Wednesday, November 19, 2008

Testing book titles using Google AdWords

My 'travel book for nerds' book, The Geek Atlas: 128 Places Where Science and Technology Come Alive, will be published in April 2009 by O'Reilly. As part of the process of writing the book I had to come up with a title. I had three titles that I liked: A Voyaging Mind, A Mind Forever Voyaging and A World of Discovery.

Ultimately, O'Reilly came up with the current title after doing their own market research, but before that I wanted to figure out which of the three titles would work best.

To do that I bought ads on Google AdWords that were relevant to the book (such as when people search for 'science museum') and set up three ads that would appear randomly. The ads all had the same text except for the main title which was one of the three possible book titles.

I let the campaign run for 30 days and then analyzed the results to see which one had the greatest clickthrough rate. There was a clear winner: A Voyaging Mind.

And for a long time A Voyaging Mind was going to be the book's title.

It seems to me that Google AdWords could readily be used for other such experiments: it's cheap, it's simple to target your experiment based on keywords so that you can choose the type of people exposed to the experiment and by setting up random display of a set of ads you can try out variations of an idea easily.

Obviously book titles are just one possibility. What other things could be tested using Google AdWords?

1 comment:

designcortex said...

did they all show up in a list like that. One after another. Because if they did of course the top one trapped 40%. People always click on the top link