Account creation is, arguably, the most important point of interaction between an app and user. Screw it up and people leave, never to return. And, yet, it’s typically the most abrasive point of contact you’ll have with a service. (Second, perhaps, to canceling an account.)

Why?

Service providers tend to confuse personal identification as a prerequisite to offering a personalized service. This is the kind of confusion that leads to questions like “what’s your name?”, “where do you live?”, “how old are you?”, etc. when all they really need to know is “do you need a key to get you back in again?”

The world got too excited about social networks. Now every time I want to try out a service, it wants to know my dating status. (I’m married dammit, leave me alone.) Stupid questions like these lead to distrust and bailout.

Here’s a question all application developers could stand to ask: Are we providing an identity service? If the answer is “yes,” then good for you! Let me know if you need some cheerleading. If not, go easy. Not only do you not need my identity, I’d argue that whole segments of apps don’t want it.

Take a look at Health 2.0 and, more specifically, [portable] medical records. The privacy risks involved are very real, very obvious. Far more so when you take a set of hyper-sensitive data, store it centrally, and then make it personally identifiable.

What if it wasn’t?

Maybe your news aggregator doesn’t need to make my attention preferences personally identifiable either. Just a thought.

Let’s ignore all the reasons you don’t need another social network and say that you do. And since we’re hypothesizing, I should tell you you are a content provider.

It shouldn’t be too much of a stretch. Every [new media] publisher I know already thinks this way. Social networks are a great way to build community, so every content provider can benefit from all the features of a social network. Right?

By default, I don’t buy it. But there are cases where site visitors would benefit from the content recommendations of friends. That’s one example.

To get the content recommendations of friends, you need friends. And to get friends, you need a social network. Right?

Err…

The success of a social network (bolt-on or otherwise) depends on members getting past the set up. Visitors need to register, elaborate, personalize, search for friends, request authorization from friends, authorize friends, and repeat.

It’s a lot of work to go through to be able to recommend content on one site.

Here’s a better way to handle this

Let the social network travel with the member rather than require them to garden yet another walled social network.

OpenID handles identity, kind of. Today, it’s more like sign-up, but it could handle a hell of a lot more. Like who I consider a friend, which communities I’m part of, and what content I find interesting for starters.

In other words, the member would own the community activities versus the reverse (the community owning the member activities).

It’s not so crazy to imagine. Each time OpenID jumps you past the registration for a site, it could store the community as a record inside your [decentralized] account. OpenID could also store records of your friends. And records of content chunks (form of a content id) you like.

From that data, you’d be able to share content with friends at any community / site you frequent. Providing, of course, that your friend(s) frequent the community / site as well.

OpenID won’t be implementing this anytime soon. But it would be very straightforward to build a companion app that lets you organize your communities, friends, and the content you find interesting (more later on why this isn’t a privacy concern).

A few adjustments at each site (authentication / markup) and the content provider would be able to provide visitors with the content recommendations of their friends. Again, one example.

Portable social networks are good. In whatever form they take.

Social networks suck. I’m tempted to stop right here, but I realize that’s just a side effect of my Twittering. Since we have more than 140 characters to discuss, I’ll keep it going and try to explain what gets me down.

It isn’t the friending issue. As much as I hate re-upping each friend / colleague I have inside each new network, support groups are already being formed by geeks with more friends than me, and solutions will find the market (real people) before the market finds the issue (social network fatigue).

It’s not even redundant profiles. The idea of creating / gardening more than one is also lame, but there’s a lot going on with decentralized presence. Services like OpenID and OAuth are already batting at the cause, and it’s only a matter of time before the browsers jump in as well. (I’m downloading Flock 1.0 right now.)

Then there’s your caged activity profile (i.e., who, when, what you interact with). This also sucks, but I can live with the idea that, beyond my friend lists or profiles, most of my data activity is meaningless outside of the context in which it’s caged. Did I get bit by a zombie or a werewolf? Maybe it was a chump. Whatever. I’m only interested in keeping the connection, Facebook can have the manner.

So what else is locked down? Publishing streams, right? I can get past that too. I don’t originate anything inside a closed network that doesn’t need that context to make sense. The crap I post in MySpace should stay there. I don’t want it on my blog. The streams I deem ready for public consumption are available for re-syndication via RSS, etc. My answer to not being able to export my Facebook statuses, for instance, is to not post them there to begin with. I republish my public Twitter timeline instead.

I’m pretty much over all of that (which is to say, I can point to and even implement emerging solutions for most of these issues).

What I can’t get past is the idea that these damn things (social networks) put a log in wall between communication channels. So any point of contact I want to make (whether message, friend request or zombie bite) means the network I’m currently logged into is just going to bleep/blorp an email to my counterpart instructing them to log in if they’d like to see what just happened.

WTF is that? It’s certainly not interaction. At least not in the sense of anything that’s come since email chess, circa 1993.

*sigh*

We bang up against requests to build social networks every day at ep. Most of the projects we’re engaged in, though, are either niche enough to keep my data caging complaints at bay, or they aren’t actually social networks at all (they’re collaboration apps). But we still battle the idea of the log in wall.

Too many people think you measure community in terms of sign-ons. And the way to build sign-ons is to blorp a bunch of teasers to their users.

But what if there was a service like OpenID that actually did act as a single sign-on model? What if people were perpetually signed in to their networks? And, what if messages and alerts and notifications across all networks were delivered in decentralized form, like that of an IM window?

Crazy right? It shouldn’t be.