Once you have set up a new pay-per-click campaign, landing page split test, or search engine optimization campaign, you have two main choices: let it sit and hope for the best, or optimize your new project. Obviously, given the amount of competition out there Internet Land, optimization is the only realistic choice.

Optimization, in very broad terms, is simply getting the most bang for your buck. Usually, in the world of online marketing, it refers to decreasing your cost per conversion (i.e., cost per sale, download, inquiry sent, etc.).

Now, there are a ton of ways to optimize any given campaign, depending of course upon the nature of the campaign, the techniques you are using, etc.. The specific ways to optimize for a given campaign is for another post. But, in very general terms, the optimization process goes like this:

a. set up your campaign
b. define some metrics (something you can measure who measurements represent success or failure)
c. establish some baseline metrics (i.e., take initial measurement)
d. make a small change
e. wait a while (to give things time to collect data)
f. measure again
g. compare new measurement to previous set of measurements
h. go back do “d” and repeat

Here’s the biggest pitfall that so many of my clients fall into: they try to test more than one variable at a time. In the world of optimization, here is the golden rule:

Test only ONE variable at a time!

Here’s why: if you test more than one variable at a time, whether the results (see “f” above) are good, bad, or mediocre, it won’t matter, for you will have learned NOTHING about what really worked, what didn’t. In other words: you won’t be able to tell for sure what made the difference! You may as well have not done anything at all.

This is one of those pieces of advice I’ve repeated to customers, colleagues, and random people on the street countless times. Sometimes it sinks in right away, sometimes it takes a while. But, if you think about it, it’s the only way to do optimization that makes sense. So: follow it, use it, live it! (And, happy testing, by the way.)