Storyful is, according to Founder and CEO Mark Little, “the first news agency for the social media age:" it sources news content from the "real-time web," authenticates it, and delivers it to an influential roster of clients, including The New York Times and the Economist.
Part tech start-up, part news agency, it is powered by a small team of professional journalists scattered around the globe – from Dublin to San Francisco and Hong Kong – who work as social media “field producers,” using a combination of algorithms and human skill to pick up early warning signals of breaking news, pinpoint sources on the ground, decide which tweets, photos and videos are “actionable,” and feed this verified social content to their publishing partners.
As news organisations plunging into online video strive to find their legs on YouTube, Storyful’s ability to discover and validate citizen footage grows increasingly relevant. This is reflected both in the Nieman Lab’s summer 2012 report titled Truth in the Age of Social Media – to which Little contributed the checklist that Storyful employs to authenticate citizen content – and in a recent study by the Pew Research Center on YouTube’s growing significance within the news industry – to which Little responded by exploring, in a blog post, What makes YouTube’s news audience click.
In a recent conversation with Editors Weblog, this visual news vanguard, who began his career as a TV journalist, imparted wisdom on the challenges of working with citizen sources, from attribution and remuneration to authentication and protection, on the shift from control to collaboration in the digital realm, and on the future of online video.
Editors Weblog: In your article for the Nieman Lab report you wrote, “authority has been replaced by authenticity as the currency of social journalism.” Please elaborate on this concept.
Mark Little: When I started off in journalism, we would judge our sources by the level of seniority. The closer you got to the centre of power, the better the source. Now, what we’re finding on the social web is that there are all sorts of people that, because of their proximity to events, are so much better at being authentic sources.
Years ago, the first thing I would want to do when I hit the ground in a foreign country was to interview a minister of the government, interview a local journalist, and interview an academic to tell me about the political situation there. So I was seeking people who have authority. Now, if I hit the ground virtually, in terms of analysing the conversation in the country, I want to find the person who is influencing public opinion most – who is actually shaping the conversation because they’re close to an event, or they have a particularly compelling point of view.
We’re analysing the U.S. presidential campaign at the moment, and you always want to find the local mayor in Iowa who is the gatekeeper – who every politician has to pass through to talk to the people in the diner in Indianola. You’re always looking for the person who is the most authentic source on the ground, closest to the event being described, and who has most influence on the conversations taking place around that subject. Usually they are not the politicians; usually they are not the academics; and they are certainly in many cases not the journalists who tend to drink in the same bars and circulate in the same places, and join each other in supporting whatever conventional wisdom is out there.
Authenticity – being close to an event, being the person who is shaping the conversation around an event – is so much more important these days than the label that you attach to your name, or the position you hold within an organisation. The social web has demonstrated this truth, and I happen to think it’s liberating, because it means that people won’t all sit in Washington D.C. by the Beltway and analyse the mood of America – we’ll actually get out there and talk to the people who are shaping the conversation.
EW: Could you please describe the model you’ve built for gathering news from authentic sources in a region where journalists can’t penetrate, such as Syria?
Little: We have essentially built data sets. We start with Twitter: we start a list on Twitter with all of the people that we feel have proven themselves to be authentic sources. We then have a piece of technology that allows us to examine the networks that they are connected with. So we can have 50 names of people we feel through our own judgements are people we should listen to, and then our algorithms kick in and tell us who the most authoritative sources within that network are. So we will tend to build our lists off a combination of human judgement and artificial intelligence.
The next step for us then is we will extract the content that’s trending among those people, so for example we can extract the content that’s coming out of their Twitter conversation. We have, obviously, a set of Facebook accounts that we will monitor on a regular basis, and we have RSS feeds feeding us YouTube accounts that are regularly updated with content.
Now, as the Syrian conflict has developed, there is a much more sophisticated network of activist groups who we’re in contact with, who essentially have become much better in helping us to establish the veracity of certain video. We started to see a couple of months back that they would hold up signs showing the date; point their cameras at key landmarks to help us identify it; and they would also have a whole range of people who are outside of Syria, for example in Kuwait, Cairo, Beirut, who will help us to translate and distinguish between certain accents.
It began in Tunisia at the beginning of 2011 when we were watching a hashtag, and it’s got to the point now where, through a combination of engagement with key sources, human judgement – that’s our judgement as journalists – and artificial intelligence, where we can essentially have a very sophisticated discovery and validation network.
I think we’re starting to see how the next generation of journalists will be able to deliver and build the most authentic contact book – so 20 years ago, I used to literally have a contact book, where I had all my contacts written down – the ones I trusted. Now we can do that through a combination of technology and journalism, and also engagement with the most connected people on the ground. That’s why I believe as a long-time journalist that this is a golden age – I’m better connected now, through Twitter and YouTube than I was as a foreign correspondent actually in Iraq and Afghanistan.
EW: What are some of the hallmarks of manufactured or manipulated content, or content that is trying to push a message in a way that is dishonest or disguised?
Little: Well we think that first of all there are the opportunists – the hoaxers – who will, for example in the case of a flood or an earthquake, pass off old footage and say that it’s current. In the past we’ve seen many news organisations falling for that kind of hoax. I don’t know why they do it – I think it’s a community of people who love to see the media make a fool of themselves, and I think they are very active, particularly around major news events or natural disasters.
I think there’s a second group of people who are politically motivated to try and push a particular line. We’ve seen examples of that during the Arab Spring: YouTube accounts set up by allegedly ‘ordinary’ people, which are pushing a pro-regime account. So if they’re supporting the current regime in Syria they’re saying, “here’s a video that describes atrocities committed by the rebels.” And what we find out in some cases is that it can be an elaborate hoax. We had a video last year in which there was allegedly a person being buried alive by Syrian soldiers. After some investigation we changed to the conclusion that it was most probably a hoax. It took us several hours – nearly a day – to establish that. We’re seeing increasingly sophisticated attempts by political organisations, sometimes state or pro-state organisations, to try and fool or discredit the established media.
Finally, there are physically altered videos where smoke may be thickened, or where certain things will be added. In those cases, I think technology is catching up. So I would describe it as a bit of an arms race, where our ability to detect hoaxes is developing quite rapidly, but then again, the people who are committing the hoaxes are also developing their technology quite rapidly. Every time we find a way to detect something, the hoaxers seem to find ways to try and promote it.
Part of the problem here is that people are not held to account for things they might do within YouTube in the same way they would be within the established news media, because there is not the same level of scrutiny, there is not the same level of established standards, and there is not even the same level of legal accountability. In the absence of all that, we have got to work much harder to establish truth – veracity – and I suppose that’s a challenge that we’re taking on in Storyful. I think YouTube themselves are also realizing that they’re much more important to the global stream of information and debate than perhaps they might have thought two or three years ago. Certainly the Pew report is proof of that.
EW: The Pew report speaks to the fact that a clear set of ethical standards has not yet developed for attributing citizen video content. Why do you think this is, and how can it be resolved?
Little: I think the problem is that there hasn’t been up until now a journalistic force in the YouTube ecosystem. A lot of the partners on YouTube are essentially trying to promote their own brands – understandably. Some of them do it very well, and I think some of them make a huge contribution to citizen journalism – like Reuters, The New York Times, the BBC – but I think at the moment what we’re missing is a common standard on the use of user content within their own products.
There is such a competitive urge right now for everyone who is in digital publishing to try and use YouTube as a source, and you’re seeing what may be called ‘grey areas’ – they’re not exactly unethical, but they’re certainly not best practice. That is, for example, where certain established broadcasters will simply take a YouTube video and put their own logos and brands on it. Some of them are quite happy to include a credit to the uploader, but I think the problem is that many digital publishers and broadcasters just don’t like the idea of having to put a credit up on their broadcast.
There are so many different standards for different broadcasters and digital publishers – I think it might be time to try and agree on a format by which the original user is credited.
I don’t think it’s a huge problem; I think these are things we can address, and I certainly think that YouTube has been incredibly inventive in trying to find ways to help the news industry not only use YouTube as a source, but also contribute their own content into the YouTube ecosystem in a way that benefits everybody. Partnerships with brands like Reuters, AP, and the BBC are really exciting, and I think you’ll see more of that in coming months.
EW: How do you attribute sources at Storyful?
Little: We have an agreed form of words that we put to every uploader that we talk to. So for example we might find a video of a tornado in Kentucky. We find the video – we connect with the uploader, and we ask them if it’s okay for news clients to use that video in return for a credit to the uploader. And then what will happen is we’ll take that video and connect with ABC News in let’s say New York, and that morning if they ran that video on Good Morning America they would include a credit to the original uploader. So what we’ve done there is we’ve acted as a mediator between the original user who posted the content and the broadcaster, and we’ve established the basis – the permissions – by which that video can be used.
In most cases, probably about 95 percent of all the people that we talk to, the uploaders are quite happy just to see their name in lights – to see ABC or The New York Times or any one of the people we work with use their video, as long as they properly attribute the uploader. So that’s the standard that we start at.
Other people want to get paid, which we respect. In that case we will always tell the broadcaster or the publisher that we work with that they will need to go to the uploader and sort out money, or sort out rights.
In certain cases we know that video has been put up on YouTube by campaigns, by humanitarian groups, by activists, by law enforcement, who want it to be seen by as many people as possible. In that case we feel it’s publically acceptable for as many people to use it, as long as they attribute it.
We are starting to see three categories emerging:
- Establishing clear attribution – to recognize the uploader’s ownership rights but also allow the broadcaster to use that video.
- The emergence of a rapidly growing group of people who are professionals, who would like to see payments for their video.
- Huge growth in what I would call public video within YouTube, and that is people who are trying to push their message out through the YouTube ecosystem.
EW: In the case that a source does want to be paid, how involved do you get?
Little: We have some people who are helping us on a regular basis, and we will offer some financial return for things like translation – for services they may provide. We don’t buy video and we don’t pay any money to activists. We don’t try in any way to pay for help they may give outside of basic professional services.
What we do try to do is, if a journalist in an established news organisation wants to use their video, or wants to engage them professionally, we provide Skype details or Twitter handles to make sure that they can make that transaction. We don’t get involved; we’re not a brokerage or news agency in the traditional sense.
We don’t own content; we don’t sell content – we essentially try to make that content useful. We want to help the 5 to 10 percent of our network [who require payment] build a model where they can sustain the quality journalism that they’re doing, but we don’t want to be in a position where we’re profiting from citizen journalism.
EW: On the opposite side of the spectrum from attribution, what are some of the measures you take to protect the identity of sources?
Little: We are very much a business-to-business service. That is, we will reveal as much as we know about a video to our broadcasters and partners, and allow them to broadcast or use that video according to their own editorial standards. Some are quite happy to push out a video as quickly as they get it; others have very rigorous standards by which they use it, including masking the identity of people who might be at risk.
At the moment I think there are very few ways of properly protecting identity. That’s why YouTube’s initiative last week to help activists blur the identity of people in video is hugely helpful for us. We don’t have the resources to carry out that level of protection.
We can obviously hide the source in terms of not broadcasting the name of an individual to our partners, who can then come to us and say, “How do you know?” and we provide them with as much data as we possibly can. We’re there to provide all the meta-data to make that content useful. Certainly if we’re asked to restrict the details for the security of the uploader, we will do that, and we certainly did that during the early stages of the Syrain uprising, where we had sources in Damascus and other towns, who were in personal contact and who asked us to do what we could to make sure that their identities were not released or were not made public.
EW: How do you see online news video content developing in the short, medium and long term?
Little: I think in the short term, most news organisations that try to get into video are going to find the financial returns disappointing. There are going to have to be a couple of years of learning what the right formats are. I think personally that the more authentic the video that you’re showing – the rawer the experience – the better. The Pew study showed that the most popular news videos viewed within YouTube didn’t feature a personality, like a presenter, or an individual politician; they were videos about an actual event. I think that’s the emphasis I would be putting in the future – finding the most authentic videos of events that people are compelled to watch.
I do think that we’ll probably in the longer term see increasing partnerships between YouTube and publishers, where there may be ways that YouTube can help their partners make a bigger impact, deliver perhaps even higher advertising revenues from the YouTube ecosystem, but that’s probably a medium-term goal; I don’t think it’s going to happen overnight.
And in the long term, I really feel, and I’m probably stepping out on a limb here, that YouTube will become the destination. It will become the place where people go in the first instance to consume news in the world, particularly video news. They won’t go necessarily to a broadcaster’s channel on YouTube; they will seek out a much broader experience.
The winners, I think, as we go forward long term, are going to be the newsrooms and the journalists who recognize the value of curation – of essentially bringing together all of this vast amount of video – we’re talking 72 hours of video every single minute. Journalists are needed more than ever within the YouTube ecosystem to help curate and make that useful, to come up with the ethical standards, to come up with the solutions to the oversupply of information. Long term, the people who can do that and who do that in partnership in YouTube are the people who are going to see a thriving business.
There are a lot of issues to be decided – it’s going to be a difficult transition – but the key understanding here is that journalists can no longer have ownership over scarce content. Now they have to be the managers of an overabundance of content that at the moment is confusing to most users. The people who make that less confusing and more useful are going to be the people who win; the people who think they can own content going forward in a proprietary are, I think, the people who are going to be the slowest to adapt, and possibly the people who don’t adapt at all.
This interview has been edited.
In case all of this text about video made you yearn for the real deal, here is a clip of Mark Little explaining his vision for Storyful at the Institute of International and European Affairs in Dublin, a few weeks before he formally left his last job:
YouTube & News: A New Kind of Visual News, Pew Research Centre
What makes YouTube’s news audience click, Storyful blog
Blurring faces makes security a clear priority, Storyful blog
Finding the wisdom in the crowd, Nieman Reports