Ideally you could create both, do some split testing and see which one converts better.

From my experience (I work at a freemium startup) I recommend going with one bundle plan. Fewer choices is usually easier otherwise they're worrying about what they need. You can always add another plan, but it's more difficult to take them away (Don't take candy from a baby!). Always try to minimize the amount a user has to think and how much pain it is to upgrade. Also, less important than how easy it is for your users, but important nonetheless: as a developer it's also more difficult to support and maintain different services/tiers

The key thing is less about what type of plan you're selling, but how well you're converting your users. Make sure you're notifying them when they're getting low on space, etc...

Yeah, good advice. I always try to simplify things, so this makes sense. One thought was that It may be easier to convert users if there are focused cheaper plans that meet a specific need. Wonder if anyone has data on this?

Have you launched yet? If not, I'd suggest just launching with a single paid plan. Something low cost. Then you can add more plans as you grow. Without analytics you won't know the best way to segment multiple paid plans.

Hard to say without more details but:

If your service has a real professional end user (i.e. someone who will make a living off your service), a multiple tier one seems a better fit. For example look at smugmug - they have their free account, a couple of paid accounts aimed at "serious" users, and a professional account aimed at people selling photography.

Evernote has only one paid service but that matches their usage base - either free/casual users, or people who have bought into the whole "second brain" idea. But note that the evernote paid account is still reasonably priced.

thanks. good point on the pro end user. I'm not looking so much for advice on "tiers", but if I should split out the premium services into multiple premium services.

eg: evernote could split their premium service into two services: - supersized uploads ($2.50/mo) - advanced collaboration ($2.50/mo)

Oh I see - you are talking about a la carte pricing. Interesting question - as a user my gut reaction is I would be happy to pay $5 for 5 features, of which I only use 3, so that way you get an extra couple of bucks out of me. But, again the key thing is that $5 has to seem reasonable to me. If you were charging $100 for 5 features and I only use 3, I would definitely want the choice to only spend $60, or else I might decide not to spend any money at all.