If you want some good data, you'll really want to read this: http://particletree.com/features/web-app-autopsy/

Wufoo & BlinkSale are interesting cases. I was surprised to see their conversion rates so transparently out there. But services like Freshbooks, RegOnline, RescueTime & 37 signals stuff are not really what I'd call part of the 'Freemium' business model. The free version is either limited in time or functionality to the point where you can't be a 'real' user (like you can with Gmail or Dropbox) without being a paid user.

Even BlinkSale is kind of borderline case. Serious users would 90% of the time need a paid account. Wufoo is also near the 'sample' end of the spectrum though it looks highly usable with a free account.

Nothing wrong with that of course, non one says you can't sell services.

Also they're all in the business tools arena. That's definitely an area where paid tools are likely to succeed. But it's an area that (as you can see from those examples) it's pretty straightforward to sell software subscriptions. But they really don't get us anywhere near a facebook premium.

Flickr is potentially a good example. It's consumer oriented & I'd guess the great majority of users are always going to be free users. But is it earning it's salt this way yet? Or is it still in the wait & see category with Google Apps Premium?

BTW I downloaded dropbox today & I can see myself paying for the service, but also see myself using the free version.

FWIW, we have tens of thousands of free users at RescueTime who seem pretty enthusiastic. We've got a mess of paying customers, too-- but I've often pondered the idea of crippling the free version a bit more.

If you're looking at selling stuff to consumers, rather than businesses... I think you're in a scary world (unless you're an iphone app and you can offer the 1 click buy). No matter what you sell, there are always a dozen idiots who will give it away and throw adsense on there-- and consumers are well-trained at this point to find the free alternative.

Have you given much thought to variations on fairly standard business models?

We've applied for YC this round, and one of the things we're trying to achieve is to successfully adapt a freemium model for a product that has been traditionally something that customers would pay up front for. The market is already one that is very profitable and our philosophy is that eventually products in our market will be free, so we want to be the ones who give it away first. We want to do this not only to experience the benefits that come with the freemium model (adoption + potential viral growth) but to also help insulate ourselves from competitors doing it to us.

Without going into too much detail, after some recent discussion (or perhaps it is just paranoia speaking, this occurred about a week ago) we feel that while a freemium model is still the best framework for our particular concept, it comes with a couple of drawbacks¹ that we hadn't originally conceived, so we have had to think a little outside the box.

Right now, we're considering if a hybrid trial/freemium² model might provide us with what we are looking for. Essentially we envision this to be where a user would receive a fully featured product at the start of their product lifecycle, which after awhile degrades into a functionally sound but less featured and free version. The absolute last thing we would want to do is annoy a user by handicapping their product into something unusable.

We think this might provide a user more of an incentive to want to upgrade to a paid-for product beyond the simple fact that they have access to more features/bandwidth/data, while still maintaining our core freemium philosophy.

Also given the nature of our product, we do have room to add in some further functionality in which the addition of a surcharge on further transactions could be warranted, but we would have to further explore the specifics, so perhaps that is another option available to you?

Notes:

1) If PG is reading this and is seriously considering our application and needs to know more about point 1 above, please feel free to email me and I can explain

2) I can't name anyone offhand who is using a trial/freemium hybrid but would love to know if anyone can point me to someone who is using this (or if we're just over engineering the problem).

From what I've gathered, there are more examples at the other end of the spectrum closer to the 'free sample' - free users get a good, usable trial, but premium users are what it's about. I'd call that more of a sales method then a business model.

Anyway, from that I'd say that a hybrid trial/freemium² is a reasonable idea. I would put more chips on the product if it's for businesses.

http://www.wired.com/techbiz/it/magazine/16-03/ff_free?curre...

I guess his book isn't out yet but it looks like he's blogging about free @ the long tail:

http://www.longtail.com/the_long_tail/

Advertising is an old one. We know that one from the TV/Radio/Magazine world. We also know it's not capable of funding every type of popular free web thing. (Great on Search, less great on Facebook).

Cross subsidies - Also pretty straightforward. Real estate site provides tools, Google gets people to like them etc. etc.

Freemium, on the other is not as well trodden. Especially if your strict about excluding free samples from the category.