huckabee.
• Mitt Romney will win New Hampshire with ease - unless a loss in Iowa derails things.
• McCain is the front-runner for South Carolina, but could get derailed by poor Iowa and New Hampshire results.
• Rudy Giuliani is very formidable in the big states which could be enough to offset early losses. This seems to be his strategy but could losing early take the steam out of the campaign? (I don't think so)
• Ron Paul's strength is the South and the West which with later voting dates, could sink the campaign. (this might be why he is going this month to Nevada - to shore up support on the West for a late surge?)

Overall, I think this race will come down to Romney needing to win early and strong so that he has momentum going into the big states against Rudy Giuliani. The other candidates have pockets of strength which could lead to platform negotiating down-stream in exchange for endorsements that deliver constituencies. Candidates like Ron Paul and John McCain will have a big impact on the end-outcome if they start to steer the other's platform to accomodate Paul and/or McCain constituents. Fred Thompson's platform is not unique enough to warrant any platform integration with the larger candidates. Huckabee currently has some momentum, but it's not clear if his positions are sufficiently independant enough to provide a constituency he can deliver through endorsing someone else (his strength - social issues - is already being stepped over by some of the largest social conservative endorsements already delivered to Romney, Giuliani and Thompson).

CAVEAT: This exercise for fun. Google Trends is directional only and may not be a reliable source of data to predict future purchase (or voter) behavior.