Great Marketing Requires Experimentation, Not Just Analytics or Perceptual Studies

I just heard Ducan Simester, professor at MIT Sloan School of Management, speak on the critical role of experimentation in the marketing decision process. He presented fascinating research that states that most marketing decisions are made pretty badly because marketers rely too much on perceptual information/insights or past analytics, without enough real-time customer input when forming plans to positively impact demand/buying behavior.

Unlike with operational data, analyzing customer data is just too hard for most organizations to get value out of it - even the largest corporations - per Simester. It takes the best people around - top Ph.Ds - to make sense out of the complexities typically presented. Corporations simply don't have the capabilities to pull this off.

Perceptual research - think focus groups, interviews, conjoint studies - is just customers' take on how they are going to behave in the future. It isn't their actual behavior. So, perceptual studies are good for generating hypotheses on what to experiment on, but not for basing strategic marketing portfolio decisions.

So, the underlying principle for making the best marketing decisions is to determine the best behavioral customer experiments - with feedback mechanisms - to enable business units and marketers to vary what they are doing with customers and their channels. Measuring everything that makes a behavioral difference in the purchasing equation during these experiments is critical.

It's also crucial to keep these experiments simple and small, because you have to implement marketing experiments real-time, with real people in the "field. " If experiments aren't simple/small, your people unfortunately will have high degrees of failure.

Here are some core guidelines for conducting effective marketing experiments:

• Behavior: The key measurement
• Simplicity: So experiments can be implemented by people in their jobs
• Customer Acquisition: The target