Separating the News from the Noise


In all the talk about the future of news, it’s rare that you actually come across an innovation that alters the way you think about reporting and news gathering. Digital journalism innovations too often focus on how stories are presented to audiences. Social media is used to generate and sustain interest in news, rather than as a source of news. Where news is sourced from social media it is dominated by celebrity feuds or political gaffs. This type of journalism only scratches the surface of what is possible. Sentiment analysis, the virality of content, all measured through social media, overshadow the real evolution occurring, often imperceptibly, across the globe.

Consider the recent Guardian profile of Eliot Higgins. Higgins, a blogger, who from his livingroom in Leicester, “has put together a database of 491 videos of cluster bombs being used across Syria, together with map references and details of the type of weapons used”. Scouring “about 450 YouTube channels from Syria every evening” Higgins has managed to beat organisations such as the New York Times to a globally significant story nearly 5000 km from his front door.

Which is what drew us to the work being done by Storyful. Founded in 2010 by former RTE Prime Time presenter Mark Little, the organisation is a ‘news agency for the social media age’. Storyful’s raison d’etre is to draw actionable news from the noise of social media. Its journalists and editors curate stories about conflicts, natural disasters and elections by verifying social media events – tweets, facebook updates, videos, images etc.

This interview with editor Malachy Browne explores the work they do, the evolution of the organisation and some of the challenges they continue to face, in terms of hesitancy from the traditional news world. Browne raises important issues with respect to attribution and the ethics of engagement on social media, which is very much still unchartered territory for many journalists.

The thirst for 24 hour news has seen some of the worst journalism practices in recent months. The Newtown school killings showed that, beyond the inaccurate reporting, there were serious questions to be asked about how journalists should engage with witnesses via social media. Also, Browne’s account of Rami Al Sayid, who’s work reporting from Homs went unattributed, echoes traditional news failures to cite the work of their local contacts and reporters in conflicts.

With over 100,000 hours of video uploaded to YouTube every day the need for and potential effectiveness of the type of journalism described below is growing by the minute. Looking back to what a similar headline to the one used above (Separating the News from the Noise) meant in 2001, it’s difficult to imagine where innovations such as this will bring us 10 years from now.

(MB: Malachy Browne, Storyful; DM: David Manning, MediaBite)

DM: How did you get involved in Storyful?

MB: At the time that Mark Little left RTE I had just set up an Irish website called with Vincent Browne. We had an idea of doing something similar in an Irish context – to bring social media into reporting and news gathering, using a mixture of curation and original reporting, while also making use of emerging technologies. So when Mark began canvasing for like-minded people, I contacted him and we found we had a shared vision.

DM: From your blog I understand Storyful are involved in training, how does that side of the business work?

MB: There’s a couple of different sides to it. Firstly, there’s ‘public thought leadership’, whereby we publish stories about what we do, giving practical examples of the theory behind social media journalism. We recently published an eBook comprising the best blog posts and case studies of good social media journalism, to illustrate what people should be doing.

We also do consultancy work with newsrooms and NGOs. We go into the newsroom and show them the tools that we use and discuss best practice for discovery and verification of social media content. Fundamentally, we are talking about information systems, how to manage news gathering.

DM: Is there a lot of sharing between those news organistations which have developed their own social media verification teams?

MB: What you find is that social media practitioners collaborate off their own back, sharing information, whether that be on Twitter or Facebook etc. Groups like the Vulture Club on Facebook include NGOs, stringers and conflict reporters, all sharing information, tips and tricks, everything down to  equipment, the best place to stay and sourcing translators and fixers, as well as breaking news based on what they know on the ground. But within newsrooms you don’t really get the trade secrets, you don’t get that much collaboration as far as I have experienced.

DM: So, for example, the BBC and CNN’s social media teams, would you see them as competition?

MB: The BBC is a client of ours, we’ve collaborated with them on a number of stories. When the recent protests broke in Tehran over the currency crisis – the currency was devaluing at an alarming rate, promoting currency traders in the Grand Bazaar to protest. We had video from there, with the location absolutely verified, but we were finding it difficult to verify the date. The BBC were finding it difficult to find the location, but they had the date.


BBC Persia were able to decipher the date using the slogans that were being chanted and from referencing the banners hung on the walls. So by collaborating we were able to verify both the date and the location, which are the two main things you want to cover yourself legally for broadcast. That would be an example of private collaboration, happening behind the scenes.

There was another instance in Syria, where a number of organisations collaborated publicly in order to verify a contentious video purporting to show a man being buried alive by Syrian forces. The video turned out to be a fake. All the information relating to that we published on Storyfulother organisations chose what they wanted to publish themselves.

DM: To what extent was that verification chasing the story? Had it spread around the world by that stage?

MB: We picked it up as it was emerging in the middle of the night. We had serious reservations about it immediately and we advised our news clients, first of all of the existence of the video and that it was going viral, and that we had justified reservations about it’s veracity. By mid morning we noticed a number of news organisations running with it as legit, at that point we decided to put a public post out. We collaborated with our friends at the BBC, another British reporter and a Syrian person who did an audio analysis of the video. We put it out through our network of activists and they had reservations about it as well.

[Warning: Graphic footage]

DM: I want to ask about Storyful’s evolution, did you envisage it being what it is now when it started? It appeared to have a stronger community aspect, whereas now it appears focused on business to business.

MB: Storyful was always founded on the principle of drawing actionable news from social media, that fundamental concept hasn’t changed. We were always going to be supported through business to business, that was our business model. We did have a public website and also a community forum in the initial stages. However, the community forum didn’t really work. It took a lot of TLC and energy to build and court the community. We also faced problems with people using it to push PR material. So it wasn’t really worth our while, we felt our time was much better spent honing our discovery skills around social media, rather than trying to attract users to the Storyful community.

We were never concerned with advertising, we were never traffic driven, it was solely about story telling in a new way. Essentially the public website was a marketing tool and we got to a stage where we didn’t need to do that, because our business to business side grew so much we gradually had to scale back the public story telling. While we continue to do that storytelling now, it happens behind a type of paywall for our clients, including all the raw ingredients they need to build and enhance their stories.

DM: With the focus on business to business, to what extent do you find yourselves leading your clients rather than responding to them? Is it a case you are pushing them in the direction of stories or are they coming to you with the stories looking for the verification?

MB: 90% of the time we are pushing them towards the stories. We set the agenda for what we go after, we make a judgement call like any other newsroom in terms of what makes the cut using our resources, which includes the discovery, the verification and the delivery out to our partners and their news clients.

One of the things news clients like about us is that they have 24 hour access to our team. If they have a query about a particular video or story, our system facilitates a two way conversation. This morning we were reporting on Storyful Pro, using our Twitter newswire, on the suicide bombing in Ankara. At that stage we didn’t know it was a suicide bombing, but ABC in the United States quickly got in touch with us about an image that was circulating. It was one we were already working on in the background, but weren’t comfortable enough to alert our clients about yet. They wanted to know if it was legitimate. 10 minutes later we were able to say that it was.

DM: How much of that 10 minutes is new media verification versus old school methods?

MB: We don’t have anyone on the ground in Ankara. We might be using people who are on the ground, who are posting videos or images. In this case, IHA, a Turkish media organisation, had sent a reporter to the scene. He was running towards the embassy and filming as he went. From the video we could see a doorway which tallied with the image which was circulating and the satellite imagery we had pulled of the area, meaning we could confirm that it was actually taken there.

DM: Are there any instances you’ve been disappointed with the types of stories that are picked up, where your editorial focus doesn’t par with those of the larger news organisations?

MB: This is a tricky one, I think it all has to do with context. For instance, a bomb was dropped on a place called Kafranbel. It was a horrific attack, completely destroying an entire block. There happened to be an activist that we knew in Kafranbel who was active around that time. We went straight to his YouTube account and found he had uploaded high definition cinematic quality video of immediate aftermath. It was shocking. There was complete silence bar the wails of women in shock, people running through the streets filled with charred bodies, smoke in the air, cars and trucks ablaze. Debris was still falling, it was that fresh. The person filming happened to be just outside the radius of the fallout went it struck.

It was dramatic footage, which we pushed out to our clients, but nobody ran with it. We couldn’t find it on websites or broadcasts that evening. It was hard to believe. The following day we found a very low quality video of wisps of smoke above a highway in Damascus. That video was used in a whole range of reports. News organisations didn’t particularly care about Kafranbel, it wasn’t considered strategic in the Syria battle. Whereas the highway to the airport outside of Damascus was.

DM: Is it difficult to be in a position of collecting pieces of evidence, and find news organistans attempting to hammer that in to a particular editorial narrative?

MB: We don’t worry too much about the context in which news organisations want to use footage. We provide story fragments, not the opinion and analysis. Purely the facts that we have established around a piece of content or event. We are an agency, the same as Reuters, the same as AP, except we do social media. It is up to news organisations how they want to use that.

Having said that, we do keep an eye on traffic relating to what’s used and what’s not used. Obviously there’s an element of serving news organisations, but that doesn’t inform an agenda on our part. We have a mix of content, with a range of dashboards, whether they want viral videos, weather videos, business news, world news videos or news specific to Syria. There is a variety of content there for people to choose from.

DM: While it is very early days in the social media news business, do you think this type of remote investigative/verification journalism is being recognised as a means to keep doing the type of worldwide journalism that is being eroded by industry economic problems, reduced staff numbers, budgets etc?

MB: It depends on the organisation. Certainly we are seeing more and more news organisations changing their attitudes to social media. You can change all the technology you want in a newsroom, but if you don’t change the attitudes you are not going to generate that change.

Digital journalism strategies vary from organisation to organisation. For instance Channel 4, at the beginning of the Syrian conflict would have used user generated or activist content, now they rely much more on video journalists who go in, choose a really tight focus and through that microcosm tell a story about the broader conflict. That’s their editorial choice. We find other outlets like the New York Times will integrate content  on their website when they are telling a story about a particular attack.

I don’t think there’s ever going to be a replacement for the reporter on the ground, but what social media does is help enhance reporting on the ground. Mark Little was a foreign correspondent with RTE for quite some time. Social media partly bypasses the many procedures he would have had to undertake to file a report, meaning you are able to get closer to the story and provide additional layers of context.

There is public tool called iWitness, which we’ve used in the past week, particularly around Port Said. You enter a location and it presents on a map all the tweets that are coming from that area. What’s very useful about this is that it gives you the people that you know are on the ground and, importantly, it gives you the specific terms they are using to describe an event.

So if you consider an uploader in Benghazi, because they are interacting with a social circle that knows they are in Benghazi, they might use the name of a square or a building, or slang for a place. By using these hyper local search terms you can then broaden your search across other mediums for content, because, you have to remember, only 5-10% of tweets are geo-tagged.

DM: Do you come up against, for example, Syrian people living abroad tweeting, not necessarily maliciously, as if they are in Syria? .

MB: On our country lists, the first rule is that the person has to be in the country. It’s no use for us following a Syria expert, like Joshua Landis, because he is not there, he is not going to be an eye-witness. Even within the list, if we’ve got somebody reporting from Aleppo, from Damascus, from Homs, they are obviously not in all those places so we would never take them to be a witness to something. Unless we could confirm their location, which you can do by asking them to for example take a twitpic or by looking at their digital footprint, checking their other social profiles, where they locate themselves and that sort of thing. A lot of it is about vetting the source.

DM: What sort of response do you get? Do the people you are trying to engage with recognise what you are doing? Is it always a positive response or can you be seen as doing something negative? For instance, in the Syrian conflict, if you were to contact a pro-Government uploader, would they treat you in the same way in the verification process?

MB: We haven’t had too much engagement with the pro-Government side, mainly because they have their own mediums to get information out. Although there are a couple. We do ask questions, and the same as anybody else, if you are vetting somebody, questioning somebody the natural response is get your back up. And sometimes people do get very defensive and you just have to deal with that, explain that you’re just trying to establish some facts. So it just takes some careful engagement. The other issue we have is that some people don’t want to be identified, understandably, and we respect that. But usually if they are a good source, we can get on Skype and they can recognise who we are and what we do and they come to trust us. We’ve had that experience with several activists who we’ll be in touch with via Skype, and we’ll know them by first name, even though they go by an acronym. We would never make that public.

DM: Through this, in much the same way reporters on the ground, engaging with people living or working in war zones, you develop relationships with people who potentially may die.

MB: We do and we have had that experience. Last February, there was a person, who’s profile I checked up on every morning, who was killed the night before Marie Colvin was killed. His name was Rami Al Sayid, he was a video activist in Bab Amr in Homs, which came under sustained bombardment for about 2 months from the Syrian military.

They completely flattened the area. He was out on the streets very close to the action, filming every day and he eventually was killed. His news reports were broadcast across the world that evening and the weeks leading up to that evening. He really drew the attention of the world to that suburb which probably led to Marie Colvin, Remi Ochlik and all the other reporters sneaking in to that suburb, because this was at a time when journalists were banned from Syria.

In her reports, the evening before, Marie Colvin used his videos. She was killed 12 hours later. And of course, her death, tragic as it was, completely overshadowed his death, because nobody knew who he was. Of course news organisations don’t really give credit to YouTube uploaders, particularly in Syria, they generally broadcast without attribution.

DM: With that in mind, and the general issue of being exposed to violent footage, one of your team published a blog post about dealing with trauma, how does Storyful address that?

MB: We have a councillor available to staff all the time and we had a session about a year and half ago, just to educate our team on trauma. So we have those services available. There are also certain ethics that you apply to your journalism, we never want to put an uploader or would be uploader in danger, that crosses a line. There are certain rules by which we abide when engaging with people who may be at a scene of a dangerous incident.

DM: From an outside point of view it is quite easy to see how your process works for instantaneous events, is it possible to apply those techniques or to adapt those process to cover slow burning events?

MB: That is more difficult, certainly you can build lists of people who engage primarily around those topics and we’ve done that around markets, business and entertainment for certain clients and there are certain analytics you can run off that, but again its a very real time system.

What we have done is feature documentaries, with for example Channel 4, and several other organisations, on a freelance basis. They pick a particular theme, they are using User Generated Content to tell that story. For instance, one of the stories we did with Channel 4 was a series, Torture Machine, about using detentions and torture in a systemic way, which Syria has been doing for decades, but has ramped up in the last couple of years. That involves going through in the region of 70 videos of torture, from instances of actual torture filmed by the guards who are perpetrating it and those phones or cameras were later lost or captured or leaked out somehow, or in the aftermath when bodies are returned.

So that wasn’t in a real time situation, it involved a lot of digging back through social platforms to try and find witnesses, to try and find information about the victims, where they were from etc. That’s labour intensive, but rewarding when you finally join the dots and bring all the facts  together.


Another recent example of that in Tahrir Square on the 29th January last year, about 9 months before that an American reporter Lara Logan was assaulted. This person who was assaulted in January looked very like her and we had a query from a client to prove that it wasn’t Lara Logan. Twitter’s search is limited, it goes back only a month or two, but using Topsy we were able to go back through 9 months to tweets at that time and try to identify the people. We actually found three eye witnesses reporting from the scene and we were able to get in touch with them and they were able to confirm the identify the woman. So if you know the right tools, doing historic analysis is possible.

DM: Presumably, all countries aren’t as active in terms of social media. Are there many dark spots where news is happening and there is just no noise?

MB: I’m sure there are and we’ve had difficulty in some countries, like Central African Republic, some Eastern European countries, China can be difficult, through censorship. We have staff in Hong Kong, with native Mandarin, Cantonese speakers. There certainly are black spots, and even within countries that have a strong social media presence, such as Egypt, there is a very high rate of illiteracy, so how representative is social media in Egypt, that is something we are aware of as well.

DM: There’s going to be an entry level to social media through mobile devices, in terms of financial resource, is there a potential to skew the way you see an event due to that?

MB: There potentially is, but if you consider Egypt and Tahrir square and the tens of thousands of people that are there. There is only a handful that are among the crowd that are broadcasting. Images are really very hard to doctor, if you have lots of them coming from a single place, then they tell the story. But there are I guess rural areas are black spots definitely, simply because of the limits of communication technology. Resources in some countries, certainly developing countries.

DM: Presumably that’s the reason why you rarely see any content coming out of rural Afghanistan, save for video from US military, official and unofficial. There’s simply no infrastructure there. Would you see a lot of noise from Kabul itself?

MB: Our Afghanistan list is actually one of our strongest lists and it is because of the people we have in Kabul and a couple of other locations. It is usually very accurate. They are very au fait with the mediums they use to broadcast. Content from Kabul is usually not a problem, because there are a couple of local networks there. Pajhwok Afghan News and TOLOnews are very good at getting on the scene quickly, but of course that’s agency stuff. So in that case we would pass on to our clients their contact information and they would contact directly with respect to the rights of content. In other areas though it can be limited. But there again, in traditional news, you are not going to have reporters stationed in remote areas of Afghanistan on the hope that something is going to happen, so in effect the net is cast wider with the growth of social media.
DM: But you don’t think it will ever replace the need for a roving reporter?
MB: Certainly not, take Mali for instance, pretty much the only content that is coming out of there is from the French forces themselves, primarily when they want to publicise videos of aerial attacks on rebel bases. There is no infrastructure, there are very few people uploading user generated content, raising the question of ICT and the availability of cheap technology. At the same time Western reporters are being held back from the front line until the French and the Malian forces have done their job. We are starting to hear some stories coming out of there, which we have yet to verify, but if they are true, it raises serious questions about what they have done and are doing.


Leave a Reply

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

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

Twitter picture

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

Facebook photo

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

Connecting to %s