I am playing around with different possibilities to solve a client problem about how frequently respondents would use a product configured with 6 attributes, with 2-5 levels per attribute. The product would be something that could be used on a daily basis under different pricing structures (i.e., pay per use, monthly payment, etc.) so the client wants to know which pay structure best optimizes usage and revenue.
One possibility I considered (but I think I have an alternative plan) would be to show a one concept per screen as a volumetric CBC compared to the NONE level (which could be renamed to be the main competitor as a constant alternative). It seems to me that this is very close to the old CVA.
So my questions are:
1. Is there any inherent problem in doing it like this?
2. Am I missing a "better" method?
3. My "better" method is just to do it as a discrete choice CBC with follow up questions in between each task asking open ended numerically how often they would use it. Then I can do two models, one with the discrete choices (to determine share vs. constant alternative competitors), and one with the volumetrics (to determine usage frequency). Any concerns with doing that?