Matilda Davies, The Times, about creating a Datawrapper workflow

Problem: The data team was siloed

[00:00:04] Matilda: Hi everyone. It's so great to be here today. Yeah, my name is Matilda Davies. I am a data journalist at the Times and the Sunday Times, based in London. And I've been on the data team for about two and a half years now, and the team's gone on a bit of a journey, as the team has grown, particularly in terms of our workflow and how we work with the wider newsroom.

So that's what I wanted to talk to you about today. The transformation has allowed us to, as we like to say, spread the data gospel across the whole newsroom. And in doing so, upskill the workforce and free up the data team to work on more in-depth data investigations and make bespoke interactives. 

So a few years ago, we were working in a really siloed, separate way. The data team was responsible for all of our data, rapid charts, visualizations, and data analysis for the entire newsroom. And so we're a small team of seven, soon to be eight permanent data journalists. So this meant a lot of our time was taken up.

With making basic charts and maps, like simple line or bar charts, sometimes the same ones over and over again. For most of our reporters, the data team felt quite separate. So we were kind of the geeks over in the corner, and they didn't really understand what we did or what we were capable of and changing that required a key ingredient.

Transforming the whole newsroom

[00:01:46] Matilda: So we wanted to get the whole newsroom to see data as part of their daily roles, not something specialist or separate from them, and that they didn't need to be experts to start using data in their reporting. So we started training up teams. Showing them how easy it is to create basic charts in Datawrapper that not only increase engagement and dwell time on their stories, but help them see where trends emerge, where records have been broken, and what the bigger picture is We utilized our graduate data journalists by, embedding them on key teams like business, sport, and property to show those teams what data could do, and how they could start using data and charts in their day to day reporting. And to make it as easy as possible for people to start building charts and improving their data skills, we implemented a custom theme from Datawrapper, which automated a consistent design and style across all of our charts.

So the changes from this approach became clear pretty quickly. In 2018, more than half of the whole company's charts were being created by one person. And as you can see here, as more people have developed their Datawrapper skills, it's really spread The load across the newsroom. so we've increased the number of people making charts, which has increased the number of charts that we can make as an organization.

And in turn, that's really freed up the data team's time to do what we do best. Data investigations, more complex analysis, and bespoke interactives. 

Four levels of knowledge: "Everyone"

[00:03:57] Matilda: So I see our workflow as kind of four levels of knowledge. we'll start with what we expect everyone to be able to do. At this point, now almost all of our journalists that work with data, even in the most minimal or tangential way, can paste data into a Datawrapper, to create a basic line chart, Here we've got an example of a chart showing Bitcoin's price after each halving event, um which created this week by one of our money reporters. reporters We also encourage individual teams to learn functions that are useful to them specifically. So here's a chart that was created by one of our sports reporters this week. Flags are particularly important on their desk, and they now often incorporate those into their daily charts. And as you can see, the custom theme comes in handy here.  Our reporters can really easily make charts that fit with our house style and use our color scheme without needing advanced knowledge of Datawrapper or graphic design. We also have additional oversight, just in case. So we use Datawrapper's Slack integration for our channel called Dataviz Review, which posts every time someone publishes a chart. So that way we can keep an eye on what's going out and make tweaks or offer extra training where it's needed. needed 

"Creatives"

[00:05:04] Matilda: So, as we've rolled out this basic training, we found that some reporters were keen to experiment with what Datawrapper can really do.

All of our graduates, whether they're news reporters, sub editors, picture journalists, they all do a secondment on the data team where they spend a few months with us, learning what we do. With us learning what we do, with more time to play around with Datawrapper's functionality, a lot of them became interested in making more creative charts.

We also found this when we embedded data journalists on other teams, like Money. When they realized just how much it could do, a lot of the reporters were excited to learn more about it. So this is a chart that was made by one of our health journalists this week, after he was dispatched to India to report on the world's largest vaccine factory. And having reporters in data heavy areas like health or crime, with advanced Datawrapper like this, really lightens the load for the data team, and means we have more time to train up other journalists and work on our own projects. This chart was built this week by a graduate who's temporarily on our team team called Jess Sharkey.

And training up young journalists to be Datwrapper pros and then sending them out into the newsroom has been really key to upskilling new teams. And, as I say, spreading the data gospel, these reports still lean on the data team. We have Slack channels, and obviously, we're there to support in person with analysis and to check the final products 

"Coders"

[00:06:38] Matilda: So other people, we found, were less interested in the creative possibilities of Datawrapper, but were excited by the technical aspects of it. Some of our journalists had tried to learn coding before, or know a little bit, but weren't sure how to implement it in their journalistic practice. And there are lots of areas of our newsroom where the same data sets get reused and recharted, kind of month after month and year after year particularly in areas like politics or health or the environment, or some of our more bespoke products like the Sunday Times Rich List. So we've now automated a lot of these using Datawrapper's API. So we have R scripts that pull and analyze the data, which then plugs directly into Datawrapper to update the charts that we use time and time again.

And it's a really good place for journalists that are interested in coding to get started. When I first started learning R, I got used to the process by running scripts that other people had built. And then I could figure out what each line of code was doing until I could create those scripts myself.

So we train people to download and use the basic functions of RStudio, and access the team's GitHub repository, so that other journalists can do exactly that. And then they can start learning R in a way that's directly applicable to the work we do every day. So using Datawrapper's API, we've now automated all of our election polling trackers, environment and climate change charts, and other kind of data heavy issues that um get on again and again. 

As far as we see it, the more people that have these skills in R and can run these scripts the better, because the more we can automate as reporters apply those skills to their own beats, and it makes their lives and our lives easier simultaneously.

"Experts"

[00:08:55] Matilda: So then we have the experts. Everything I've talked about so far about upskilling the rest of the newsroom and redistributing the workload, means that the amazing data team that I'm super lucky to be a part of, um gets more time to work on bespoke projects and investigations and in-depth analysis.

So my colleague Venetia Menzies recently conducted a data-led investigation that uncovered the anxiety drug that has the fastest rising death toll of any drug in the UK, pregabalin.

 My colleague George Willoughby mapped the shocking spread of Japanese knotweed across the UK over the past hundred years by combining QGIS analysis with Datawrapper's amazing mapping features.

And I've been able to use my extra time to experiment with new Datawrapper hacks to tell equally hard hitting stories like how successful every Taylor Swift album is, or how many recycling bins people have in every corner of Britain. 

So encouraging people at all skill levels to get more involved with data analysis and data visualization has been pretty successful.

We're still spreading the data gospel and teaching people how to use data more effectively and utilize Datawrapper's incredible functionality and incorporate data more into their daily work. But what we're ultimately striving to achieve is to make the newsroom not view the data team as the geeks in the corner, or even as an editorial support desk, but as specialist journalists in our own right. We're still always on hand to help our colleagues when needed, but using data and Datawrapper in particular, we think should be a part of everyone's job every day. 

Yeah, and that's me. Thank you so much for listening. It's been such a pleasure to be involved in Datawrapper's conference, and the other talks have been brilliant as well.

You can get in touch with me here, I'm more than happy to answer any questions or link up after the conference. I don't know if we have time for questions now?

Q: Buy in to invest in training the whole team?

[00:11:22] Shaylee (host): Thanks Matilda. We have time for one question, but I did want to say: excellent talk. It was very interesting and well thought out. We do have one question that I wanted to talk about. This one is from Yanika Borg. I hope I'm pronouncing that correctly. And they say: how and why did you get the buy in to invest in upscaling the whole team? 

[00:11:45] Matilda: That's a great question, Yanika. As I say, it all started from the way that we were spending our days. Much of our days were spent building the same line chart over and over again. And I think a lot of the data journalists got a bit frustrated with it. Because, we're journalists, with our own ideas for big projects and things like that, that we just didn't have the time to do. So that was the spur of it. And the more of those projects we were able to get out and prove what we could do, the more receptive the senior management were to us changing the way that we work and, as I say, upskilling people in individual teams as well. Yeah, so that's how we started it.

[00:12:41] Shaylee (host): Fantastic. Makes sense. Well thank you once again.