So you run 300,000 banner impressions on this site. Then when that ad campaign is over you look at each of the three banners and see these results:
Banner A 0.7 CTR
Banner B 0.3 CTR
Banner C 0.9 CTR
What do you do next? Well the first thing you realize is that if you are going to run a banner campaign on another web site, you are going to give them Banner C since it gets three times as many visitors as Banner B, all for the exact same price (if it's a CPM campaign).
But you don't want to rest on your laurels with Banner C. You want to look at it and see what may have made it successful and the others not so successful. Then you come up with two additional banners D, and E,
Maybe D takes what you think are the strong points of Banner C and improves upon them and maybe with banner E you go in a completely different direction.
So, you get another ad package for 300,000 impressions and run banners C, D, and E and get these results.
Banner C 0.9 CTR
Banner D 1.2 CTR
Banner E 0.7 CTR
Are you starting to see the power of incremental improvements? You can now take Banner D and use that to advertise on other web sites. What if, originally you had used just one banner and it was Banner B from the above example? You'd be getting a ¼ of the visitors for the same price!
All right, so by using this method you can increase the number of people who come to your web site. But how about once they get there? Are there things you can do to improve your conversion rate? Absolutely, the testing doesn't end once you get someone to your web site. In many ways, it is just beginning.
Let's keep going with the example above and say you have a video game web site. Let's say your main page (your home page) features all the different type of video game systems out there (Nintendo, PlayStation, XBox).
Let's also say you found out (from your testing above) that a banner featuring the Nintendo GameCube receives the best CTR of all your banners. You also know from the banner campaign that when you point that banner to the home page, you get a 1% conversion rate. It's time to focus on increasing that conversion rate.
So, the first thing you do is instead of featuring all 3 video games systems equally, you make the graphic for the Nintendo twice the size as the other two video game systems. Viola, your sales of the other two systems stay the same but your sales of Nintendo GameCubes go from 1% to 1.3%. Nice work - you just started making 30% more sales!
But don't stop there. What happens if you offer free shipping on the GameCube? Bam, orders go from 1.3% to 1.5% (more than offsetting the shipping charges). You're on a roll.
Let's try increasing the size of the PlayStation picture to see if we can increase PlayStation sales the same way. Oops - PlayStation sales stay the same but Nintendo sales go down. Not a problem, just undo the changes and try something else. Not every change you make, maybe not even a majority of them, are going to be an improvement. But you don't keep the bad ones, only the good ones, so you are always getting better.
It's important to remember that you can make incremental improvements to your site to increase sales if you are a content site as well, it's just as important.