We often run allocation conjoints where the respondent can allocate as many items on the screen as they want, or they can select 'none'. We then create a synthetic none in the data which is the max they allocated minus the amount allocated on that screen. In the end we have a utility/preference score for none. In our simulators we weight respondents' share estimates by the max they allocated to get share of volume.
It can be difficult to explain what the 'none' share of volume actually represents. Especially if you are simulating the best possible scenario where every tested product is available and at the cheapest price- in theory the none share here should be 0% because we have removed people who chose none on every screen, however the none always gets some share. How can we explain this? Any guidances/best practices for handling the none/explaining what it means?