They should do their best to stick to the "earn it or buy it" design of the best F2P models. Now, many games have different ways to "earn it" and I think that rift is somewhat limited in how they can structure their "earn it". But first a few things.
Rift is a very PVE, group focused MMO and as such it needs to make sure that its model facilitates that as much as humanly possible.
So how should Rift do this?
Well, for starters, they need to make sure that forming groups stays as easy, or becomes easier than it currently is.
One way Rift can do this is to not limit queues for F2P accounts, but instead give subscribers priority queues. This would have a dramatic effect on subscribers ability to find groups quickly for dungeons and gear up with greater ease than they do currently. Most importantly, this would improve the queue time for subscribers who queue for a dps role, which is notoriously long. How? Well, lets assume that the player base grows proportionately, in terms of roles. In low level queues there are probably 10 tanks, 15 healers, and 100 dps (subscribers). Now, if rift goes F2P there will be lots of new players, so add 40 tanks, 60 healers, and 400 dps to the queue. For tanks and healers (subscribers) the queue might improve a little, perhaps instant rather than 5 to 10 minutes for healers, but for dps their queue time should be cut from 1 and a half hours to 20 to 30 minutes, probably less. This has another effect. The 400 F2P dps will feel compelled to switch to tanks and healers, or at least have a spec of one that they queue for because subscriber DPS will constantly cut in line ahead of them, which will further expiate the queue time for subscriber DPS. I suggest having an alert for F2P accounts that says "The queue time is faster for tanks and healers" when they queue as dps or support only. This, in the long run, will help queue times for the game as a whole and improve the experience for subscribers.
What should rift sell?
Rift is somewhat unique in that it has some extraordinary customization in both appearance and player housing (spec too but that's something else). Both of these are veritable fields, ready to be plowed, planted, and harvested by Trion. Countless costumes and dimensional items could be sold for cash currency. Now, some of you might feel that this means development time is being sapped from the rest of the game. However, programmers don't model hats. Neither do game designers. Neither do level designers. And, most of the time that sort of task (unless its tier gear) is passed off to the army of unpaid interns who work insanely hard in hopes that they might actually get a job. I am friends with several un/under paid freelance / interns in the gaming industry from when I lived in California and my wife worked for three companies doing it before she decided on a new career path. They pay next to nothing for fancy hats and justify it by saying, "Well, I am giving those unpaid interns real life job experience that will help them in their career later on". If anything, selling a hat that cost next to nothing to develop for 10 dollars to thousands of people helps them pay their real staff enough money to keep them on the project.
Another potential source of revenue is selling raid keys (not passes like swtor). Instead of paying for a subscription for access, players could buy their content piecemeal. Players would buy permanent access to, say, grim tidings for 25 dollars. Why 25 dollars? Because the next raid will probably be out 2 months after that. This sort of model is attractive to players who don't know exactly how much they will be able to play rift. They might not what to have a subscription ticking away when they aren't able to play, but being able to buy a key they know that when they DO have time, they can always go do it.
In addition to these, there is always the staple of a cash shop, mini pets and mounts. To start, I would place seasonal pets and mounts in the cash shop. Personally, I like waiting and earning these pets / mounts during events but others are willing to spend money and I, for one, and not oppose to them sending money to the game developers who make a game that I play. A fast way to churn out these pets and use as little of the intern army as possible is to simply shrink the models of NPCs. There's already a number of pets like this in rift, like dead Simon. Flood the shop with mini w/e'rs and reskins of existing mounts. People who want to have a rare mount model will fork over money, but it will still lack the rare mount skin so the elites who have it will be able to say "Well. I got the real one."
Now on to the bummer part. Restrictions.
If you go over board with these you will turn off new players. The trick is to never get in the way of the functionality aspect and to always give players a taste of what they could have. Swtor has done a good job of outlining what not to restrict. Professions, group queues, action bars, hide helm / cloak ect ect ect. Do not touch those. Those should not be restricted. Something that Trion could get away with, however, is limiting the number off specs you can have as F2P down to 2 or wardrobes down to 1 or the dye system down to primary colors only or barbershop costs cash currency unless you are a subscriber or you only get 2 random instant adventure bonuses a day instead of 7 or you don't get the bonus item from your first random dungeon. And probably, most importantly, you need to offer them a way to remove these restrictions through cash or hard work (work that helps subscribers have fun). Small things that aren't necessary. Ideally, you want to offer carrots to subscribers instead of sticks to F2Pers.
What carrots should you give?
For one thing, free porting for subscribers. Improved currency gain like +5 infinity stone gain on world events, +2 vanquish mark for random experts, ect. Improved rep gain, +20% rep gain for subscribers. Improved leveling speed +50% experience for carnage quest completion +10% on story quests +20% on rifts, ect. Another thing you could do is give subscribers a +15% run speed outside of combat and outside or war fronts / conquest active zones. Reduced cool down on soul recall. Reduced / no cost to repair your soul. No cool down on guild banners. Reduced cost of lures. You get the idea.
I really can't emphasis enough the need for a tiny stick and giant carrot. The fact is there is going to be a large population of people who will enjoy the game enough to play it but will be unable to justify subscribing due to RL issues like work, school, moving, relationships, relatives, money, other games, friends, ect ect ect. There are a million reasons why someone might be unable to justify a subscription. If you use too big a stick it will outweigh the enjoyment of playing rift and they will stop play. Which means less people in your queues, less people looking at the cash shop (it goes players THEN item sales. you need the players FIRST), and less people helping you close rifts and down world bosses and less people buying your stuff off the auction house and less people to join your ghunt group ect ect ect. You need people in order to have an mmo be fun. Small stick, giant carrot.
Now, the most important part is the currency.
One of the staples of a good F2P model is the ability to earn currency. The ways in which people can earn currency will, to some degree, dictate how people spend their time. If you offer trion coins or w/e from doing the daily ghunt quest, more people are going to do the daily ghunt quest. If you offer coins from doing the daily dungeon, more people are going to do the daily dungeon. I suggest giving small coin sacks contain like 5 to 20 coins for almost every daily activity that requires 5 or more people. Then, another huge carrot, subscribers get a buff that increases their coin gain by 200% so they loot 10 to 40 coins. Have minor rifts drop 1 coin, major rifts drop 3, ghunts drop 5, rhunts drop 10, ect ect ect. And have all these gains doubled if you are a subscriber. Then price things, depending on the item, pretty high. Within a determined F2Pers reach, but just outside the reach of a lazy one. That way, people who want to unlock another wardrobe slot can say "I can earn this!" and they will go out and do lots of group activities, making your subscribers happy (as they will be able to find groups easier and faster) and lazy ones will say "well, its only a few dollars."
Ideally, you want to have F2P be an improvement for your subscribers. If something in your model isn't improving the life of your subscribers... It probably shouldn't be in there (looking at sticks right now... Some tiny ones are meh but... Don't be a swtor.) Another thing is, you really want to understand your potential customers. If they aren't playing rift right now, will they want to play it when it's f2p? Depending on how far you go with sticks, prices, and currency gain that answer will change. If you have too many sticks people won't bother. If your prices are too high people will say "well, i don't like grind fests just to be able to play the game". If your currency gain is too low people will again, feel its a grind fest. You can almost always never have "too few" sticks but you can sometimes devalue items by underpricing them. You want them to feel valuable, hard to obtain, take work. That's why people like rare mounts.
(pst, elrar. you don't have to comment on anything but if you read this just write "I have read this")
http://www.penny-arcade.com/patv/epi...rotransactions <- basically video version of this.