The archive contains older posts which may no longer reflect my current views.

Going beyond the hashtag - using implicit social graphs within Twitter.

hashtagAs I posted last week, the recent release of mobile application Color has caused quite a stir, not because of the app itself but because of the idea it brings to the table: the implicit social graph.

Implicit graphs are being touted as the way forward for the social web as they are more firmly rooted in the here and now, more closely mirroring real life as we are constantly moving through implicit graphs throughout our day.

Implicit social groups are formed around locations, events, ideas, topics and ad-hoc conversations exist for the duration of the event around which the group is formed in direct contrast to the way Twitter normally operates.

Twitter users have already demonstrated ways the service can be used: from elections to sports events to natural disasters it has taken its place at the forefront of instant communication and connection but this has not been structured.

Beyond the hashtag

By going beyond the hashtag and introducing implicit social graphs within Twitter more targeted conversations can be created. We can go beyond just location and group tweets together based on our interests and activities.

One man’s meat is another man’s poison and multiple people at the same event can create noise within timelines; how many times have we seen threats to unfollow people because they are attending a conference such as SXSW or CES – these would be ideal candidates for targeted conversations.


The obvious benefit of implicit social groups is keeping track of all related tweets within one place and there should also be an easier way to contribute. By hiding the conversation from the main public timeline we would also reap the benefit of reducing noise.

Revenue potential

Some conversations such as those surrounding sports events and conferences occur repeatedly so the owners of those events may wish to have an element of control over the conversations that occur; this would allow for the possibility of registering your event for a fee to gain this control.

So, how would it be done?

As the post title suggests implicit social graphs could be created by going beyond the hashtag; using some form of code. To keep the nomenclature consistent I would suggest the hashcode. The hashcode would be the target word surrounded by hashes, for example: #SXSW#.

Twitter with my6sense

The My6Sense extension for Google Chrome illustrates that additional tabs can be created for the #newtwitter web UI – this same behaviour could be used for implicit graphs created using hashcodes.

Tweeting just a hashcode on its own could open a dedicated tab for that conversation which will display all Tweets tagged accordingly.

Twitter with hashcode

The group tab will be a self-contained timeline and any Tweet posted here will not be visible in the main public stream, thus reducing the noise for those not attending the event or discussing the topic.

Tweets will be tagged as belonging to that group if they are either: posted directly from within the tab or from the public timeline provided they are preceded by the hashcode. Tweets could also, perhaps, be tagged within the UI itself to indicate the ad-hoc conversation to which they belong, as below:

Tagged tweet

Just as with @ replies, the tweets would show within the timeline on your profile should it be viewed directly so that a permanent record can be kept.

Just as with other tabs a drop down could provide the option to close the tab once you no longer need it.


Many conversations will be one-off events and could be created by anyone but, as described above, events such as conferences are repeat events and the opportunity exists to charge for the registration of the hashcode matching that event.

Registration could provide the ‘owner’ with an element of control such as locking the hashcode (so that it can’t be used unofficially) and treating it more like an account. A hashcode could have an avatar and description and the ability to block users from accessing it – perhaps subject to an official review from Twitter staff.


In short, hashcodes would provide the ability to create ad-hoc implicit social groups within the traditional explicit framework of social networks granting a focus and flexibility not currently available as well as a means to reduce noise within the public timeline.

I had intended this to be pitched directly to twitter but have not been able to find a way to do so. You can view the full idea pitch in this PDF: Going Beyond the Hashtag

Lead image by Jeff