let’s get two things out of the way.
- i love twitter and am frustrated with its service of late
- i love friendfeed but don’t think it’s a twitter killer
if the two above statements mean nothing to you, that’s fine, it’s not important. i want to talk about micro-blogging – the short status like statements that individuals are making and recording via the internet and mobile devices – and the twitter/friendfeed debate has a tendency to get in the way. i think that’s a shame, particularly when so much is happening in the micro-blogging space. over the next 6 months to a year, i expect to see convergeance around micro-blogging inputs (“friends”) and outputs (presentation). in particular, richer metadata will become an implicit part of the interface, enabling users to refine an ever richer stream of information from sources both known and unkown. location information and friend classification will be the drivers of change over the near term and from this, i believe, one dominant service will emerge. from there, micro-blogging services will require further parsing and prioritization capabilities such that the right information arrives exactly when needed. that is a topic for another day. for now i want to focus on the immediate horizon.
my friends: who? what? where?
twitter profile information (location) and post information (what) are separate entities.
no discussion about microblogging can be had without discussing twitter. twitter put micro-blogging on the map, and simplicity is its calling card. users submit 140 character posts which the service then presents in reverse chronlogical order for the set of self selected people a user subscribes to, or “follows”.
each post contains a limited amount of metadata: the sender’s name, a timestamp, a location proxy, and an optional rating. beyond this, the community has added 2 features: replies and hashtags. replies are made with a simple syntax (an @ symbol followed by a specific user’s name to which the reply is directed.) hashtags, a character sequence beginning with a “#” included in the post, provide additional metadata for a post, usually for a topic, location, or event.
each twitter user has a profile with the service. this is independent from posts, so although the profile includes a user’s location, it is neither associated with a post nor immediately available to the user’s audience. location, then, is best guessed from the device which sent it. “mobile” posts imply that a user is out and about, but obviously, this is unrealiable as a location proxy. instead we must depend on a user explicitly stating their coordinates in a post to know where they are. a user’s profile also identifies their “favorite” posts. favorites are a form of rating that allows a user to favor certain posts with a check so they may be referred to later.
to twitter’s detriment, its well documented service problems have prevented it from expanding on its metadata. in the meantime, other services have emerged with different takes on the medium.
plurk specifies an optional controlled vocabulary for posts.
plurk is a new service, similar to twitter, that expands the twitter functionality with several notable innovations: 2 classifications for followers, cliques, a controlled vocabulary, threaded conversations, and a sophisticated user interface. let’s examine each more closely.
- friend classification
- plurk followers (in twitter lingo) are either fans or friends. a fan has subscribed to an individual’s posts. it is a one-way relationship. a friend designation denotes a mutual relationship. two individuals are fans of each other.
- cliques
- cliques are groups of friends. posts directed to a clique are seen exclusively by its members.
- controlled vocabularies
- a post in plurk may optionally employ 1 of 15 qualifiers: loves, likes, shares, gives, hates, wants, wishes, has, will, asks, was, feels, thinks, says, is. the qualifiers are optional, a user may ignore them and post “freestyle”. “share” is noteworthy in that it easily facilitates sharing multimedia such as photos within the service’s conversations, something that the base twitter service does not do. the qualifiers are an interesting spin on posting, and i would be love to aggregate them so that i could view what some group of people – known or unknown – “wants”, “likes”, etc. at a given time. such functionality is not currently offered.
- threaded conversations
- when a user replies to a specific post on plurk, the user interface reflects that, making conversations much easier to follow. that said, threading “buries” responses, such that refinding particular posts is difficult.
the service does not currently offer search, which might remedy this situation.it does have search, it’s new, and not an obvious part of the interface. - user interface
-
the plurk user interface expands on twitter’s simple page-based presentation in two ways. first, all posts are displayed on a timeline which scrolls from right to left. that may sound odd at first, but it allows the user to read from left to right and scroll further to view historic posts. the timeline context is maintained at all times. - second, when additional information – threaded conversations, account maintentance, photos, etc. – are presented, they are presented on a layer above the timeline. the user interface is much more playful than twitter’s, yet this is largely a matter of user preference. at this point it is conjecture as to whether the interface influences adoption of the service.
brightkite friends can be seen on a map interface, but posts can’t.
brightkite is a location based micro-blogging service. posts are manually associated with a given place by the service’s users. to address privacy concerns, friendships come in 2 levels. “trusted” friends are able to see much more granular details about where an individual is. posts are divided into 2 categories: those from my friends and those near me. posting on brightkite involves first “checking in” to a location and subsequent posts are made from there until the user checks in elsewhere. checking in requires manually searching for your present location. these searches can be done by address (all levels of granuality are supported: city, state, zip, country, etc) or business name.
the availablility of location information creates interesting opportunities for viewing. not only can one see their friend’s posts, they can see posts in their vicinity, and their friend’s locations on a map. brightkite has 2 big opportunities to improve it’s offering: location awareness and further use of maps. location awareness would greatly simplify the search process used when checking in. including more maps would visually identify popular places by showing where greater numbers of posts occur, either in realtime or historically. adding friends to this equation would enable identication of popular places amongst a user’s friends.
others
several other services, existing or recently announced, deserve mention. in july, the next edition of the iphone, with GPS (Global Positioning System) capabilities, will be released bringing location awareness to a much larger segment of the population. already several services have been announced to take advantage of this, including loopt and nrme. loopt superimposes your friend’s location coordinates on a map using gps coordinates. nrme’s focus is proximity. users will be made aware of all posts happening within a 9 block radius.
likely outcomes – part 1
i expect the best features of the above services to begin cross pollinating. in fact, brightkite was built to interact with twitter from the start. brightkite posts can automatically appear on twitter and can also update the location field of a user’s profile. brightkite and twitter are redundant, one form will emerge. each could learn from plurk’s interface, and i expect microblogging services to present alternative views: archive lists like twitter, map-based like brightkite, and some variant or enhancement of plurk’s timeline. the real opportunity though, exists around metadata.
friend grouping is the first piece of metadata i expect to see in all services. while broadcasting is an important part of microblogging – think newsflash – over time directing messages to groups of people will become essential to reduce noise and maintain relevancy. location information will become a part of all services, adding a second layer of filtering, and replacing some hashtags.
while i urge caution (for complexity’s sake) in the use of controlled vocabularies like hashtags or plurk qualifiers, the potential of such a mechanism will tempt services to include it in their base functionality. hashtags are voluntarily added by the twitter community, and i believe this optionality is the reason for their success and adoption.
all microblog posts should include an option for rating. it should be a simple one-click “like/dislike” system, like twitter’s. i am not convinced of the utility of a five star system. it is more difficult for the user, ie “is this post 4-stars? or 5? who cares, i’ll skip it.” in the case of rating, simpler is better.
this new metadata – friend groups, post location, and ratings – will provide for advanced searching capabilities, including real-time alerts for critical information. very soon, one service, or suite of services, will incorporate all these features, and that service will attract the largest community. in my next post, i will consider the next iteration of “friends”, particularly entities not people, and the important role trust plays in maintaining relationships on these services.