I'd recommend testing your hypothesis. A pure split test (A/B) doesn't quite seem right, because it'll be more difficult to compare w/o doing some sort of mult

I'd recommend testing your hypothesis. A pure split test (A/B) doesn't quite seem right, because it'll be more difficult to compare w/o doing some sort of multivariate regression. I'd suggest this: peel off a small segment (10%, 20%) randomly out of your user base. Give them the option for monthly / annually. Compare the revenue from that group to the revenue from your control group. Even if you do anger a few of your customers, or the monthly plan costs you a lot of sales / revenue, you've only done it with a small section of your audience. The only downside to this is that you might find some customers prefer the monthly plan, even if you only want to offer the annual (after testing, you find that annual is more profitable to you). Each site is different. Each customer base is different. This seems to me like a situation where you should collect data, rather than look for high level advice.