The purpose of a Pay Per Click (PPC) landing page is to incentivize visitors to get deeper into the site. Let’s say you want to optimize a PPC landing page and design a completely new layout which you A/B test. The results show that conversion rates went down a little but 12% more visitors reached product pages. Should we draw the conclusion that the landing page failed because it didn’t lift sales or should we say it was a success because it brought visitors deeper into the site? If the landing page brought people deeper but those visitors didn’t convert it means the steps after landing page need to be optimized.
Let’s look at another example. You want to increase sales of a particular product so you rewrite product description and run an A/B test. For such a test we should set conversion goal as visits to cart page. The primary job of the test product description is to inspire visitors to add this item to their shopping cart.
My point is that each test has a very specific purpose and test success should be measured based on how well the variation page is able to complete that purpose.