Volume 27 Issue 1, 2012

The twitterisation of ABC's emergency and disaster communication

Julie Posetti

University of Canberra social media researcher, Julie Posetti (who's working on a PhD titled 'The Twitterisation of Journalism'), asked the ABC's national social media co-ordinator, Ping Lo, to detail the ABC's experiences in an online interview for AJEM.


A report released in January, on the first anniversary of the Queensland Floods, highlighted the important and rapidly emerging role of social media in emergency communications. The report, #qldfloods & @QPSMedia: Crisis Communication on Twitter in the 2011 South East Queensland Floods (Bruns, Burgess, Crawford & Shaw), found that the Queensland Police Media Twitter feed was the most visible presence in the flood of information emanating from Twitter. In second place was the ABC’s Twitter account @ABCNews.

Twitter bird

The ABC serves an important function in crisis communications as Australia’s official ‘Emergency Broadcaster’. So, how is the ABC approaching social media during natural disasters like #qldfloods (the Twitter hashtag by which the Queensland flood disaster became known)? What are the benefits, risks and pitfalls? And what can other organisations learn from the ABC about deploying social media?

University of Canberra social media researcher, Julie Posetti (who’s working on a PhD titled ‘The Twitterisation of Journalism’), asked the ABC’s national social media co-ordinator, Ping Lo, to detail the ABC’s experiences in an online interview for AJEM.

Q1: What's the ABC's approach to using social media in emergency and disaster situations?

The use of these media during the Black Saturday bushfires of '08 – where a staff member used twitter to great effect – helped pave the way to making social media activity a more routine work practice. It became evident to the broader organisation that employing well-organised social media could have a profound impact in an emergency situation. Having set a precedent during a disaster which saw media coverage very closely scrutinised afterwards, we felt it was the ABC's responsibility to continue to use social media during emergencies thereafter. It's a continual work in progress – as is all coverage of emergencies – as broadcasters, emergency and response services, government agencies and the public all strive to better inform.

Q2: It's a year since the Queensland floods and the subsequent cyclones. Can you take us through what the ABC did in terms of deploying social media communications during that large scale emergency?

Our emergency strategy is built on established daily routines. Rather than funnelling our social media activity through one or two key ABC accounts on different platforms, we have established many ABC accounts on these platforms. These accounts represent programs (such as 'triple j Hack'), stations (such as ABC Brisbane), networks (such as Radio National), genres (such as ABC Environment), and divisions (such as ABC News).

This is a resource intensive approach but enables passionate program makers to tailor their activities to suit those whose ABC preferences are as individual as they are.

During an emergency, having so many accounts offers opportunities and challenges. While it's helpful to be able to scale our social media coverage accordingly, it is obviously challenging to coordinate the efforts of these accounts, with the staff managing them spread across the country and based in completely different areas of the ABC.

For the floods, the divisional social media coordinators and I discussed likely coverage strategies and talked with some management and staff about resourcing relevant accounts, editorial responsibilities and managing the tricky juggling of social media on top of increased station/newsroom tasks. A new emergency-focused account was established on twitter (@abcemergency), and radio and news staff concentrated on covering the floods on multiple social media sites throughout the event itself and the ensuing recovery period. At that stage, divisional social media coordinator positions had only just been filled, and there'd been little opportunity for those pivotal roles both to connect with divisional staff and management, but also to collaborate across divisions. With every emergency, ABC staff become more confident in how they can utilise social media, but the biggest change has been the development of very effective channels of strategic cross-divisional communication – in the lead up to, during, and in the aftermath of an emergency – mirroring the communication which already occurs in the ABC's radio, TV and general online coverage of that emergency.

Q3: Did the ABC see social media as a separate set of communication platforms to be deployed and staffed during #QldFloods, or just as adjuncts to its broadcast functions?

During the floods it was more the former than the latter, bearing in mind that with so many different accounts in operation, some realised the critical importance of proper planning and resourcing of social media activities sooner than others (— and in the spirit of learning from each emergency, by the time Cyclone Yasi was bearing down on the Queensland coast, all of those relevant account 'owners' recognised the importance of giving each account the thought and attention it required).

Very often social media and traditional broadcast media activities are discussed as separate spheres of activity, whereas in reality, for the ABC the approach is (rightly, I believe) much more blended than that. Social media platforms external to the ABC site are clearly a separate set of communication platforms but they serve the ABC and those connected with us in ways that are changing all the time, depending on what's going on. ABC staff have been quick to recognise that it's problematic and limited to approach social media platforms as 'support acts' for the main coverage on broadcast media. That said, in emergency situations we have core responsibilities that have long been built up on the key, robust platforms of radio, online and television. We now have those same responsibilities with social media, and it is up to us to embrace them as a standalone resource, not simply as an adjunct.

Q4: How did the ABC try to integrate social media with its traditional outputs during the crisis?

This varied from account to account, as each has its own character (and expectations and responsibilities) within any particular social media platform. But generally speaking, on social media platforms we published key messages being delivered on air, summarised key on-air announcements, linked to key online articles and linked to other emergency services agencies and relevant organisations in the same spirit in which we connect with them on air. On our traditional outputs we fed crowdsourced information into on-air messages (with clear explanation, treating them in the same way we do public talkback), referenced the social media activities of other ABC and emergency services accounts, and used search tools to listen to what people were saying so that we could get a better feeling for what was going on in different places.

Some ABC accounts, such as @abcnews, already had a high level of integration with traditional outputs (the @abcnews account is managed by the senior producers on the News Online desk) so when the floods occurred it was more a case of providing more resourcing to enhance what these accounts were already doing.

Q5: Social media is a very rapidly developing field and often it demands experimentation. How adventurous was the ABC in deploying social media during the floods? What new approaches did you try?

Beyond finessing our integrated social/traditional media approach, we undertook one significant experiment. The timing of the floods coincided with our trial of a map-based crisis-crowdsourcing platform Ushahidi. The trial centred on collecting feral animal sightings across Australia, but was conducted with emergency and other story coverage possibilities in mind. This platform, developed in Kenya as a result of growing unease over fraught election processes, is all about harnessing the power of numbers – in this case, numbers of people willing or compelled to contribute location-based information about a crisis through a number of means such as SMS, tweet, online form and email. We had been discussing the potential of utilising this platform for some time, having seen it used effectively in a number of weather-related emergencies further afield: the floods in Pakistan and the Haiti earthquake being two prominent examples. When an emergency happens here, the phone lines of stations are in meltdown with members of the public wanting either to contribute or to request information or assistance. With that in mind, it struck us that harnessing a platform such as Ushahidi could give any member of the public who felt so inclined the opportunity to offer information via this alternative means, while at the same time giving us the opportunity to collect and potentially draw attention to this information in a new, useful way. This was a highly adventurous but thoroughly considered experiment. It was agreed that of all emergency types, the slower-moving, more drawn-out nature of floods offered us the best opportunity to test the platform with as few risks as possible, people being unlikely, in such an area, to rely solely on the information to make critical decisions.

There is a compelling tension the ABC has to tackle if it is to assume some responsibility in this rapidly growing field of emergency coverage: that of timeliness balanced against verification. On the one hand, most of all it is verified information that everyone is aiming for – for this can mean the difference between life and death. But then again, so can timeliness: getting enough useful information out there, quickly, for people to refer to in order to make decisions. The two can't and won't always coincide. On top of this, the sheer ease and facility of social media means that excellent, succinct pieces of information can and do become lost in a sea of unintended or even malevolent information — with mass duplication of it all.

Ushahidi has a huge potential to make sense of this 'noise', but it requires of its participants that they accept a certain level of uncertainty in exchange for the ready availability of information that is 'most likely' to be correct, based on a whole suite of logical algorithms. It's not unlike taking a talkback caller whose information cannot immediately be verified by emergency services but is deemed urgent and potentially useful enough to be put to air with appropriate caveats. Nevertheless, it has the potential to assume a huge scale, given the escalating use of social media during emergencies by the public, so we didn't take it lightly. We contributed a large proportion of the information ourselves which we were able to mark as 'verified' by cross-referencing with emergency service and ABC reports, and then undertook a strict assessment of public-contributed data which meant that any uncertain pieces of information, however likely, were nevertheless still tagged as 'unverified'.

To allow ourselves the opportunity to learn about the platform and how it could work in an Australian emergency context when managed by us, a broadcaster, we kept the promotion of the Queensland Floods Ushahidi map low-key and largely confined to online (i.e. to an audience already engaging predominantly with online) instead of or in addition to traditional media.

Q6: What approaches worked best in terms of information dissemination and community response?

It might seem obvious, but clear, simple questions are the most effective means to elicit response from the community. At any time, loading a status update with a heap of information and ending with a question means that all of that information is effectively devalued. This is especially the case during emergencies, when people don't have the time to sift through a lot of text to find the crux.

Timeliness is key, as is consistency. It is only reasonable for the public to know what they might expect from a social media account, both as regards update type and frequency. Too inconsistent, too irrelevant, — and people will disconnect from you. Engaging in conversation with people to elicit more information and connecting with 'power users' in the community on the ground in the midst of the emergency is also important. As is pointing to and intersecting with emergency services, response and recovery services, government agencies and other relevant ABC accounts.

Pointing to non-ABC and non-official sites such as a crowdsourced photo gallery on Flickr created by members of the community, and tailoring strategy and tone to each platform, on a per account basis are significant developments, too.

All of these approaches highlight the need to remain flexible enough to adapt strategies ‘on-the-go’, in response to community needs, or a development in the emergency. Also vital, is recognising that there are two different audiences relying on ABC emergency coverage: those in communities immediately affected requiring accurate, clear information, support; and interested observers who might have family or friends in the region but are not personally at risk. The requirements for both vary but the latter group's interest and willingness to be involved or provide assistance needs to be acknowledged and managed concurrently.

Q7: What approaches were less effective? And why were they less effective, do you think?

One problem was too many updates, each not holding enough value in and of themselves to warrant contributing to the noise. Another was unnecessary repetition among accounts, especially those whose followers overlap, which contributes only to the noise and runs the risk of accounts losing followers.

Issuing untimely information, such as pointing to an overly generalised emergency information article (however interesting) at a critical point is also problematic — and a potentially dangerous distraction. Inconsistency in the frequency, tone or approach of posts is also an issue. At critical times, there should be no doubt as to what an account is going to cover and attempting a cover-all approach to all social media platforms, wherein one message is published across a number of different platforms can be a problem too. This is likely to result in fewer responses and less reach across the board.

Q8: One area of experimentation you focused on was what's known as 'crowd-mapping', where you ask your community or audience to locate problem areas, such as rising floodwater, on an interactive map (see Ping Lo’s answer to Q5 above for more detail). How well did that approach work?

Bearing in mind we kept the promotion of this trial low-key, the map nevertheless garnered a lot of attention, mostly from people not directly involved in the crisis but presumably interested in what was unfolding and how it was being covered. As we were putting a number of features of this approach to the test, we allocated resources to input verified data into the map alongside inviting the community and other agency contributions.

Approximately 1,500 reports were submitted to the map, of which we estimate one-third were contributed by the community, with the rest either created or harvested from different sources by ABC moderators. While information was able to be pulled from a number of different sources (email, SMS, online form, twitter, blogs), the online form built into the map interface was the most popular means for submitting reports.

The site received 230,000 visits over a 24 day period, with the peak occurring following the flash flooding that happened in Toowoomba. Despite the visits being so high at one stage that they toppled Ushahidi's servers, the average time spent was very short, indicating that most visitors were looking at the map out of interest but not exploring it in much depth. This ties in with the fact that most traffic came from references made to the map by Google, the BBC and the ABC.

We hadn't expected many public contributions given that there'd been no opportunity to build awareness among the community or indeed relevant ABC staff about the possibilities of the map. However, we were surprised to find that upon hearing about the map some people felt compelled to contribute their experiences of the disaster, for example providing evacuation advice based on their own experiences.

Q9: Some organisations are now reporting that they rely on Twitter and Facebook, for example, as the fastest way to gather and disseminate emergency information. What's the ABC's experience? How did this play out during the floods? Was radio still king?

From the organisation's point of view, radio is still central to our efforts, and it's a matter of continually lifting our social media game rather than diminishing any focus on radio. Radio is an enduring, critical medium for good reasons: the network is resilient, fairly durable and far-reaching; the technology is simple and easy to access; the message is clear and, as far as the ABC is concerned, the audience has a reasonable idea of what to expect.

That said, the ABC recognises the continual increase in the use of online, and within that, social media, by the Australian public. In particular, during emergencies we are noticing an increasing number of ABC Online readers, ABC radio listeners and TV watchers choosing to use social media in the lead-up to, during, and after an emergency. There's an immediacy in them that lends them to emergencies. They’re quick and allow people to dive into the eye of a storm, so to speak, from wherever they might be in the world. It's no coincidence that platforms such as Twitter and Facebook see big uptakes in their service in the lead-up to and following a major crisis.

At the ABC we certainly recognise the value of social media in gathering and disseminating emergency information quickly and to a vast potential 'audience'. To that end we are continually refining strategies for emergency coverage on social media that includes (but is not limited to) live breaking events and press conferences, crowdsourcing and clarifying misinformation.

Q10: There are obviously risks involved in allowing publication of unfiltered, community-supplied information – particularly during emergencies which are traditionally the focus of risk averse communications – what does the ABC identify as the main risks and how does it evaluate them?

Key risks in using a tool such as Ushahidi have to do with the trustworthiness of the information being submitted to the map and the timeliness of that information given that conditions during emergencies change all the time. However, these risks need to be balanced with the community's changing approach to online activity during emergencies and their growing expectations about finding information online.

Ways to tackle these risks include raising awareness about the possibilities of such a tool and how they might best make use of it as contributors and/or audience members, involving official emergency sources in order to verify data, being clear about verified/unverified content when providing data, providing clear time stamps on submitted data and highlighting most recent updates and making the crowdsourcing process as transparent as possible.

It's important to note that challenges such as the trustworthiness and timeliness of data are foremost in the minds of those driving platforms such as Ushahidi. They are constantly working on ways to sift through the noise of social media and other online content to extract the most reliable and most timely information possible. During the floods we trialled a tool being developed by Ushahidi called Swift River, whose aim is to extract the best information possible by using a mixture of logical algorithms coupled with human assessment.

There’s more work to be done around who should spearhead such an approach, and who should be involved in order to ensure the best possible outcome. There are a number of community initiatives such as Bushfire Connect (which also uses Ushahidi), Google's Crisis Response, as well as emergency services developments which need to be taken into consideration when thinking about how best to serve the community online during an emergency. It's everyone's responsibility to minimise confusion at these critical times.

Q11: It can be a difficult balancing act trying to weigh up the potential risks of social media deployment against the potential (and often unanticipated) benefits – especially in a fast moving disaster. What kind of approach to this problem do you suggest?

Provide training to staff, so as many as possible feel comfortable with navigating social media. You also need to build awareness in the organisation about what colleagues in other areas are planning to do, and of how their activities are likely to figure in the broader picture of ABC coverage of an emergency, as seen from the point of view of a member of the public.

You should have a strategy in place, and service it responsibly. Consider the commitment you need to make to the plan, such as adequate resources. It’s important, too, that consideration is given to the ‘type’ of emergency and what communications will be available to those on the ground. The 2010 Brisbane floods, for example, were anticipated with a few days to mobilise resources and implement strategy. Other emergencies such as bushfires do not always offer a window for preparation and planning.

The ABC contingencies include ensuring any staff potentially responsible for social media activity during an emergency have undergone editorial training and are confident of their ability to navigate the likely tides of information flowing in all directions. In addition, ABC Radio's Multiplatform team has created a trial role, Executive Producer for Emergencies, to facilitate this process and ensure staffing and resource capabilities are being met. This position is currently active for a six month period.

Q12: Is it possible for organisations involved in emergency communication to avoid social media in 2012?

It's possible, but it's unadvisable. To avoid social media is to be missing an ever-increasing and vital portion of the puzzle.

Q13: Finally, what are your top three tips for organisations using social media in emergency communications?

  • Be strategic about your social media activity. Plan ahead. Allow for flexibility in approach.
  • Build awareness internally and externally about what your social media emergency plans are.
  • Consistency is vital. People need to know what they're getting, and when, where and how they're getting it.

You can follow Julie Posetti and Ping Lo on Twitter, they are @julieposetti & @pinglo

About the author

Julie Posetti is an award-winning, internationally published journalist and academic who lectures in broadcast journalism at the University of Canberra. She is currently writing her PhD thesis on The Twitterisation of Journalism. A former radio and TV journalist with the ABC at the national level, she has been a regional news editor and a Canberra political correspondent. She now writes about digital media transformations for the prestigious PBS Mediashift website and commentates widely about journalism. Additionally, She also consults on social media strategy and online community building, providing training and policy advice to major media and corporate clients including the Sydney Morning Herald, SBS, UNSW, the Commonwealth Ombudsman and CSIRO. You can follow her on Twitter twitter.com/julieposetti. She keeps an occasional blog at J-Scribe (http://www.j-scribe.com)