There are certainly times when you have too many options. Specifically, I’m talking about you AdWords campaigns. Making the right choices could potentially mean netting thousands of additional users. Thanks to split-testing, you can be more certain that you’re running the most effective campaign possible.
Split-testing is simple enough. Essentially, you run the campaign as is, and you also run it with one elemental change. Both ads are randomly displayed to users and eventually, over thousands of views, you discover which is more effective.
You change in your campaign can be a variety of things. Maybe the two ads being split-tested are drastically different in color, text or another way. Or, the ads could direct users to different landing pages.
To achieve your split-test, you’ll need to set AdWords to rotate your ads evenly. This may not be as effective as the ‘optimize for clicks’ option that AdWords suggests, but for the sake of testing it’s the only way to get real results.
Alistair Dent details the split-testing strategy at Search Engine Watch. Essentially, you should set aside one campaign strictly for split-testing, which allows your other campaigns to capitalize on what is already working. Copy the campaign you wish to improve into the split-test campaign, then turn it off in the main campaign.
Remember, it will take time and thousands of users to minimize potential variables, so don’t rush through the testing phase. If you are patient, you’ll get a clear answer on how to tweak your campaign.