Comments on: Causation vs. correlation http://www.lunametrics.com/blog/2006/05/24/causation-vs-correlation/ LunaMetric's blog on conversion rate and web analytics Tue, 06 Jan 2009 13:07:24 +0000 http://wordpress.org/ By: Mr. R. http://www.lunametrics.com/blog/2006/05/24/causation-vs-correlation/#comment-84 Mr. R. Sat, 27 May 2006 05:01:00 +0000 http://72.22.16.69/blog/?p=148#comment-84 The problem is that in scientific world you can have controlled variable to trully test if A infuences B. In real world you can't control the variable, so you have to "educately guess" The problem is that in scientific world you can have controlled variable to trully test if A infuences B. In real world you can’t control the variable, so you have to “educately guess”

]]> By: SiteSpect A/B Split and Multivariate Testing http://www.lunametrics.com/blog/2006/05/24/causation-vs-correlation/#comment-85 SiteSpect A/B Split and Multivariate Testing Fri, 26 May 2006 21:18:00 +0000 http://72.22.16.69/blog/?p=148#comment-85 This is a fruitful observation, and it's true: traditional web analytics report on "what happened" historically. They enable us to segment traffic so we can measure correlations such as PPC traffic for keyword A converted at X%, vs. keyword B converted at Y%, etc.<BR/><BR/>However, causality (vs. correlation) is not unattainable. In fact, A/B and multivariate testing <I>do</I> give us the ability to establish <B>causality</B> [note: you knew I was going down this road, didn't you? ;) ]<BR/><BR/>Through A/B and multivariate testing, marketers/analysts gain control over their visitors' experiences and can learn how site changes (e.g. a different layout, promotion or copy element) impact behavior. The capability to intentionally alter user experience and measure response lets us establish if certain content caused a change in behavior. This is the essence of why the things we learn through testing are so highly actionable.<BR/><BR/>-Dave from <A HREF="http://www.sitespect.com" REL="nofollow">SiteSpect</A> This is a fruitful observation, and it’s true: traditional web analytics report on “what happened” historically. They enable us to segment traffic so we can measure correlations such as PPC traffic for keyword A converted at X%, vs. keyword B converted at Y%, etc.

However, causality (vs. correlation) is not unattainable. In fact, A/B and multivariate testing do give us the ability to establish causality [note: you knew I was going down this road, didn't you? ;) ]

Through A/B and multivariate testing, marketers/analysts gain control over their visitors’ experiences and can learn how site changes (e.g. a different layout, promotion or copy element) impact behavior. The capability to intentionally alter user experience and measure response lets us establish if certain content caused a change in behavior. This is the essence of why the things we learn through testing are so highly actionable.

-Dave from SiteSpect

]]>