Testing can be conducted sequentially or in parallel. In sequential testing, often the easiest to implement, the various versions of the webpages are made available online for a specified time period. In parallel (split) testing, both versions are made available, and the traffic is divided between the two. The results of sequential split testing can be skewed by differing time periods and traffic patterns in which the different tests are run.
Imagine an autoresponder that doesn't just send emails, but allows you to track which channels your visitors are coming from, segment them based on actions they take and who they are (what they do in your funnels, how socially connected they are, what they purchase and more!), create custom follow up sequences (email, text messages and more!) for each visitor and FINALLY see the TRUE Lifetime Value of each of your customers!
Of course, the address itself won't be enough to estimate the value of a home. It just denotes the home's neighborhood. That's why the next page follows with more questions about the property itself, like number of beds and baths. Below, you see the copy "Tell us where to send the report" -- with a disclaimer that, by entering this information, you're agreeing to connect with a real estate agent. This is a great example of a company giving value to their visitors from the get-go, while setting visitors' expectations about what will happen as a result.
I'm not sure how the algorithm works (or if there's one at all), but while I was filling it out, I had some anxiety about not qualifying. Once I found out I did, I was excited to fill out the form, which I'm sure most people who are in debt and using this tool are. By making this offer seem more exclusive before the form appeared on the landing page, I'd bet that Bills.com increased conversions pretty significantly.
Open-ended experimentation. This approach is similar to closed-ended experimentation except that more variations will be added for testing and experimentation will not stop when a winner is found. This method is used by large corporation to dynamically improve their conversion rates and improve user experience. Landing page can also be adjusted dynamically as the experiment results change to further enhance user experience.