I'd argue you should measure how active does a user use a product and only charge them when you see they actually make reasonably heavy use of it. An inactive

I'd argue you should measure how active does a user use a product and only charge them when you see they actually make reasonably heavy use of it. An inactive user doesn't cost much and can convert into an active-and-paying user once. A slightly-active newcomer doesn't cost much either and can actually bring value by reporting bugs (incl. UX bugs like unintuitiveness). I wouldn't pay for what I had no chance to try. A free month is also useless: I just sign up and then see it expire as I fail to find spare time to actually try it. Just measure how active the user is and tell them they will have to start paying soon once they reach a certain threshold in median daily interactions or in storage/traffic/whatever. Don't forget to provide an option to export and delete their data.