18456Re: Offermatica / Test and Target results analysis
- Jul 21, 2008As a rule of thumb, you should always let the test run for at least 7
That way you're covering an entire week cycle, and the results won't
be skewed by weekday vs weekend traffic.
--- In email@example.com, "thorynstephens"
> Offermatica / Test and Target results analysis
> I've been instated analyze our Offermatica / Test and Target results
> and have a few questions regarding interpreting the data.
> At a recent Advanced SiteCatalyst training class I spoke with a Test
> and Target rep regarding data interpretation, and he recommended the
> best practice of running the test for a minimum for 3 days with a
> minimum of 90% confidence.
> As we're running tests in different markets (UK, France, Spain) with
> multiple media types (search, display) we're seeing different
> traffic levels due to the media spend and size of the market.
> That said can anyone recommend further best practices/papers for
> interpreting this data, realizing each market is different? What
> are sufficient traffic levels before we pick our winner? For what
> T. Stephens
- << Previous post in topic