When making decisions for a webpage design, some are purely aesthetic or purely functional decisions like page organization and color scheme. But other decisions lie in the scary realm of user experience. For example you might wonder, 'How big should the donate button be so people can easily recognize it?' Or, 'What should the follow-up text on this signup page read?' Even, 'How many exclamation points are acceptable when announcing we've exceeded our fundraising goals?'
Most of the answers to these questions are studied by researchers who specialize in understanding user experience online. You could plow through stacks of research to find out optimal size and color for donate buttons, or, you could just A/B test it!
A/B testing was pioneered and is still widely used by Google to test two or more options in user experience. The researchers and programmers at Google readily admit that there is just no reason that some things work on a website and some things don't. What A/B testing gives Google is real feedback in the form of hard data on what works and what doesn't.
But you don't have to be Google to get your hands on A/B testing. Optimizely is a service (of varying pay scales) that offers A/B testing services on any webpage. You can create 2, 3, 4, even 6 variation on a webpage, and Optimizely will pick a random sample of your traffic to sample each page, giving you real results in no time.
We like this service because a user will never know that they are using a test site; there is no indication on the page that it is part of an experiment. And results can be used in the long term for future projects, so both our team and our clients learn the best ways to set up donate pages, petitions, sign up pages, and tell a friend pages.
So when you can't decide whether that new page on your site should have one header or two, skip the debate and headache! Just A/B test it and let the data be your guide!