We are leaving Google Wallet

The problem then becomes that all your subscription logic is handled via Stripe, so at that point you have to move your data to something like ChargeBee, or you have to set up your own in-house solution that handles working with PayPal's own recurring payments authorizations for some accounts, or Stripe's subscription features on the other.

We honestly haven't gotten far enough to decide what we're going to do about this yet. Moving to something like ChargeBee would make things simpler, but it adds an additional cost on top of what we're already paying Stripe/PayPal, plus we have to migrate all of the thousands of customer's data out of Stripe and into something else...

Firstly, I wanted to bill in multiple currencies according to where customers come from, and Chargebee required multiple API accounts, so that there could be no unified reporting of activity. Stripe, on the other hand, is perfectly happy with USD plans/subscriptions living alongside those in EUR or GBP.

Secondly, if you start with just Stripe, you can add other added-value providers like Chargebee on top later, but once you're committed to managing everything through those providers, it's hard to go back.

And finally, it would be easy to assume that you will need all that fancy Dunning process and workflow magic, but setting it up is an additional up-front time & complexity cost which might never pay for itself. Better IMO to start off simple, write a little code to track failed payments and expiring cards, and email the customers directly the old-fashioned way. This way, you're also connecting personally with your customers.

What I'm wondering though, is they say Wallet is being focused on mobile devs, but what if I'm a mobile dev that also wants to offer a subscription service or something else that they identified as a problem for them?