Last weekend, I became involved in a project called Tweak the Tweet. I wanted to put into writing my perspective on the project, its long-term viability, and the role I see it playing in the crisis-response tool chain.

In the broadest sense, I’m discussing the notion of tools built on the Twitter platform specifically for use in emergency response and crisis management. More specifically, I’m discussing the various technology efforts currently organized under the Tweak-the-Tweet banner. To be absolutely clear, I speak only for myself, not for the direction of the project.

What’s wrong with pure-SMS crisis systems?

In an emergency, broadcast SMS is the easiest way to tell “whomever is listening” about urgent needs and conditions. Pure-SMS systems such as Ushahidi’s 4636 short code in Haiti and Instedd’s Geochat program provide a valuable outlet for similar crisis information. How should Twitter-based emergency messaging fit into this ecosystem, and what merits would it have beyond emergency-specific SMS systems?

My perspective is that Twitter emergency messaging systems should act as a channel for non-urgent emergency messages, in order to keep the signal-to-noise ratio of crisis-specific short code systems high. With this division of utility, non-urgent statements, requests, and offers are routed through the Twitter system and the short code takes on the text equivalent of 911.

Specific advantages of Twitter as a platform

Because Twitter is a peer-broadcast medium, users of the system can find and contact each other outside of the purview of the emergency system, which will facilitate rebuilding, community self-organization, and organic infrastructure networks.

Because of the open nature of Twitter’s data and API, any number of agencies or organizations could consume the data most appropriate to their missions, allowing for easier interoperability. Our current implementation targets both push (outbound API integrations posting data to other systems) and pull (XML-based open API for requests) publication mechanisms.

Why structured syntax?

Our specific efforts seek to extract structured tweets in a particular syntax designed and promoted by the EPIC lab in Colorado, although it is a high priority for the technology to allow drastic changes to the syntax without any changes to the code. I believe that usage and adoption will temper idealism and prescriptive syntaxes in the case of a crisis with high Twitter usage. Listening technologies such as Tweak-the-Tweet must adapt rapidly. A requirement I personally take quite seriously is the need for all crisis-specific information to be updatable and deployable by a non-technical administrator.

Promoting a structured format at all makes what otherwise would be a deluge of information into a stream of data that can be auto-triaged and routed to appropriate organizations.

The adoption hurdle is an inevitable risk for prescriptive technologies like the Tweak-the-Tweet syntax. The strongest argument in that regard is that it is difficult to promote a Twitter syntax in a time of crisis unless there is technology listening and routing syntax usage. Therefore, the hope is that if we build it, they will come.

Urgency and Utility in Haiti

Currently, very few people are using Twitter on the ground in Haiti and certainly none are using the prescriptive syntax. However, the public visibility of the Haitian Quake and the public awareness of the need for a modern crisis toolchain provide a valuable opportunity to build momentum, gather requirements, and raise community criticism for a project of indubitable long-term value.