The part about 1,440 more new trial signups equaling over a million dollars per year is plain silly. Not all trials will convert into paying customers--especially when plans start at $49/month.
The main lesson you should take away from this is to A/B test your hypotheses. (By the way, that means starting with a hypothesis.)
Also, keep your expectations in check. Reading stories about a single test resulting in +$1,000,000 will delude you into thinking that every test will result in huge improvements. In reality, most tests don't have a conclusive winner. Optimization is a process... It is not a one-day or get-rich-quick affair.
Edit: Apparently it's not OK to mention I'm a marketer...? I removed that bit. If you're downvoting for some other reason it would help me if you explained.