Freemium isn't a business model, it is a marketing plan. In some cases giving away a subset of your product for free will cost you less per paying customer acquired than any other method of advertising / sales.
If that isn't the case then you should stop giving away the service and spend that capital on the other more profitable methods of acquiring customers.
There are some technical advantages to charging users from the outset that he does not mention. A whole class of spam problems goes away, and your app doesn't fill up with ghost accounts from people who signed up just to take a look around, never to return.
He doesn't but that's because he's talking about a model where you give away something for free and charge for a better or more featureful version of it.
Charging everyone (even a nominal amount) from day zero is great but has a narrow application - it mostly works when people perfectly understand what your product does. In the case of pinboard, 'del.icio.us minus the suck plus some other stuff'. The original del.icio.us would have got exactly nowhere charging users from day 1.
The original delicious wasn't trying to pursue a 'freemium' business model, though. It was pursuing the 'grow big fast; get bought' business model, and caught it.
I think for the lowly startup that is actually trying to extract money from customers, extracting it right at the outset is an attractive choice, even if it's a token amount just to filter out spammers and deadbeats.
I'll confess to being totally confuzzled between people continuing to think there is some huge, categorical difference between freemium and free trials, which strikes me as a mostly semantic distinction. My confuzzlement aside, I think there is probably an alternate causality for having godawful terrible conversion from beta users. It is: "My business recruited beta users from TechCrunch readers."
There's a very real difference which is that free trials end and free use of some feature subset does not. The difference isn't very big when the supposed 'free' product is so crippled it is next to useless compared to the pay product. But that is just a (lousy) implementation choice. The free Flickr, for instance, is a full-featured, perfectly usable photo sharing service.
Traditional free trials aren't always crippled; it used to actually be standard in software that the free versions were useful in their own right, and the rise of "crippleware" that was basically useless on its own was initially fairly controversial in the shareware community.
I mean, Doom could've been seen as a "freemium" FPS, where you get the entire FPS engine and several levels free, and only pay if you want to buy bonus levels.
(Admittedly, that's software v. service, but in the same era, some BBSs ran on a service-based freemium model as well, providing free access to anyone, but letting users pay for certain features like extra download quotas or time quotas.)
I didn't say that. In fact free trials are often fully functional. But trials expire. That makes them different from 'freemium'.
Shareware is a little different. There are at least two kinds - shareware that is limited and its purpose is largely demonstration to entice you to buy the full product. Doom was that kind of shareware, the stuff you paid for was not the 'bonus levels' it was the actual game.
There's also, less commonly, shareware that's useful as is but certain extra features cost money. This is the closest analogue to a 'freemium' web service.
The original question, though, was 'what's the difference between a free trial and freemium'. And the answer remains the same, free trials expire, freemium done right is a useful standalone product in its own right that you can use for as long as you like.
Does "freemium" strike anyone else as "shareware"? Limited functionality until paid for? Is it just an attempt to shrug off the bad image shareware has acquired?
edit: no, really. I'm asking. Is there a conceptual difference, or is it just using a new buzz-word to boost perceptions?
Shareware doesn't have a bad image because developers ask for money. Shareware has a bad image for numerous other reasons: perceived lack of professionalism on part of developers, low quality, (largely inaccurate) association with viral infections, a mindset determinedly stuck in 1996, and the fact that shareware has no meaning to most of the people on the Internet.
I would never, ever say "shareware" within earshot of a customer. (Though "freemium" doesn't sound to me like a customer-friendly word either.)
Shareware doesn't have a bad image because developers ask for money.
Wasn't aware I was implying it did. But it certainly does have that malware'd crapware image (however incorrect it may be for each individual app. There's plenty of good shareware out there).
It's pretty similar, I have another comment on this thread talking about this so I'm going to try not to repeat myself but a basic difference is that 'shareware' is a much broader term. Shareware can be anything from fully functional 'nagware' to a useful product with extra features in the pay version to a product with limited features with the limits removed in the pay version. And there are probably another 30 variations.
Freemium is a web service where ideally the free version is useful and you pay for certain additional features or a reduction in the limits of existing features. This makes the model like that of some shareware. You're generally buying a service so the something-ware name becomes even less sensible.
That's not to say that 'freemium' is not one ugly gargoyle of a word that I hope drops into the same oblivion pit where we left 'folksonomy' and other monsters.
If that isn't the case then you should stop giving away the service and spend that capital on the other more profitable methods of acquiring customers.