Towards an SMS Code of Conduct for Disaster Response

Picture this: it’s October 7, 2011, and a major hazard hits a highly vulnerable population resulting in a devastating disaster. The entire humanitarian response community mobilizes within 48 hours. Days later, the cell phone network is back up and dozens of SMS systems are activated by large and small organizations. Two or three of these systems use short codes thanks to rapid collaboration with the country’s national telecommunication companies. The other SMS systems all use long codes.

That picture concerns me, a lot. The technology community’s response to Haiti has demonstrated that using SMS to communicate with disaster affected communities can save lives, hundreds of lives. Humanitarian organizations and NGOs have all taken note and nothing will prevent them from setting up their own SMS systems in the near future. This wouldn’t worry me if coordination wasn’t already a major challenge in this space.

Let me elaborate on the above picture.

Picture further that one organization decides to send out regular SMS broadcasts to the disaster affected communities to improve their situational awareness and prevent panic. This is an important service during the first few days of a disaster. But imagine that this organization does not provide a way for users receiving this information to unsubscribe or to specify exactly what type of information they would like and for which locations. Next suppose that three NGOs set up long codes to do the same. Now imagine that two major organizations independently set up an alerts SMS system, asking individuals to text in their location and most urgent needs.

This is an information disaster in the making for communities in crisis.

So what are we going to do to prevent the above picture from turning into reality? The group Communicating with Disaster Affected Communities (CDAC) is probably best placed to support a coordinating role in this space. But before we even get to this, our own community should start drafting an “SMS Code of Conduct for Disaster Response” for ourselves. I can think of no better way to start the process by using distributed cognition (aka crowdsourcing). This blog post on lessons learned and best practices may be informative as well.

Here are a few ideas to begin with:

  • Set up a complaints mechanism
  • Do not duplicate existing national SMS systems
  • Set up a single clearing house for all outgoing SMS broadcasts
  • Ensure that SMS messages are demand driven in terms of content
  • Enable receivers of Disaster SMS’s to unsubscribe and to specify alert type and location

There are likely dozens more points we could add. So please feel free to do so in the comments section below. I will then create a more structured Google Doc out of your replies and send this out for further peer reviewing.

Patrick Philippe Meier

21 responses to “Towards an SMS Code of Conduct for Disaster Response

  1. Agree fully. Failing in Chile seems to be lack of short code. Most reports initially coming by web from Santiago.

    As cell service is restore in the south SMS may well be primary means of reporting/communication.

    International community must understand what you have illustrated above if efficiencies are to be realized.

  2. Patrick – a few things come to mind:

    # clearly publicize – over and over – the function of your service, and do not shift your mission without reassessing the entire ecosystem and notifying the public

    # do not launch an SMS service unless you have the ability to act on incoming information — e.g. someone asks for something and you can respond with the information, service, or service referral that they need

    # aim for consistency in language, response times and quantities, and such, from day one

    # at launch, identify a time period in which your service will be useful and your resources available

    # at launch, even with loose consortia, identify the decision-makers and establish protocols for making those decisions

    # plan regular and professional contact with telcos to avoid swamping them with ad hoc requests

    # consider that telco incentives will not necessarily align with yours, and while you’re happy to volunteer your resources, they may not share your sentiment/ability/etc

  3. While I recommend that there are some shared guidelines; the starting ones suggested above are a bit quaint…
    I would stick to 4 main themes and understanding there are PATTERNS to go after rather than one-size-fits-all guidelines. (And this is from experience deploying nation-wide mission-critical SMS systems in many countries)

    – Design focused on value and simplicity to beneficiaries over work it may cause on the backend (hard to have objective tests; and was a fail in haiti)

    – Support of national institutions and processes (then again, folks like MSF might disagree)

    – Compliance with local SMS regulations (then again, some may disagree and see a crisis as a way of pushing these regulations forward)

    – Support ecosystem diversity (central clearinghouses are a great way to stifle rapid problem-solution iteration. And if central=remote, all the worse. But then again, aspiring central clearinghouses may disagree)

    I will not add all the other ones here …you may know we are preparing with the Haiti telcos a retrospective analysis to capture the learnings and look forward to incorporate experience-based feedback into these community guidelines.

  4. Just had a chat with Nigel and Imogen who brought up an an even bigger point:

    The IFRC Code of Conduct:

    Rule #1: Do No Harm. Given that the technology community played a central role in the humanitarian response to Haiti, we may want to incorporate some of the IFRC language into an SMS Code of Conduct.

  5. I think at some stage this should be a UN responsibility

    if you could imagine what a UN managed system would look like … might help inform the design of the present ad hoc systems and help in the the evolution of a permanent system

    and further imagine what the local system, telcoms, local and State govs would have to look like….. so they readily adapted to an outside system

    • @quinny, if you could imagine what a UN managed system would look like … might help inform you what NOT to do 🙂

      @patrick, one suggested guideline: share your data freely with other legitimate orgs working on the same disaster. Dividing the requests for help between 4 orgs at random reduces the effectiveness of all. If you do divide up the messages, do it from the aggregate data set with the agreement of all parties based on your speciality.

  6. I strongly support the idea of all groups in this space endorsing the code of conduct, and making sure that do no harm is at the heart of programmes and activities.

    For instance: there’s a whole bunch of issues around what it means to actively solicit input on needs without setting up ways to link to those who can help fulfil them, or who have the responsibility to do so at the very least.

    I think this is a great place to start a really important discussion.

  7. Good post Patrick. You certainly raise a number of good points here. Here are a few comments:

    1) You focus mainly on the humanitarian response community, but one could argue that in many cases this should be the role of the local government. We way to often dismiss them as an important player in the response activities, thinking that only us super-knowledgable and experienced responders from abroad can do this work. I would argue that you need to think of how to drive this kind of effort as a collaboration between the government and the humanitarian community.

    2) Why not aim for a universal disaster information shortcode, similar to how the 112 number has become a standard in Europe for calling for help (same as 911 in US). It would make things a lot easier if the same number would be agreed upon. Maybe one could work with the disaster people at ITU to help drive this kind of effort.

    3) I agree with the above comments that it is not enough to have a number, you also need to have the ability to process what is coming in. And you need to be able to provide people with some service level agreements. If you are trapped underneath rubble, calling 911 or sending a text to a relative/friend might be much quicker than sending a text to a service that maybe answers a few hours/days from now.

  8. Pingback: Crisis Response and SMS Systems Management for NGOs and Governments « iRevolution

  9. Pingback: The impact of crowd-sourcing project | Diary of a Crisis Mapper

  10. 1. Public Education:
    I think that it would be wise to tackle the public education side simultaneously. There’s not much point in working out all the technical issues if the affected public cannot make responsible use of the facilities. We must absolutely guard against only the “loudest voices being heard” because this could totally mis-direct response resources. We’ll need to come up with some simple and standard ways to express threats and needs by SMS (fire/ hazmat/ #injured / water/ food / shelter/vulnerable population/ entrapped) And, esp. the importance of including some standard spatial coordinates!

    Guidance could be put into a 5 minute online-lesson including for widest possible online and mobile distribution. It would make sense to partner with a large annual community-wide drill like California, Bogotá, etc. where the system could be tested and improved continuously, simultaneously getting the wider public involved. We should always think big (eg. earthquake/megacity).

    2. Local Government:
    I strongly second the HUGE importance of working with existing local governments. Most places will have more ‘there, there’ than Haiti. We will have to push them with civil initiatives, but we must take pains not to exclude and bypass them. ICLEI ( would be the best place to take this up for good leadership, standard-setting and broad buy-in.

    3: Developing and Promoting Code of Conduct with Humanitarian Networks:
    It would be valuable to get participation in developing (and then promoting) Code of Conduct via key ngo networks like InterAction (US) and Bond (UK), as well as with SPHERE Standards. All of these have important networks and experience in developing and promoting standards and codes of conduct.

  11. Good Points Identified.

    Keep Posting Patric.

  12. Patrick, this one concerns me.

    “Set up a single clearing house for all outgoing SMS broadcasts”

    I totally get the need for coordination, but it’s quite easy for systems like this to become blocking mechanisms, intentionally or not. Coordinate, but don’t squash the potential for innovation.

  13. Pingback: Data Protection Standards 2.0

  14. Pingback: Finally, A Decision-Support Platform for SMS Use in Disaster Response | iRevolution

  15. Pingback: Finally, A Decision-Support Platform for SMS Use in Disaster Response | iRevolution

  16. Pingback: Launching: SMS Code of Conduct for Disaster Response | iRevolution

  17. Pingback: Kenya: one election, 7 phone services, 3 maps and some confusion! | Diary of a Crisis Mapper

Leave a Reply to Marla Petal Cancel reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s