We tried both ways, and the free trial was a clear winner. I think this could depend on your service, however.
Something else to consider: even though we had more people sign up and pay, we also had far more people sign up and try without paying. Several of those users also gave us free, insightful feedback. Especially if you are starting a new service, this is worth more than the paying customers who give no feedback.
Cost of provisioning and supporting was already mentioned - that's big .
Another key point to consider is the amount of effort required from the user in order to derive value from your product. If the product can be used easily and passively (e.g. Pandora) than a free signup makes the most sense.
If your product requires higher user engagement to be useful - i.e. they need to dedicate an hour or two to working with it, or need to integrate with something, than a refund or out may make more sense - many people treat free trials as worth what they paid for them - nothing. If someone has to pull out a CC, they have more likely overcome the hurdles necessary to engage with your service.
Even if someone has split-tested, the specifics of how much the product costs, what it provides, how much time it saves the user, and its perceived value all make a huge difference on where the split-testing ends up. Apples and oranges.
Comments
I'm asking HN because I'm hoping someone HAS split-tested or can share from experience.
We tried both ways, and the free trial was a clear winner. I think this could depend on your service, however.
Something else to consider: even though we had more people sign up and pay, we also had far more people sign up and try without paying. Several of those users also gave us free, insightful feedback. Especially if you are starting a new service, this is worth more than the paying customers who give no feedback.
Cost of provisioning and supporting was already mentioned - that's big .
Another key point to consider is the amount of effort required from the user in order to derive value from your product. If the product can be used easily and passively (e.g. Pandora) than a free signup makes the most sense.
If your product requires higher user engagement to be useful - i.e. they need to dedicate an hour or two to working with it, or need to integrate with something, than a refund or out may make more sense - many people treat free trials as worth what they paid for them - nothing. If someone has to pull out a CC, they have more likely overcome the hurdles necessary to engage with your service.
Even if someone has split-tested, the specifics of how much the product costs, what it provides, how much time it saves the user, and its perceived value all make a huge difference on where the split-testing ends up. Apples and oranges.