In the very first blog post I wrote for Monetate, I talked about the “garbage in, garbage out” notion of computing programming. In other words, the value of what you get out is inextricably linked to the quality of what you put in.

The point of that post was to convey the importance of proper planning in website testing and optimization. Specifically, untargeted tests (i.e., those shown to all website visitors as opposed to a key audience segment) have little value to the organization. There is no “One Perfect Page,” and so good tests—and conclusive analysis of those tests—almost always require you select a meaningful segment of your website traffic before the test begins.

To be clear, website testing is a system of moving parts—and one that only works when all inputs (i.e., who you’re targeting, and what you’re showing them) align to produce the right experience at the moment it’s needed.

But one input is more valuable than all the others: the customer segment you’re targeting.

In my first class, and on my first day, of college, my professor began by stating, “If you want an A in this course, you need to teach me something I don’t already know.”

In website testing, it goes without saying that your ability to teach the organization what worked, what didn’t work, and why, is valuable. But the real value to the organization—what has intrinsic worth as an input—are the segments that feed these tests, and the optimization vendor’s ability to:

• Identify new segments
• Teach the organization about ones it did not know about previously
• Make these segments available for applications beyond just testing

You’ve heard it said that success isn’t about what you know, but who you know. The same is true of website testing and optimization. (Yes, life imitates computation.)

To know who your customers are is to control commerce. The next time you design a website test, consider devoting as much time assessing who you’re targeting as you do what you’re showing them.

Remember–garbage in, garbage out.