Interview with David Ottewell

Q&A with David Ottewell


David set up, grew and runs Trinity Mirror's multi-award winning data journalism unit. they use data to create exclusive agenda-setting stories for regional and national titles, unique interactive gadgets, and highly-acclaimed projects like their Real Schools Guide. Previously he was a news editor, chief reporter and political editor at the Manchester Evening News, the UK's biggest regional paper.

We caught up with David ahead of his appearance at this year's Digital Day



First up. What is Data Journalism?

In it's broadest form, data journalism involving finding and telling stories from data. That's not very illuminating, I know! Generally it means we analyse lots of spreadsheets, and use coding and other techniques, to break important news that lies within. Data journalism is also about having skills to help us find data in the first place. While plenty is simply published by government and public bodies, we also scrape data from websites and make a lot of use of Freedom of Information laws. Finally, data journalism is also about how we tell stories. We use a lot of visualisations and build a lot of interactives to not just display important data, but to help our readers explore it themselves.

Trinity Mirror launched a specific digital journalism unit with a focus on data journalism back in 2013. How has the unit progressed over the years?

We started with just two journalists experimenting with data. Since then we've grown rapidly and now have 12 full-time members, including two graphic designers, two coders, and a videographer. Obviously that has massively extended the scope of what we do. We now write an average of about 1,500 stories a month based on more than 100 different datasets, for the various local, regional and national titles in the group. In addition we do a lot of long-form data investigations and set-piece projects, like our annual data-based schools guide. The basic ethos has remained the same, though: we've always believed some of the best stories out there are hiding in data.

What kind of sources can be utilised at the hands of journalists to uncover good stories?

There is no shortage of data! In fact, you could easily spend all day, every day as a journalist simply analysing data routinely published by governments, departments and other official sources and do very well indeed as a journalist. It's remarkable that there still aren't enough journalists out there with the knowledge or skills to analyse these datasets, break them down into local components, compare them over time, and tell the stories that emerge. We've always thought the problem in journalism is people trying to hide stuff that the public needs to know. To some extent it is. But there is another problem: stories hiding in plain sight because journalists don't have the time or skills to find them in data. 

In terms of other sources, pretty much everything online can become a source of data. Some of the most popular websites in the world are based on data, after all: Trip Advisor, Rightmove, Just Eat, IMDB... the list goes on.  You can scrape websites for specific information, and turn it into a dataset; you can compare previously-used data in new ways to see what lines emerge.

We also work with businesses and third parties to obtain interesting datasets - one of the projects I'm most proud of was some analysis we did for the 100th anniversary of the start of WW1, which systematically broke down the casualties, for the first time, into the towns and cities they were from. Using data journalism techniques we weren't just able to tell people how many people died, but put names and other information to the oldest to die, the youngest to die, the first to die, the last to die, and so on. None of that would have been possible had we not got the original data from the Commonwealth War Graves Commission, who have been collecting it diligently for decades.

Have there been any specific tech developments or policies which have helped/will help to uncover more truths for your readers?

There's been a huge debate in data journalism over the years about whether data journalists should learn to code. Increasingly my answer is "yes", but not for the reasons others give. I don't want my journalists building interactives or working as front-end web developers. But coding is also useful for extracting and analysing huge datasets, and to automatically monitor data and be alerted when potential stories emerge. It allows data journalists to work much faster, and on a much bigger scale, than by using spreadsheets alone.

Data journalism is all about transparency and holding the powerful to account. With such a wealth of freely available information, how do you go about deciding what’s newsworthy and what’s not?

It's one of the hardest parts of being a data journalist, actually. I'm from a traditional reporting background, and rose up through a regional newsdesk, so I've been forced to develop my news values during my career. They are the same as always: a truly great story needs to be strongly in the public interest, but also be something people actually choose to read, engage with, and share. Where data journalism is concerned, that's about two things: the ability to tell a story, but also the data we choose.

Clearly, we have a lot of audience data available to tell us what stories do well, and to hint at why. We've found, for example, that people love postcode-search interactives which allow them to personalise a story, to get data for their city, town, neighbourhood or even street.

Still, though, we have to choose the right data in the first place. There remains some reliance on old-fashioned news sense. It's also one reason I try to make sure that the majority of our stories are based on ideas, rather than the calendar we keep of upcoming government/public-sector data. At any given time, all our reporters will be working on at least one long-term investigation, too, because we know that is something that matters, and which gives us control of the agenda and an ability to focus on issues we know readers do - and should - care about.

What’s been the biggest story for you/your team since taking the reins of the team?

I worked out recently that we get an average of 100 front pages a year (yes, our work still goes in print as well as online). Every one of those will be a story that would have gone untold without the data unit; every one is an example of an editor saying "today, this is the most important story in our patch", which is exactly what we strive for. Then there are the projects, too, which are more than just "stories". It's impossible to choose a favourite! 

David on Twitter

Get your tickets for Digital Day 2018 here.