Before is more likely to put potential subscribers off, but would keep people that have no interest in keeping the service from using resources. After is poten

Before is more likely to put potential subscribers off, but would keep people that have no interest in keeping the service from using resources. After is potentially more likely to yield a higher conversion rate (if your service has enough hooks to keep the customer in) but you're opening the service up to abuse. Perhaps another thing to test is this: Let the user sign up with a card, and during the trial period ask the user for feedback on how the service could be improved, either through a survey or free form. This at least lets you engage in a conversation with the customer and identify whether or not they're likely to convert. Have a call to action to convert to the full service so the decision to enter details is in the user's hands. Measure the responses versus the other method, and post back here :)