It is interesting to watch buzzwords appear and disappear. Many live just for a few months. Usually, when they are gone, they never come back. Sometimes, however, it happens.
Freemium is one of those buzzwords. It is the model of giving away something for free, get people to use your service (hence, making the free version quite interesting) and then upsell a minority of people on additional features. A combo of the words free and premium.
It is the model used by Flickr, for example. You can use the service for free, but they will start hiding your old pictures if you do not pay, once you have a lot of pictures. Similar for LinkedIn and others.
Users like it. If the value you deliver for free is good, they use it happily. And there is nothing better than free. If at a certain point the value you deliver can be even better and is worth paying, people are happy to pay. They do not feel like you are robbing them, or nickle-and-diming. They are genuinely happy to pay (at least, that works for myself, I like free but I do not mind to pay for a good service, like Flickr for example).
The other model, of course, is advertising. But it is intrusive, in most cases. And you need to give up your personal info to an advertiser, which many do not like (I do not, sorry, I will keep using Adblock, I do not care if it hurts the content provider, it is my life you are playing with...).
Now, freemium was very hot a few years ago. Then it became very cold ("you cannot create a large company with freemium models"). Now it is back. LinkedIn is making a lot of money, and it will IPO soon (my prediction). Some interesting variations include having people pay for virtual goods (nobody beats Zynga in it, Farmville and Fishville are delivering hundreds of millions to them).
Guess what? You can build a large company based on a freemium model.
One place where freemium could work very well is mobile, in my opinion. I do not believe the pay-per-app model can create many large companies (or even one). What the AppStore is delivering are $0.99 apps, and nobody can make money with it, excluding a guy in a basement who is fine with making a decent salary. However, if your apps are free and you can charge along the way (freemium), that is a different story. Your adoption rate will be 1,000 times higher, and you just have to find a minority willing to pay a premium (note that only Apple supports this model, Android does not - yet).
Many believe the people who pay for Farmville are just dumb. I don't. If you buy a Wii game, you pay $50. And nobody thinks your are dumb. If you spend a lot of time in Farmville getting entertained (ok, that might be dumb, I get it), it seems very reasonable to me that you spend $50 for it, even if it is one dollar a week. And if you have 400 million people that could play, then it is easy to see how your revenues can be in the hundreds of millions.
Mobile is just an extension of your desktop life, with the slight difference that you have your mobile device with you every single moment of your day. Forcing people to pay, maybe on a per-month fee, it is a good dream. You can definitely do it in the enterprise, with SMBs and prosumers. But when you go to the large crowd, it is going to be harder and harder. Still, the market is enormous...
In my space, Apple charges $99 per year for MobileMe. There is a lot of room to cut that price (and it works only for the iPhone, so good luck if someone in the family has a different device). At the other end of the spectrum, Google charges zero for Google Sync (albeit it is quite a bad product, sometimes free can be of a depressing quality...). How do you move between these two extremes if you are a carrier? Per-month, per-year, free, advertising or freemium?
I say, for now, stick to per-user per-month on the high end of the market, and check freemium for the masses (they are coming).
Long post just to market Hal's last paper for mobile operators "Using Free-nomics to Avoid Pipe-ification". It is free as in freemium, since he expects you to pay one day, I guess. Check it out, it is worth it.