It seems that for every successful online game that holds a large, dedicated user base for a long period of time, there are ten that are virtual ghost towns 6 months after release. This holds true across every genre: from FPS to MMO, there are more abject failures than there are hit franchises. What does it come down to? What separates WoW from Auto Assault, Counter Strike from [insert one of many barely played FPS' here]? To determine this, I decided to look at the Valve model and break it down into the steps that the developer has taken to keep players interested in their product, bring new players in long after release week, and get players to continue spending money on the respective developer's games and services.
Lets look at Valve:
Valve has a proven track record with their games enjoying large amounts of user support even almost a decade (see: Counter Strike) after release. But it's beacon of continued support, it's current flagship of fanatic fanboys, is Team Fortress 2. After almost 2 years on shelves, there are still 1562 servers, by my count, with users playing on them in the middle of the day. They must be doing something right.
Teach a man to fish.
When the game was first released, it shipped with only 6 maps. They could have been the 6 most well balanced, insanely fun, incredibly varied maps in the world and it still would have been a weak effort on Valve's part. Of course, Valve's reputation and the merits of the game and the original 6 maps made it much less of a problem. But there is always dead time after a game's release and the first patched-in maps. If this dead time lasts too long, users will become bored with the game and move on to the next big release. This is where user-created content saves the day. The Source Engine SDK is not the most approachable piece of software to the regular user (something Valve SHOULD fix at some point), but there have been a great number of developer quality or (gasp!) better custom maps pumped out since TF2's release. The benefit is twofold : people keep playing the game without too much complaint of lack of content, and the developer gets a chance to see what, how, and why the users play the game, then they can create new maps based on the player base's habits and wants.
Another patch? I just downloaded one!
With (according to Steam) 80 updates in less than two years (4 patches a month by my count), it's hard to put a version number on each patch for TF2 (I'm sure Valve does so to some degree, but the user is generally oblivious to it). As problems arise they are fixed, and new content is added on a semi-regular basis. It's easy to find what is changed with a new patch, and best of all every patch is downloaded automatically from Steam. Awesome. Which brings me to my next point.
Location Location Location
Would you want to buy a house if you had to drive 10 miles to buy food and 20 miles to use the phone? You want the things you use close and easy to access. The same principle applies to Steam. It's been a long, winding (annoying, frustrating, x-fire filled) road to the friends list, voice chat, in game chat/community/web browser overlay, and other necessary features in this day and age, but Steam is now an all-in-one, free platform that is all most gamers need. Yes, it is missing features ( ex. selling your old games ... somehow) and great developers. But lets get back to how it affects TF2. Without Steam, do you really think Valve would be able to release 4 patches a month for a single game? Without Steam, would it be so easy to hop onto the server a friend is on, or to compare achievements and stats with said friend then rub their face in it over and over again? Of course not. While I think that every developer should release their games on Steam, it's not realistic. So make something better. Stardock has Impulse (I haven't tried it, I've heard good things though), other services exist, and competition is only a good thing for the market.
NOTE TO EA: Try harder. EA Downloader isn't that great.
If you've got it, flaunt it.
Game advertisement is generally a simple thing. A few commercials, a website, word-of-mouth, full page ads in magazines, media-exposure, and annoying banner ads are pretty standard. But most games are only advertised for a few weeks after release. In contrast, the large class updates and accompanying free weekends keep TF2 on the media radar, and things like the still ongoing Meet the ... series are genuinely entertaining. The free weekends are especially large boosts of advertisement, and it would interesting to see the TOTAL sales stats that accompany these (Valve only releases retail sales stats, not Steam sales stats). Sure it costs money to keep adding meaningful content to your game, make new videos showcasing it, and keep the game's website updated, but this cost is then outweighed by all the free advertisement TF2 receives in gaming news and all the new users that end up picking up the game. You have to spend money to make money.
Oh yeah, the game also has to be great to start with, and the added content has to be top-notch. No pressure.
That's about it. I think I've hit all the high points, but I'm sure I've missed important components of success. Also pictures. And less commas. Comment as needed. read