Paul Bradshaw and Turi Munthe discuss crowd-sourcing journalism

 by Martin Belam, 11 March 2011

Yesterday I posted the notes I took of Paul Lewis talking at a panel event about crowd-sourcing journalism. It was part of an afternoon entitled “Data and news sourcing” jointly organised by the Media Standards Trust and the BBC College of Journalism, and hosted by the Royal Statistical Society. Also appearing on the same panel were Paul Bradshaw and Turi Munthe. Here are my notes on their opening talks

Bradshaw was talking about his experience in running Help Me Investigate. He said it had exceeded his expectations, as he had thought he was just going to get “failure for free”, just as lots of potential stories die a death in front of the editor at news conference. The site had, he argued, unearthed significant stories, including the overspend of Birmingham City Council on their website.

He had identified five key factors of a successful crowd-sourced investigation:

  1. An alpha user who pushes the investigation forward - a pattern repeated across the web where the “90/9/1 rule” suggests one active user for every ninety lurkers. Without that one alpha user, an investigation dies.
  2. Momentum. However small, you need to keep the investigation rolling along. Uncovering lots of different pieces of information makes it look like there is continual progress.
  3. Modularisation. An investigation which can be broken down into small parts enables more people to participate, and helps to keep momentum up.
  4. Publicness. Twitter was instrumental in an investigation into the London Weekly because it drew people in. Paul felt they could have maybe used Facebook more.
  5. Expertise and diversity. A good investigation needs people playing the role of journalists, and may require FOI experts, or experts with a specialism like planning, photographers, and sometimes simply people who happen to be in the right location.

Like Paul Lewis before him, Paul Bradshaw sounded a note of caution and highlighted some problems with the site and crowd-sourcing journalism. These chiefly centred around legal issues, which mean they can't make the whole of Help Me Investigate public. Whilst this limits the legal risk, it also limits the possibility to get involved. Paul finished by saying that now that the code running the site has been open sourced, they are effectively crowd-sourcing the future of the site.

Turi Munthe, CEO of Demotix, followed Paul. He picked up on the theme of the alpha user, explaining that the crowd already functions according to a pyramid structure, with leaders and influencers and followers. Crowds are rarely just an indistinguishable mob.

He was concerned that the selective demographics of participation on the internet could skew the journalistic narrative. He cited the “revolution” in Iran in 2009, when a lot of the reports of unrest and dissatisfaction with the regime and election results were coming from a handful of sources, in a country where only 0.3% of the population had a Twitter account. When pushed by Chair George Brock on how the coverage of Iran in 2009 could have been improved, Turi said that, essentially, having people on the ground was the only way.

He thought there was a marked contrast with recent events in Egypt, where the demographic of people involved in Twitter was much wider. He also noted that communication with Demotix contributors had almost entirely shifted from email to public and direct messaging via Twitter. Having people who could speak Arabic and get involved with the Arabic-speaking Twittersphere had also been an advantage.

He echoed the concern of Paul Lewis, that without traditional journalism skills going hand-in-hand with new communication forms like Twitter, there was a risk that a vocal minority could drown out the true picture. Mind you, in a later part of the session, during Q&A, Paul Bradshaw said: “I think the media is an even stronger example of something dominated by vocal minorities than the internet”. Later that evening during his inaugural lecture at City University, he showed a slide representing the traditional media, and said that is was no coincidence that they were all white males.


I’ve still got some notes to blog from a second session at this event, which featured a discussion around the reporting of crime statistics, and from the Paul Bradshaw lecture that I attended later that day. I’m sure those will be appearing on the site in due course...

Keep up to date on my new blog