Using an 1890s theory (Price Elasticity) to drive up your website profits

But a few tips:

a) Attrition, again, is the name of the game. Use that as your metric, look at trailing trends. Typically we've observed the highest drops in between the first 3 billing cycles, after that attrition rate seems to stabilize. If you're lucky attrition stabilizes after the second rebill cyle and you can better manage front end testing and customer aquisition strategy.

b) You're at something of a disadvantage w/price elasticity testing w/SAAS because you're going to get press, people are going to hear about you / hear about your pricing and write reviews early. If you change pricing UP, the internets will scream loudly :). Plus you have to worry about current customers getting upset, always implement price changes DOWN accross the board, always grandfather price changes UP to current users (and tell them about it so they feel warm and fuzzy).

A strategy we've used before is to put together a meaningful ad budget for PPC, and test the heck out of the front end pricing. Treat it like traditional elasticity testing, expecting that rebill #s will hold relatively steady on the back even though you can't predict that. The point is to get the highest front-end you can and then work it down, it's obviously much, much easier to lower pricing than raise.