Do you think that if you A/B tested your favorite choice from this service that it'd meaningfully change your success?
Probably not, honestly. I'm currently in the process of setting up a graphical A/B test for all my conversion buttons, which I think will majorly move the needle and give me ways to move it in the future. (Tweak button texts, tweak button colors, tweak button orders, etc etc. These are all hard for me in the status quo because I don't have the buttons ready to go, and I have the graphic design ability of a drunken lemur.)
This might not be obvious, but all A/B tests are not created equal in terms of implementation difficulty. The cheapest A/B test is a change to page copy: it is literally one line and done. I started three of them last night, testing microcopy in my shopping cart such as "You do not need an account to pay with a credit card through Paypal."
Buttons are a wee bit harder to A/B test because they're often seen in many pages throughout my site and, stupidly, they're not currently implemented as some sort of render :partial => '/static/easily_replaceable_buttons'. That means it is going to take me a couple hours to get the button tests ready sitewide, and I have to disable page caching to do it.
Yanking my logo is harder still. In addition to being on every page in the site (which means I have to turn off page caching entirely, which is going to require some thought about performance consequences), there are a few places where it is used offsite, for example in the top bar of my Paypal payment pages. Changing logos right at the shopping cart is probably not a great idea. Changing that sucker will require an hour or two of work, for boring technical reasons.
Probably not, honestly. I'm currently in the process of setting up a graphical A/B test for all my conversion buttons, which I think will majorly move the needle and give me ways to move it in the future. (Tweak button texts, tweak button colors, tweak button orders, etc etc. These are all hard for me in the status quo because I don't have the buttons ready to go, and I have the graphic design ability of a drunken lemur.)
This might not be obvious, but all A/B tests are not created equal in terms of implementation difficulty. The cheapest A/B test is a change to page copy: it is literally one line and done. I started three of them last night, testing microcopy in my shopping cart such as "You do not need an account to pay with a credit card through Paypal."
Buttons are a wee bit harder to A/B test because they're often seen in many pages throughout my site and, stupidly, they're not currently implemented as some sort of render :partial => '/static/easily_replaceable_buttons'. That means it is going to take me a couple hours to get the button tests ready sitewide, and I have to disable page caching to do it.
Yanking my logo is harder still. In addition to being on every page in the site (which means I have to turn off page caching entirely, which is going to require some thought about performance consequences), there are a few places where it is used offsite, for example in the top bar of my Paypal payment pages. Changing logos right at the shopping cart is probably not a great idea. Changing that sucker will require an hour or two of work, for boring technical reasons.