AdSense split-testing (a/b testing)
One of the easiest ways to increase your AdSense earnings is to optimize your page ads.
Basically you can try different ad sizes, colors, types, and location. Obviously you can also try different ad providers, not just AdSense.
The best way to see which is the best combination is to do split-testing (sometimes called a/b testing). Split testing means that you keep changing between combinations - usually with a simple ‘random’ way (or a round-robin). This means that the next visitor will get ad type 1, and the one after that will get ad type 2, and so on. This is better than changing the ads once per day, or a week, because it will be very difficult to find out if the performance is indeed better in one combination over the other - maybe the traffic profile changed over the days/weeks? (for example, the day of the week makes a lot of impact, and the traffic source is also important).
So making the combinations change ALL the time makes the test more reliable.
Following the results in AdSenseLog is easy - you can simply add a custom table report with all the combination channels, and see the difference. Obviously you should wait a little bit till you have enough traffic… otherwise the stats might not be relevant.
If you want to read some more about split-testing, see this post in ProBlogger.