I'm currently building a small online service and have a problem when trying to decide the pricing model. As far as I can see, there's three possible models:
1) PAYG: you get an invoice monthly which lists what you've used, and you pay it. Can also do pre-pay (Tarsnap, AWS)
2) Tiered pricing: you pick a tier, and pay a fixed amount every month that gives you a usage limit. (Github)
3) Tiered-As-You-Go: a bit of a blend of the two, it's like PAYG but the usage billable 'units' are fairly chunky, so you get a mostly consistent bill but with the flexibility to use more on-demand.
I've read a fair bit about why pricing tiers make a lot of sense from the business side. There's lots of stuff around psychology and maximizing profit out of each customer and that all seems to make sense. However, I can't get away from the fact that as a consumer, I hate tiered pricing. I usually end up feeling that I'm being ripped off, unless I use everything that my current tier offers. The tiers hardly ever fit my needs, and it's generally not a nice feeling. And once I do get to the limit of my tier, the next small thing that I need equates to a big jump in cost, because suddenly I need to be a super-platinum customer, even though I don't need most of what that offers.
So my question is this: If you were building a startup now, what factors would influence your pricing model? Is there anything that would make you rule out PAYG or pricing tiers completely?
I know how you feel about getting the 'ripped off' feeling from tiered pricing. I think as developers we're getting used to the AWS model of paying exactly what you get.
But at the same time, you might be in a market where customers might think that in a PAYG model they're being nickel-and-dimed for every little thing. It all depends on their expectations.
So in the end I'd figure out what your customers are accustomed to. I'd also read, if you haven't yet, Patrick's feedback on Tarsnap's PAYG model: http://www.kalzumeus.com/2014/04/03/fantasy-tarsnap/ He brings up some good points I hadn't considered.