Hitwise.com – Is a 10mm Sample Size Enough?

Today Hitwise, a subsidiary of Experian, announced that they were making big change in their tool by offering the ability to differentiate between paid and natural terms that create traffic.  I am very excited about this change as I have been asking for this since I became a client about a year and a half ago.  For those of you who don’t know Hitwise as intimately as I do, let me share. 

Hitwise is a competitive intelligence tool that boasts a sample size of 10mm searchers.  Now, I know that sounds like a big number, but if you sell to a niche like most, that equals about 15 people a day for you entire industry.  By looking at the sample size from a statistical perspective, most industries (not including porn,  internet searches, and music downloads) are not getting a large enough sample size to warrant a P factor acceptable to justify using the data in key decision making.  Especially considering that a yearly Hitwise subscription is actually higher than the average annual salary here in SC.  If the data is not statistically relevant, then it’s just a market-share checker…an expensive market share checker.

 I can tell you that their customer service is great and that their analyst’s blog is very fun.  In the same way ‘Freakenomics’ looks at economics, they look at web analytics.  The blog reports that the tool can predict the American Idol winner and the winners of Australia’s recent federal elections…too bad it doesn’t always offer data that results in clients making money.

* Hitwise.com Update – Jan. 11, 2008 – I had a call with one of the reps from Hitwise.com yesterday and she showed me the new features and some of the features that I have not been using in Hitwise.  In the next couple days I will post another review of Hitwise considering some of the vast improvements they have made and some of the features I have never explored.