Testing & Personalization Kill Best Practices

Back in October of last year I wrote about lazy marketing or deciding to implement creative and design strategies based on so called best practices. Last week MarketingShepa published a case study of work we did with a client on button testing that once again proved that even for buttons, best practices need to give way to intelligent inconsistency as predicated by testing at each step of the funnel.

This makes sense when you understand that while there is an overall goal of your media or website each step of a user flow has a specific goal, so the messages, images presentations and yes, even buttons, are likely not optimized if they are the same across this flow. As mentioned in the original post this gets magnified as we start to look at segmentation to understand and address the different needs of users.

I’d like to reiterate an important piece from the original post:

This gives birth to a new best practice.

ABT…always be testing.

The absolute results of this will be the great cycle that delivers LIFT:

Learning
Improvement
Failure (faster failure, a good thing!)
Thinking

(Note to self…no more acronyms for a few months)

It’s also worth noting that since the multivariate test results of our initial work with DYMO (that in addition to the Sherpa case study can be seen in more detail here) they have progressed into behavioral targeting. We’re proud that they will be at eTail 2007, (along with another OTTO client, Fingerhut) to deliver a case study on a personalization campaign to owners vs. non-owners that resulted in a huge lift in conversion rate.