|
After creative freedom in brainstorming ideas, we
need to add some discipline and structure to the process of
defining solid test elements and reliable metrics. The goal is
to focus on the highest-potential marketing-mix variables and
the right data to quantify success.
Your ideas provide the “potential energy” of
the test, which must then be converted into actionable test
elements. Yet even a simple idea like “brighter envelope” can
become many different variables with myriad levels. The
challenge is to balance conflicting objectives:
- Test more to learn more, but do it
quickly
and cheaply
- Test bold changes, but within reason
- Test what you know is important, but
don’t ignore off-the-wall ideas
- Quantify everything… except what you
can’t
- Select distinct levels for each variable,
out of the infinite possibilities
Click here to see the step-by-step process for
selecting clear, bold, and powerful test elements.
Measuring test results is similar to measuring
advertising effectiveness. You need the right data to see the
impact of each ad or test recipe, while reducing error from
outside factors and random noise in the marketplace.
Metrics may be clear and reliable—like direct
mail response rate or landing page conversion rate—or may
require some in-depth analysis and planning before launching a
test. Retail and advertising metrics, in particular, often
require careful up-front analyses of store-by-store sales,
regional sales, and call center or Internet data collection
methods. Yet even a metric like “average order value” for a
catalog or e-mail campaign requires statistical preparation
before results can be analyzed.
Click here to learn more about selecting
reliable key metrics.
After adding scientific discipline, the next
step is to leverage statistical power in the test design.
|
/images/MsSpacer.gif){kind=spacer}
