How to measure success in data journalism and other tips from experts at the Data Journalism Unconference 2017

Last month we put 45 data journalism experts from around the world in a room in central London and got them talking about today’s challenges in the world of data-driven storytelling. This article is a roundup of what we’ve learned at this one-day event hosted by the BBC.

Data journalists, editors, NGO workers, programmers and other experts gathered in London for a one-day unconference

Data journalism experts from around the world gathered on 23 May 2017 at the BBC’s headquarters in London, UK, to tackle data journalism challenges. Organised by the Global Editors Network in partnership with Google and Chartbeat, the Data Journalism Unconference gathered guests from 16 countries, representing the five continents.

Folks from the Financial Times, DW, Al Jazeera, Condé Nast and The Economist took part, but also people from organisations such as Kiln, DJChina.org, or OSF. That gave us a great range of expertise and led to some insightful sessions on the state of data journalism worldwide.

So what happens when you put 45 data journalism experts in a room and get them talking?

Here is a compilation of what we’ve learned.

All-in-all we had lots of fun discussing the issues many data journalists have had in the back of their minds for a while such as: how to convince their boss that data journalism is worth the resources, how to measure the success of their story, what bad practices should they stay away from, and so on.

Tristan Ferne (BBC), Peter Bale (formerly CPI) and Kristina Knaving (University of Gothenburg) led a session on how to measure impact and outcome of data journalism stories

What’s success for data journalism? A good mix of impact and metrics, experts say.

The first session of the day was about measuring success for data-driven projects and was led by Peter Bale (formerly CPI), Tristan Ferne (BBC R&D) and Kristina Knaving (University of Gothenburg). The surprise was that not many people in the room had an official way to measure success of their data journalism projects. Of course there are metrics, yet not everyone agreed on which ones should prevail in order to state whether a data story is successful or not. But more importantly, it was argued that the success of a data journalism project does not rely solely on how many clicks or what retention rate it triggers.

Data journalists and editors in the room agreed that the impact a story makes also needs to be taken into account when measuring success. Did your story help holding the government accountable? Did it help people better understand a topic that’s usually mis-represented in the media? Did it lead to policy changes or did it reveal wrong-doing from the powerful? Everyone in the room agreed that an official guide with tips and checklists on how to measure success in data journalism would be very useful. Hopefully someone will make that happen soon. If you feel up for it, do get in touch.

Are big, resource-draining data visualisations worth the investment?

Many people also argued that the big interactive data visualisations aren’t always the most successful items on news websites. Although they do drive some traffic and can do well on social media if they’re built with mobile in mind, more often than not, simple maps and charts are more popular.

John Walton, Data Journalism Editor on the BBC’s Visual Journalism team, gave a presentation on election data

6 tips from the BBC visual journalism team on election data

When dealing with data, specially at the start and during an election race, you don’t necessarily know what your end project will look like. So the BBC visual journalism team, represented by John Walton, came up with this list of tips they wish they’d had when they first started:

• Keep things simple
• Talk and plan: good communication and planning will save you a huge amount of time
• Don’t use tech for the first time on election night
• If using live graphics, think about all possible scenarios
• Don’t forget there will be a lot of traffic
• Expect surprises

How to pitch or explain data journalism to the rest of your team, editors and others

Do you need approval from your editor before you can jump on that data-driven project but that person doesn’t know anything or has never been very cheerful about data journalism? Do you need to coordinate with the tech team in your organisation to build an interactive graphic but you don’t speak “developer”? Do you want the rest of the newsroom to stop taking your team for a service desk and wish they understood that, no, this responsive map of all election results in history across Europe does not just take 2 hours to build from scratch? So did many people attending the unconference.

And here is a list of tips to help you rally support in your organisation, explain what you do and communicate better as a team during projects:

• Show results: run an example of the kind of story you want to do with data and show your team and editor its success, impact, or outcome
• Show and explain why is your data project worth it
• Try and use common language when explaining your work. It’s important to educate others in the newsroom about what data journalism is so they better understand what you do and what it takes to achieve what types of graphics
• Accentuate on audience impact potential when pitching your story
• It’s also about making their job easier
• Understand what puts them off in data journalism in the first place and find counter-arguments
• Empower reporters by teaching them how to use simple visualisation tools to produce day-to-day graphics so that your team can focus on more intricate data-driven projects or investigations

The Data Journalism Wall of Shame

Best practice in data journalism has been tackled in many conferences these past few years. But what about the worst? It’s often said that people can learn a huge deal from their mistakes (and those of others). So data journalism experts who took part in the Data Journalism Unconference 2017 decided to make a light-hearted list of the most common bad practices in data journalism.

Here is what made the list (whether you agree with it or not is another matter):

• Designing for yourself or peers
• Not designing mobile-first
• Dataviz for the sake of dataviz
• 3D (Pie) Charts !!!
• Thinking big shiny sexy projects are the most popular
• Not catering for colorblind (accessibility in general)
• Not listing sources (or sharing methodology)
• Not doing your research
• Data Porn (not asking the right question)
• Not enough emphasis on UX

Did you attend the Data Journalism Unconference and want to add to any of the lists mentioned in this article? Get in touch at mbouchart@globaleditorsnetwork.org

The Data Journalism Unconference 2017 took place on 23 May 2017 at the BBC Old Broadcasting House in London, UK. You can follow coverage of the event on Twitter via the hashtag #ddjunconf or by joining the DJA Slack team.