Hey colinp, your backend AWS costs have come down, so you SHOULD be charging less for tarsnap [1].
Hey you (yes, you), ignore everything you've read about surviving as a startup. You SHOULD be charging your customers based entirely on your cost base and ignoring any conversation about the value your customers receive.
[1] Is that even possible?
Tarsnap is tragically underpriced, and we have years of its business track record (compared to every other well-known backup provider) to see that the strategy of delighting the kinds of people who account for their storage costs to the picodollar does not result in runaway success.
Colin does fine for himself, but that's because he pours an avalanche of competence and expertise into what he's doing. If I were him, I'd be furious that I was getting lapped by people who do a much poorer job. But probably also hesitant to accept that my own low pricing had something to do with that.
You'd only be getting lapped if you think life is a race to make money.
I'm doing Tarsnap because I wanted to build a good product, and by that metric I think I'm far ahead of the "competition".
Colin, I don't think I've had a chance to acknowledge you for supporting / allowing Patrick to write up the article I linked above. That was a brave call, and few business owners are bold enough to stand up while someone calls their baby ugly. (I think it was a masterpiece article on his behalf, and that Thomas makes a good point in his response here too AND I support your ongoing choices. I'm on the record as a long-standing fan of the picodollar.)
Tarsnap is, at least on HN, a great case study in the pricing and product choices startups have. I would make different choices regarding pricing (hence my grandparent post), but I would never have the capability to create a product remotely comparable - and the product is more important.