This week, the topic of social networks is all over the place; indeed it even stars on The Economist (once and twice). Just like when a shoeshine boy starts talking stock tips, that means the topic has gone definitely mainstream. Only this time it’s not a completely bad thing.
Still, the Economist has fallen on this year’s truism: they repeat the oft-heard arguments about people wanting to either connect with people who are not registered in their walled garden of choice, or wanting to move their whole profile and friend list to another service. They are both false, IMHO. And the solutions touted (independently managed personal profiles, open ID management) are tangential to the issue.
Social networking services are essentially (let’s coin a definition) permission-based mutual awareness and communication services. They allow us to build “white-lists” to filter out people, and keep only those we trust and appreciate enough to share delicate (or personal) information with, or to allow to intrude upon our time.
That is a social network, redux. The rest is incidental to the focus of the network: it can be social, creative, or professional in varying degrees, and will thus offer filter-associated services to further those ends.
In other words, a social network service is an utility, not a content provider. It’s closer to an operating system than to a web portal. The differences are serious, as we will see.
What the journalists talk about
The shoe… sorry, the article authors criticize the (apparently constraining) need to visit each network’s home page to make use of their services. This is not a universal truth: witness Linkedin’s plugins for several of their services. It is simply a consequence of some of the network’s revenue generation models… and of the technology they use, to a lesser extent. There is nothing in the essence of a network that requires the visit – even if it’s often the most convenient way to access it – and the evolution towards distributed access is already marching on.
Something else The Economist comments as news is mostly old-hat too. Most existing online services are already drawing heavily on the primary networking tool, the email: contact lists are routinely exploited to find people you want to add to your network, if not to grade the type of information that is shown to them.
That’s where the Economist hits a real bugbear of social network users: they often wish they could (discreetly so as not to give offence) rate the type of “friend” that someone is, in order to allow them more (or less) restricted access to their information and their inbox. This is currently in its early infancy, but the need grows larger as the main networks lose their original specialised character and attempt to mix contacts from several spheres (social, business, family, casual). As far as I’m aware, nobody’s yet offering a non-offensive, effective grading system.
The false bugbears
Now let’s see the truisms, because they’re rather important.
First, it’s been stated that users would like networks to be transparent (be a member of one, freely access members of others). Even if each network’s services were the same, I don’t think it’s true. You use a network because you like the way it handles the filter, and because the people you want to allow in are mostly using that service. You don’t go about putting profiles just anywhere (unless you’re a compulsive networker or a total newbie), and making your profile acessable from anywhere would be just the same. Imagine your sister’s friends in Facebook asking for contact to your LinkedIn data. Or viceversa.
The second (jump network with your full friend list) is scarier still. Imagine the shock of finding yourself in your friend’s new network service, accesible to another (and possibly totally different) sphere of his contacts, subject to different access filtering rules. No way.
Garden walls are there for a reason
What the users (we) want is a way to handle profiles in a coherent way. Interconnection, but definitely not transparency. Allow your preferred network to “encapsulate” and handle your data (or maybe just part of it, to make it more complicated) but also allow it to speak to other systems so you can add a contact whose data is not housed in the “home” network; your network should only be allowed to access and present the data that your friend has allowed for that kind of connection (i.e. the brief resumé, or the Flickr feed, or the blog feed, or whatever). And viceversa: once you have connected two profiles (put them in each others’ white-lists) it would be nice to be able to message from one to the other.
But that does not require transparency. That requires standards (for cross-identification and access to data)… and a very coherent and strong way to filter and grade access. The standards are easy to build on today’s technology (Mugshot already allows members to see an aggregation of your friends’ activities on other networks, Facebook communicates with any external email account, Google is pushing common building blocks). The filters are not here yet, as already mentioned. With that in place, things like Flickr feeds or LinkedIn miniresumés (already available to the masses) could get their versions for “friends in other networks”.
OpenID, or any unified login system that allows a person to have a single set of data for identifying themselves in several services, is a reasonably good idea (whatever the security risks of a single point of entry for so much data). It would help to build those Mugshot-type aggregations.
All in all, the walls are more important than the gates: the value of the network is in the quality of the filter. If the services it allows are relevant for your goals, your people will happily join. But if you can’t restrict access to your data, you can’t publish it safely, and the network loses its point.
As far as I can see, it looks like we’ll either see services that are able to coherently aggregate other, specialised services (the way Mugshot is very slowly going), or generalist services that allow much more precise and efficient filtering of information and access according to “friend profiles”, thus becoming completely different environments depending on the use they’re given.
The first way (integrating outside data) is urgently demanded, if imperfectly supported. The second is just as urgently demanded, as the large network services swell beyond usefulness into spam-generation machines.
And of course, there is the actual services offered on top of those white-lists of contacts and contacts of contacts. There needs to be a lot of innovation in that field if the social network companies hope to become really serious businesses in their own sake. Because they mostly aren’t, yet… and some bubbles will eventually burst. That’s one point The Economist got right.