There has been a common theme in my last few blog posts regarding both educating and incenting marketers to spend more time on in-site conversion optimization. In short, eCommerce professionals have all the tools at their disposal to convert more of their existing traffic; why aren’t they doing it?

Over the last week or so, I have seen some of the answers to that question in a series of great blog posts by Jim Novo regarding Control Groups.

While his purpose is to speak specifically to email marketing,  I think it also directly speaks to some of the core issues that marketers face in the evolution from their current task orientated responsibilities to being more goal oriented - in line with corporate objectives:

• Proving the direct impact to the business of any in-site marketing or technology initiative.
• Justifying the opportunity costs spent on doing something new, rather than the traditional traffic generation/post-sale marketing currently being done.
• Dealing with the fact that analytics data is never 100% accurate, and needs to be used more like a barometer than a calculator.

Quick example: Marketer decides to bring on an advanced merchandising management tool, never a bad decision. Three months later the vendor shows them fantastic results in the reports generated by their tool, but marketer’s analytics show no direct lift. Without being able to directly prove revenue impact, the vendor gets fired and the marketer gets egg on their face.

Imagine instead if the marketer had done two things differently:

• They had implemented the merchandising management tool so that it only affected 50% of the unique sessions on their site.
• They built separate conversion funnels for the ‘control’ and ‘optimized’ pieces of their traffic in their analytics tool.

Like Jim Novo says in his articles, the marketer will now be able to both definitively prove direct impact in terms of sales (product shown, product clicked on, product bought) but will also be able to show the ‘halo effects’ on conversion that showing more relevant products has. The control group provides a real world, real time baseline (Think Bowflex ‘Before’ and ‘After’ picture).

In short, conversion optimization is like high school math. You only get full credit when you show your work. A properly executed control group is a great way to ensure that happens.

Relevant Articles
Jim Novo: Why Use Control Groups? - http://blog.jimnovo.com/2007/12/02/control-groups/

Jim Novo: Control Group Benefits - http://blog.jimnovo.com/2007/12/11/control-groups-2/

Related Posts

• Should Dayparting be a part of your day?
• Why accept low conversion rates?
• Day of the Long Tail