Anyone care to comment? I can think of some potential ones, but have no first-hand experience with them, so I'll withhold comment.
However, my biggest fear is derived from self-awareness. I run a startup and, no matter how good I think our code is, we fuck stuff up. Plenty. Oddly, in this sector, I trust an older company full of idiots more than a brand new startup full of brilliant hackers, if for no other reason than the idiots have already made their mistakes and are now just not gonna fuck with it. If they made it a few years, ironed out the kinks and then just they leave it alone, it's less likely to break :-P
The biggest one is Stripe does little beyond actually charging a card. If you're a business, you likely need to provide your customers with receipts via email at the very least, and ideally have a portal for them to look up their payment history. You probably don't want to bother with Dunning emails, either. If you go with Stripe, be prepared to implement all this yourself.
I can't speak fully for the others, but I have used Chargify and Recurly. Chargify finally has statements. Recurly has a large emphasis on making sure your customers are retained -- so much so that they added their own payment gateway to help make sure that's the case.
So, Stripe may be a low cost option on a per transaction basis, I just wouldn't brush aside the development costs that some of the other providers save you. In the end, Stripe may still be worth it for you -- it is for us. But that's the catch.
I said "If they made it a few years, ironed out the kinks and then just they leave it alone, it's less likely to break". I have the same concerns about all new software from any provider. The same concerns applied to Google Checkout and Amazon Payments when they were new.
I just happen to run a startup & have, as expected, observed this fact about software engineering.
However, given the smaller teams and conflicting demands on attention for a startup, the problems may be more pronounced. Furthermore, billing is an industry filled with high degrees of both incidental AND inherent complexity. Navigating that complexity requires institutional knowledge that must be built over time.
Anyone care to comment? I can think of some potential ones, but have no first-hand experience with them, so I'll withhold comment.
However, my biggest fear is derived from self-awareness. I run a startup and, no matter how good I think our code is, we fuck stuff up. Plenty. Oddly, in this sector, I trust an older company full of idiots more than a brand new startup full of brilliant hackers, if for no other reason than the idiots have already made their mistakes and are now just not gonna fuck with it. If they made it a few years, ironed out the kinks and then just they leave it alone, it's less likely to break :-P