Our first series of web optimizer tests (we have done over 20 test for our first client) have been a tremendous success. We have been focusing on optimizing a client's SEM landing pages with the hope of driving more traffic from the landing page through to the product description page. Our tests experimented with different body text, headlines, calls to action, and banners.
It was fascinating to see what worked well for certain product's landing pages didn't work well for others. It was also interesting to see the difference that each test component had on affecting conversion rates. In our experiments altering the banner area had very little impact on conversations, the body text and headlines had a moderate impact on converstion rates , and the call to action (no surprise) had a large impact on the conversion rate.
Overall we have been able to drive about 100% more traffic from the landing pages through to the product description pages (in some cases over 175%). That means without increasing their PPC marketing spend at all we have been able to drive twice as much traffic to the product description pages. That makes for a very happy client.
Interesting notes from running the experiments: Pages that don't get a lot of traffic can take a very long time to get any results, so you should test fewer variables and fewer versions on these pages. Otherwise you might never complete your experiment. Don't assume because something works on one page that it will work on other or all other similar pages. You've got the tool now so test everything.
Now you will be able to test the CEO's great idea without any fear of it being good or bad as you'll have the data to compare to the other ideas that come out of marketing, sales,…