Bella Hurrell: From hackathon to front page
Bella Hurrell presented the year-long story of how the ‘Appy Helper’ went from a winning prototype in the 2017 Editors Lab Final to a powerful tool for keeping audiences abreast of some of the BBC’s most complicated news coverage. She shared how her team found collaborators, persuaded key stakeholders, and improved on their idea using audience analysis and Google Trends.
The challenge
At GEN’s Editors Labs, journalists, developers, and designers are given two days to come up with a solution to a specific problem in journalism. The 2017 Editors Lab Final’s challenge was to find a way to keep long-running, complex stories understandable and compelling for an audience that can’t always stay up to date with the latest news developments.
‘The challenge that was given to the teams who were at the hackathon last year which was around making compelling contextual content to explain large scale long-running and complicated news stories’, said Hurrell. ‘On the premise that people often lose track they don’t remember what happened and say they don’t necessarily engage’.
The proposal
The BBC team’s solution was a module with a conversational interface that could help readers understand the context of the story and engage with the BBC’s coverage. Hurrell explained that they went about this using two elements.
- One was meant to catch the reader up by providing ‘chunks’ of information through summaries of past coverage. These updates were delivered through a chatbot interface, visually similar to a messaging app.
- The second was a more traditional, interactive timeline that provided context by going through the events that led up to the story. Between these two components, the user would be able to learn what they need to engage with the article in front of them.
The timeline
The transition from hackathon to newsroom was not without its obstacles. The UK’s snap election and the summer holidays postponed the team’s unveiling of their new tool.
This gap between creation and implementation saw a changeup of the team behind the Appy Helper, as well as the decision test the tool with a variety of stories and styles and to appeal specifically to audiences under 35. The hackathon team also found collaborators in the BBC’s News Labs, who shared a similar goal and idea.
‘So what we did was think, “Let’s collaborate, let’s come together on this and together we can make a better product, and we can leverage it more we can give it more exposure in the organisation.”’, Hurrell said.
Before rolling out the tool, the team was also careful to have enough resources to fund the project for a minimum of six months, enough time to test and measure audience response.
The results
Since first integrating their tool into BBC’s news coverage, Hurrell’s team has created 12 chatbots and four timelines for a wide range of stories. These include hard-hitting reporting such as a chatbot to answer questions on BBC’s ongoing coverage of Brexit and a timeline detailing the recent US and North Korean relations. They also created Q&A chatbots to engage audiences with more niche stories, such as one on baldness in young people.
According to Hurrell, keeping the user in mind was critical to the development of these Q&A chatbots, anticipating what audiences wanted to know, not just what a journalist thinks they should know.
‘So we look at a whole range of different things. We looked at Google Trends, we looked at what comes out when you auto-type into Google search for Brexit’, said Hurrell. ‘We talked to our marketing team. We talked to audiences team and we came up with a range of questions that people were really asking and then we tried to answer those in an informal way that would not alienate people who didn’t know a huge amount about the subject.’
The feedback
Hurrell reported that among audiences younger than 35, the tool was ‘pretty popular’. They specifically appreciated the informal tone and brevity of the chatbot interactions, making the overall experience more accessible and less intimidating.
The other aspect that young audiences enjoyed was the ‘faux-messaging’ format of the Q&A chatbots. Hurrell found that promotional messages on social media captured the attention of younger users that their team tested who said that they would not otherwise have sought out a story on topics like the gender pay gap.
As for criticisms, Hurrell said they repeatedly heard back from their audience that they wanted to have a more of a natural conversation with their chatbot and ask their own questions, not just choose from a list of options.
‘Now, we’re not at that stage yet. We would like to be. It’s certainly something that we’re going to continue to work with our News Labs colleagues on’, said Hurrell.
Along with the positive feedback, most respondents also supported seeing the tool in future news coverage. When asked what stories they should add the chatbot to, readers recommended possible stories from Syria to elections to reporting on slavery after Kanye West commented that slavery was a choice.
Moving forward, Hurrell says that her team will be taking this feedback and looking to see where their team can improve the user experience going forward.
‘We’re going to look at more social media integration. We’re going to do single entry context expanders, as well, so we can put in things just like the timeline but just single entry’, said Hurrell. ‘And we can explore with News Labs how we can eventually get to the stage where people can ask their own questions’.
The advice
‘[W]hen you go back to your news organisation how can you get your idea made How can you make it a GEN success story? How can you take it further?’, asked Hurrell.
- Persuade your key stakeholders — You need to know who they are and you need to be able to explain your idea. Keep a realistic focus on what you want to achieve. Look for potential partners or collaborators in your organisation.
‘And one way of doing that might be to say, “Can we resource a limited pilot? Can we do something for a specific period of time that maybe hand-build that we just keep going just for a limited period of time to see what happens.”’, said Hurrell.
- Find out what the users think — Once you can get a pilot out, you will then be able to gather and analyse data from your audience. This information will give you an idea of what works, what can be improved, and what your next step is.
- Don’t put on blinders — ‘You need to be realistic about your idea’, said Hurrell. ‘Maybe it won’t work out, and it may not be the next big thing, but it could lead on to your next big thing. And it will be a valuable experience’.
