Update: Contrary to reporting, I’m not “overseeing Twitter’s products”. I’ll be a product manager. One of many people contributing to Twitter’s future.

A couple of weeks ago, I announced that I had left Raptr to explore other options. Raptr was an incredible experience for me and I learned a lot there. I watched and helped the company grow from 8 people to the near 30 it is now. (They’re hiring for more, by the way.)

At Raptr, I took on more roles and responsibilities than I have ever had before in my professional life. The team that I left behind consists of some of the most talented people I know whom I hope to work with again in the future. I have a little bit of insight into what’s coming for Raptr and, believe me, if you play any videogames, the service is going to be even better than it already is.

But for me, it was time to take my newfound skills and apply them to new challenges.

Today, I signed an offer letter from Twitter. I’ll be a product manager there working primarily on the web client. As you might guess from my Twitter username (@k), I’m a pretty big fan of theirs and see a lot of potential for the medium next year—for that’s what it is; a medium, not a service.

I’m looking forward to what’s ahead …

1. “Why would I want to know every little detail about what my friends are doing from Twitter/Facebook/Friendfeed?”
2. “How do you not get overwhelmed by all the people you follow/friend?”

My short answer is I don’t treat it like a fire hose I have to drink down, I treat it like a stream I dip my feet in every so often. To explain this statement, I need to first talk about friends and travel.

I was born in Vancouver, spent my formative years in Hong Kong and then returned to Vancouver for the last two years of high school and college before moving to Austin for my first job. All this moving was a mixed blessing. I was able to experience many different perspectives and made a wide range of friends from all over the world. On the other hand, there were few friends that shared my experiences throughout.

Whenever I visited Hong Kong, which is roughly annually, I meet up with my childhood friends. Conventional thinking would say that, because we haven’t seen each other for a year, we’d have a lot more to catch up on than say, someone here in San Francisco that I saw just the day before.

Anyone who has experience with this can tell you that it simply isn’t true. When you’re apart that long, conversation topics feel like they need to be a minimum level of significance to be worth discussing: career changes, marital status change, buying of property, perhaps a new family member, etc. A sample conversation might be like this:

Friend: “So how’ve you been?”
Me: “Great. Things are going well. I got engaged!”
Friend: “Congrats! You still doing that computer thing?”
Me: “Yeah. Still at the same place. You still at the same firm?”
Friend: “Yeah, 3 years now.”
Me: “Wow …”
Friend: “MmHmm …”

In contrast, the friends you see every day or every week are the ones you can talk to for hours. Why? Because any topic is fair game. You don’t feel like you have to filter out the more mundane topics because it’s such a significant event to be catching up with the person. How was that movie? Did you go climbing yesterday? Did you see that crazy YouTube video? No topic is too trivial.

So how is any of that relevant to the information overload of Twitter and Facebook?

To me, Twitter and Facebook updates represent the mundane, everyday conversations that I could and would have with everyone if I could. By seeing the stream of updates from my friends, I have much more context into their lives, and a feeling that I can converse with them about smaller things. To use a clichéd term, I feel more connected to them.

When I see these friends, even after many months apart, I still feel like I’ve been talking to them and keeping up with them to some extent. Conversations flow more naturally and are much more rooted in the present than trying to bridge the gap since we last interacted in person.

I disagree with the fire hose terminology because it’s not something that is pointed at me. It really is a stream of information which I can look at anytime I feel like. When I don’t dip my feet in, the stream flows on, I’ve missed some updates, and it doesn’t matter.

We can look at it a different way, too. Whether we know it or not, each of us probably have at least 300 people we know and like enough to want to keep in touch with. If you saw each of these people for dinner one friend a night, you would see each person once a year. One solution is to simply forget most of these and hang out with the same dozen friends week after week. Realistically, there are far more than a dozen interesting and inspiring people worth interacting with regularly.

So going back to the original two questions:

1. “Why would I want to know every little detail about what my friends are doing from Twitter/Facebook/Friendfeed? ” Using it helps me stay closer to more of my friends in a way that’s impossible to scale with in person interactions alone.
2. “How do you not get overwhelmed by all the people you follow/friend?” I don’t try to read everything.

Don’t drink from the hose. Dip your feet in the stream instead.

It seems that Twinkle is taking advantage of their popularity in a completely irresponsible manner.

Before I go into details, let me quickly recap how Twitter’s conversations work. When you publicly reply to a person on Twitter, you type @USERNAME. So if my username was kevin, you’d type “@kevin that’s so true!”. Twitter and pretty much all Twitter applications support this syntax by providing a “replies” view which shows you every public Twitter that starts with @YOURUSERNAME. So you can see that it’s fairly important for these usernames to remain unique.

Enter Twinkle, who turn out to not just be a Twitter application, but a social network of their own. A person who uses Twinkle doesn’t have to be a Twitter user as well. They might just be Twinkle users, with Twinkle usernames. Thus, there are now two sets of namespaces with duplicate identities that might belong to two different people (a person named Peter on Twitter and a different person named Peter on Twinkle).

Normally, this wouldn’t be a problem. I have usernames all over the place. However, Twinkle decided that the way a Twinkle user should reply to other Twinkle users is also with the @ syntax and if the poster happens to use Twitter as well, that reply goes to Twitter.

What does this mean?

Let’s say there is a Twitter and Twinkle user and his username is Peter on both services. He has a friend who is on Twinkle only and her username is Jane. Suppose Jane says something witty on Twinkle and Peter decides to respond:

@jane LOL. That’s pretty brilliant, mate.

As a user of both services, his response goes out on both channels — Twitter and Twinkle.

At the same time, JaneB, a Twitter user who has never heard of Twinkle and  the owner of the Twitter username @jane, logs onto Twitter. She checks her replies tab and notices Peter’s message. Except she has no clue who Peter is, nor what he thinks that her last Twitter message about her aunt being ill is all that brilliant or funny.

Essentially, Twinkle’s poor product design, or if you give them less credit, their irresponsible product design, is hijacking Twitter usernames. I’m not the only one who’s noticed this, either. Their Get Satisfaction page includes a number of threads like this:

The poster is right. It is completely unacceptable. In an environment where we’re trying to find solutions like OpenID to consolidate our identities, Twinkle has managed to find a way to create an application that now muddies the definition of who owns a username. The question is, what can really be done? My suggestion is that Twinkle doesn’t ever cross post responses to Twinkle users to Twitter as well.

But the real question is, what’s to stop anyone from coming onto Twitter and creating a similar kind of clusterf**k, polluting or spamming every @name there is?

Update: It looks like in addition to using the @ syntax, they also allow spaces and special characters (e.g., commas, dashes, etc.) as usernames. So now if your username is @kevin on Twitter, you may be getting responses in your replies tab for people who say “@kevin spacey” or “@kevin.cheng”. If you wish to voice a complaint to Twinkle and Tapulous, you may want to add your thoughts on this thread.

As it turns out, iilwy is a fascinating site in terms of its experience design. There’s a lot of interesting aspects of the site I could discuss but I’ll focus on three:

- creating a barrier such that it takes effort to have social contact
- the emergent behaviour of an open system
- using a points system on a social site

##It Takes Effort
The premise of iilwy is that dating sites are full of crap to wade through – especially if you’re a hot girl. I don’t have any personal experience of sites like Match.com but my understanding is that for a monthly fee, you earn the “right” to message x number of people a month. So as month’s end approaches, guys will feel that to get their money’s worth, they should contact as many people as they can.

The end result is that popular people on Match end up with the equivalent of spam or as founder Dan Albritton says, like every guy at a bar hitting on a girl without shame.

Iilwy is based on “games” each person creates. Players use the in game point system to “bid” on games. The top 5 bidders at the end of the game are eligible to be picked as the winner by the game owner and the winner and game owner are subsequently contacts who can contact each other through the site messaging system.

This filter puts the power back in the hands of each individual, no matter how popular. They can pick and choose who they actually want to be able to have contact them effectively creating _effort_ to have somebody in your contact list. Not surprisingly, that jives extremely well with Dan’s “hittng on a girl at a bar” analogy.

As my coworker [Sarah Cooper][2] at Yahoo! mentioned recently, social network “friends bins” are terrible analogies of the real world. You don’t make a friend by clicking a single button. It takes effort, communication and most of all _interaction_ to get to know someone.

Iilwy’s game system is actually a great system for meeting and filtering new people well beyond just the dating realm. In this case, a barrier that makes doing something harder is _desired_ and very deliberate in its design.

##Emergent Game Behaviour
The move away from just being another dating site is actually fairly pronounced on iilwy. The subtitle of the site is: “find. flirt. bid.” [Dan Albritton][3] says that they don’t encourage the games people create to focus on resulting in a date but that’s not reflected in the site. When you start a new game, there are a number of suggested games such as:

> “I want to learn the merengue”

> “Let’s go see some standup”

> “Looking for a running partner”

All of these suggest the winner of the game will be picked for some sort of activity. Luckily, iilwy abides by a common design principle of late: leave it open, see how the users use it. [Twitter][4] is a great example of this. Twitter has but one box with a prompt, “what are you doing?” Many people new to Twitter take this literally and only update to answer that question – which can get dull pretty quickly.

But because there was no defined syntax, nor rules, Twitter became a soapbox for just about anything. People could express their opinions, tell people where the party is, ask for help or even use it to announce engagements.

In much the same manner, iilwy has evolved and the definition of “game” on the site has become more and more broad. The games are much more creative now, ranging from basic “what’s your favourite movie?” to intricate “[haiku challenge][5]” games to even more demanding “[video dares][6]“. Some, instead of offering a promise of a date, offer something in return such as a [drawing][7] or a [greeting card][8].

As an experiment, I created a game that crossed the boundary between points and actual US currency.

##High Score
There are a lot of inherent dangers of setting up any community with a points or scoring system that’s explicit but also a lot of inherent benefits. Obviously, one can incent their users to participate in certain activities by dangling points on a stick and equally, users can be dissuaded from other activities when the cost is prohibitive.

I’ve been playing MMORPGs such as [World of Warcraft][9] since ’98, including numerous betas. Not surprisingly, there are a lot of parallels in the design of a persistent world’s economy and in game point systems like iilwy.

One problem that many of them face is the crossover into real world wealth. Games attempt to design such that the economy is self contained and is not affected by how much “real” money a player has. In WoW, this has become a rampant issue as many sites (mostly operating in China) are offering “farming” services where you pay them to make you in game gold.

Currency, even in a game, requires time and effort. Time is worth money. So [my experimental game][10] was to see what kind of exchange rate I could set for the points in iilwy. Instead of paying for the points directly, however, I put the money [towards][11] [charity][12]. The result? One of the highest scoring games I’d seen on the system at 2060.

Another issue commonly faced by these games is the issue of inflation. In the real world, economics is defined as the science of unlimited wants and limited resources to fulfill those wants. Games inherently have unlimited resources, however and so artificial “money sinks” are necessary to compensate. In WoW, new “features” were introduced specifically to deal with such issues – armour and weapons went through decay and would have to be repaired and higher level spells required reagants to cast. Each of these sinks forced players to spend money in such a way that they were essentially removed from the world – no player reaped the benefits of the money spent.

This issue is one I think iilwy is going to find themselves facing. There are very few money sinks within the site at the moment: inviting a player to create a game costs 200 point, selecting a winner only yields 90% of the points bid, and adding someone to your watch list costs 10 points. But logging into the system daily yields 50 points each time and even with the cut, bidding on other players is still transferring 90% of your bids to another player, keeping the wealth inside the game.

Already, there’s evidence of inflation. Initially, inviting a user to create a game cost 10 points but this amount was too little and presumably, people were still getting spammed with invitations. As time progresses, the overall pool of points within the site will continue to increase. Unless iilwy sets up new ways to spend points that take the points out of the economy entirely, they’ll find themselves with an inflation problem and will need to increase the price of inviting users to create games over and over – making it prohibitive for new players to even consider doing so.

Overall, iilwy is incredibly well designed in its interaction. It’s not rich in features but chooses to do each little thing in a very distinct and deliberate manner. I shouldn’t be surprised that one of the better designed point systems was implemented in what’s essentially a dating site but I am. Hopefully, other sites that decide to incorporate game mechanics are taking notes.

[1]:http://iminlikewithyou.com
[2]:http://sarahcpr.com
[3]:http://www.calacanis.com/2007/04/14/calacaniscast-24-beta/
[4]:http://twitter.com
[5]:http://iminlikewithyou.com/game/detail/4021
[6]:http://iminlikewithyou.com/game/detail/4495
[7]:http://iminlikewithyou.com/game/detail/4034
[8]:http://iminlikewithyou.com/game/detail/3987
[9]:http://worldofwarcraft.com
[10]:http://iminlikewithyou.com/game/detail/4530
[11]:http://flickr.com/photos/kurioso/502595850/
[12]:http://flickr.com/photos/kurioso/502595854/