27 Internet Marketing Rules for Success. Rule #15: Track Your Results

February 17th, 2010 by admin

If you follow skilled, successful marketers, you’ll notice that they frequently exhort new marketers to “test and track.” And this is great advice, but unfortunately, it is seldom followed.

New marketers usually either don’t have a strong understanding of how they can test and track; or they are not confident in the ways in which they are 1) setting up the test; and 2) tracking and interpreting the results.

If this is where you stand currently, understand that this can change. And if you’re willing to let it change—to learn how to test and track—you can truly gain mastery over your own results by becoming better at understanding and controlling the results.

Your first step should be to get a website statistics tracker that will provide you with a sufficient amount of detail. Regardless of which hosting service you use, there’s a good chance you already have one. Simply login your administrative account and check the control panel. You should see one or more icons on your panel that look like a graph. These will provide a comprehensive break-down of your traffic statistics.

If your site does not already come with a free website traffic statistics tracker, create a free account at http://www.statcounter.com/. This will give you a sufficiently comprehensive break-down of the traffic statistics for your site.

Once you have a tracking system in place, the only thing that remains is to carefully design tests. Whenever you do this, you will want to keep two things in mind: 1) as many things as possible should be held equal; and 2) you should allow for a long time horizon to pass before you reach a conclusion.

For part one, you can do this by using PPC traffic exclusively for your tests. This will allow you to be certain that the quality and relevance of the traffic you’re sending is roughly equal for you’re the two salespages or squeeze pages you’re testing.

For the second part, you will want to keep in mind that you will need a large set of observations before you can draw an valid inferences from your sample about your target population. Consider running your campaign for several weeks with identical traffic and different ads to see whether there are significant conversion rate effects.

Finally, and most importantly, remember to make only small changes when you split test two ads. This might mean only changing the headline on one salespage; and comparing the results to the original salesletter. This will make it much easier to be sure that you are capturing the effects from the headline change when you view your results (rather than some other change you made simultaneously)

Table of Contents