Alan Smith (@theboysmithy) leads the Finanical Times' newsroom team of data reporters and visual journalists. A data visualisation specialist, his TEDx talk Why You Should Love Statistics was featured on TED.com in 2017. Alan is the author of How Charts Work, a handbook on designing with data using the Finanical Times' principles. He is also Honorary Professor of Practice at UCL's Social Data Institute. Before he joined the Finanical Times', he was head of digital content at the UK Office for National Statistics where he was awarded an OBE in 2011 for services to official statistics.
Episode Description
Journalists are often tasked with making complicated information accessible to a wide audience. Given mere minutes or inches to communicate information that might have taken a researcher pages to explain. Financial data can be particularly tricky to translate as it can require not only a comfort level with numbers but also some understanding of financial markets or economic principles. Reporting on financial information is the focus of this episode of Stats+Stories with guest Alan Smith.
+Full Transcript
Rosemary Pennington
Journalists are often tasked with making complicated information accessible for a wide audience, given mere minutes or inches to communicate information that might have taken a researcher or analyst pages to explain. Financial Data can be particularly tricky to translate, as it can require not only a comfort with numbers but also some understanding of financial markets, or economic principles. Reporting on financial information is the focus of this episode of Stats and Stories, where we explore the statistics behind the stories and the stories behind the statistics. I'm Rosemary Pennington. Stats and Stories is a production of Miami University's Department of Statistics and media journalism and film, as well as the American Statistical Association. Joining me as always is regular panelist John Bailer, emeritus professor of statistics at Miami University. Our guest today is Alan Smith. Smith leads the Financial Times newsroom team of data reporters and visual journalists, a data visualization specialist and the author of How Charts Work, a handbook on designing with data using the Financial Times principles. He's also honorary professor of practice at UCLA Social Data Institute. Before he joined the Financial Times, he was head of digital content at the UK Office for National Statistics, where he was awarded an OBE in 2011, for services to official statistics. Alan, thank you so much for joining us today.
Alan Smith
Well, thank you so much for having me on. I'm really pleased to join you.
Rosemary Pennington
How do you go from working for official statistics to a newsroom?
Alan Smith
That is a very good question. And I think actually, it's a really interesting place to start. Because when I started my data visualization career, there weren't really data visualization careers, right. I mean, I think that's the really kind of funny thing with it is that I started off at the statistics office in the UK running a mapping or like a cartography service. And eventually, I became increasingly interested in the way that the very early web back then was going to make it much more interesting for us to present and show information in ways that people could engage with. So my masters focused on data visualization in a much broader context, not just maps. And with that, I mean, I don't know if you've ever seen the film Jerry Maguire, where he writes the big notes. It says, this is the way the organization should work. And everyone agrees, and then he gets sacked the next. I wrote about Jerry Maguire sort of a letter, but amazingly ONS was so supportive and allowed us to set up a small little data visualization team and said, Yeah, let's explore this. So just getting to do data visualization for ONS for a statistics office was, for me, a real step forward. And then I guess what I hadn't realized is that the way that this data visualization on the web was allowing information broadcasters to really rethink what they were doing, actually brought two very quite different kinds of information broadcasters into much closer orbit. So, you know, a government statistics department and kind of Fleet Street journalism, traditionally didn't really have much by way of crossovers in Korea. But by the time I joined in 2015, it seemed like such a natural thing to do, you know, I'm doing data visualization for a statistics organization. And I'm now going to do the same sort of thing for a newspaper that cares about giving readers quality, reliable information. So it was quite a short hop in the end. And I kind of liked the idea that data visualization can allow and facilitate those sorts of career trajectories and get out of the traditional silos that people traditionally might have felt constrained by.
John Bailer
So, you know, we'd be thinking a lot about the idea that with these new ideas and new tools, there's an evolution of the way reports are presented. But there's a secondary component to this, which is there are now novel ways in which information is going to be presented. So it was at least initially, to start with, can you talk a little bit about how some of the reports that you worked on with ONS or maybe some of the early stories with the Financial Times Changed with some of the tools and ideas and perspectives you were bringing?
Alan Smith
Yeah, that's a great question, John. So I feel like in the early days of our work at ONS, the remit really focused heavily on experimentation, right, and experimentation and learning, learning best practice by doing I mean, in fact, it was really interesting that when we set up the data vis team. Originally, it wasn't placed in its communication department, right. It was actually placed in the statistical methodology department, which made a lot of sense to me, because it was kind of what, you know, Hans Rosling would have called the final six centimeters of the survey cycle. It's all about getting information back into people's heads. And so we had a big remit to experiment ONS. And the thing that was really lovely about that was, everything was new, right? Like, I mean, it wasn't like ONS had this big kind of legacy of data visualization, that it was kind of looking to change. It was just a completely brand new area. And so we experimented a lot with really interesting ways of just using data in kind of maps and charts, and just seeing what things that interactivity and animation could do for perception and kind of understanding of graphics. And that was a really nice place to play with it. And what I think really helped was, once we published some of these things, a lot of our users, some of whom were in the media, started to realize this stuff was game changing, you know, like, and so they would syndicate it. They would feature it on TV, and suddenly, I was sitting there one day, talking to senior managers thinking, well, this is amazing, right? Like, this is like a fast, a shortcut, into getting data into the public arena, you know, like at that stage, the stats office were really keen to inform what it called sort of evidence based debate. And so we found that was a really good way of doing it. I think the challenge at the FT is similar but slightly more nuanced, in that you're suddenly in a different place, you are much more central in the debate. As a news organization, you have the impartiality of the government department, which can say things about data, but only to a certain point. And I think the thing with taking data visualization mainstream in newspapers is that you really need your content to kind of stand up and be noticed in what is an increasingly more crowded marketplace. So you use the word novelty, right, which I think that's definitely one angle on how to do that. But then you've got all the other components as well, right? Like, you know, making sure that you're using reliable techniques to even source and compile the data you're presenting. And, and kind of marrying that with visualization and techniques for accessing data in ways that work for readers. And so I mean, I think I learned a lot from ONS. And I'm continuing to learn a lot at the FT because we're talking about very different audiences in some ways.
Rosemary Pennington
I was gonna ask you, I'm looking at a collection of stories that I think your team was involved in. And it's quite a selection, you've got data visualizations on the wars on Ukraine and Gaza, obviously, lots of economic visualizations, how do you decide what is going to make for a good visualization story? And then how do you go about sort of producing it?
Alan Smith
That gets to the heart of what we do, I think very nicely, that question. I mean, I think one of the things that we're increasingly trying to sort of industrialize as an approach in the newsroom is, I think, in the past, in the news industry, charts and maps were often sort of like decorative parts of the story, you know, the page, right? So, you know, reporter writes 1000 words, files it to an editor who, close to publication, will request, you know, a picture and a map and a chart, and that will be used largely to break up the text, it's a very sort of, you can imagine a very sort of print oriented workflow that has kind of spilled over into online in the past couple of decades. And you know, that model has been prevalent for a long time, right. So if you're trying to change that model, you're pushing against quite a lot of inertia and history. So I think we do several things, which are increasingly allowing us to move in a slightly different way. The number one thing is for reporters at the FT, whether they're in my team or not, to become increasingly comfortable looking at data before they're writing about it. Right. So in that sense, visualizing data before you are laying out, kind of final story. So using data visualization more as an exploratory tool for helping to find those stories. That is definitely a habit that we're trying to sort of form more of in the newsroom. And then I think the second thing is, once you have a story that you want to tell someone, I think in terms of the visuals, a good rule of thumb is, you know, does it surprise you? Right, like, I mean, if it doesn't surprise you, how many readers is it going to surprise, right? And then who wants to show someone something that they already know? Right? Like it feels to me that what you're trying to do with visualization much of the time is to contextualize and enlighten. Right? I can. And I think if that context, if that kind of attempt to show the data in a certain context doesn't surprise anyone, then you have to question its value, right? Like, I feel like, that doesn't mean that you then go searching for every sensational metric that you can find. But I think it's a good rule of thumb. If it surprises you, it's likely to surprise the reader.
John Bailer
I find it interesting that this historic model, that the graphics and you know, sort of the other analysis, part of things were kind of this ornament that was hung on to the piece at the very end, I find for my own particular workflow, when I think about research, I found that I needed to have the visualization before I started the writing. Because if I couldn't visualize the story that I wanted to convey, I wasn't sure it could be done. And I remember we had a guest many, many episodes ago, that was a financial reporter that talked about statistics as narrative elements. And that statement really just still resonates with me. And it seems like I really love the fact that you all are also sort of flipping this workflow on a TED, as part of trying to understand these complex stories about the the impact of war, the impact of financial change, what what's been one of the hardest stories that you've tried to visualize.
Alan Smith
So I think, I mean, it's impossible to have a conversation about the last five years without talking about COVID, because it kind of covered all of the challenges of doing quality data visualization on a fast moving and evolving news story that basically affected everyone, right, like in a different way or another. And I think the thing with COVID, that was really good is that you could go through with a big list of things to do with COVID. And say, this was slightly different. So for example, like, we found ourselves in data gathering mode more than we ever had before, because there was no single initially, at the start of the pandemic, there was no single one shops, a one stop shop for all the data that you would need to tell a COVID story, you had to go and do it. And people like John Byrne Murdock were incredibly quick off the blocks , kind of looking at what are alternative data sources that we could use to help try and understand some of the things that we're seeing as they evolve. And so things like the use of excess deaths, for example, and compiling comparable excess deaths data internationally. That's not what news journalists normally do, right? Like this kind of massive kind of aggregate and kind of sorting and aggregation exercise with data to start with, then, of course, you've got so many different stories building out of this one overarching topic that the question is like, where do you focus next? Like, what is the story? You know, given that there are so many stories, what do you focus on? And I think there was one story that we published about six months into the first year of COVID, which kind of struck me as being a real step forward for what the team was doing. We did a piece on what we have understood from the last six months? Let's look at this through charts and maps. And what if we understood the global crisis in data we called it and the thing that was interesting when we were structuring that piece was we said, pretty much in this story, there's just lots and lots of paradoxes. And you could basically pick a paradox per chapter in this story and look at each paradox from a data perspective. So it was things like, I think the first chapter was our comm. China was in the middle of the biggest human migration on the planet, but managed to avoid a massive outbreak that subsequently hit Korea and Italy, for example, you know, so what about that paradox, you know, when is a death not a death, right? All the different systems that were being used to capture and record COVID deaths, for example, and why? Why nowhere in America was like America, because of course, you were looking at things, statistical concepts, like how well averages describe real people's lives, right? I can. So there were just some really interesting things in COVID, that kind of took you through a whole load of I would call kind of statistical thinking, in a real world story, right? Like, I think that's a really interesting thing, even the charts we're doing, which feature kind of log scales to understand exponential growth. That was really just unusual for a news organization to be doing that sort of stuff.
Rosemary Pennington
You're listening to Stats and Stories. And our guest today is the Financial Times Alan Smith.
John Bailer
Well, I have to confess that I'm a huge fan of the visualizations. And, you know, as I had mentioned to you before we started recording today, that when I was teaching data visualization classes during the pandemic, I found that the resources that were being produced by your team were just fantastic in terms of being integrated with text, but also very effective standing alone. And it truly was amazing to see the ability to toggle between, you know, original scale and then and then the log transform scale I. I'm sure that there were many people, you know, sprinting back to their middle school days where they may have first encountered such ideas. You mentioned in other contexts, the idea that charts should be things of wonder as a quote that I remembered of yours that I really loved. But you also thought that we need to be promoting greater efficacy, if an idea is sort of in parallel with numeracy and literacy of various forms. Could you talk just a little bit about what does graphic efficacy mean to you?
Alan Smith
Yeah, I think so. I mean, I think it's a really interesting concept. It's about, you know, we're used to talking about numeracy and literacy. But of course, when we're talking about data visualization, you're actually talking about more than both of those things. It's about being able to understand information visually, and being able to pass it in a way that makes sense, right, that allows you to turn what you're seeing into sort of actionable intelligence, actionable information. I think the secret, to some extent to what we did during COVID was the fact that actually in the run up to COVID, and in the few years before COVID, we had spent a lot of time working on graphic efficacy in the newsroom, right? Like, we spent a lot of time not just within the visual and data part of the newsroom, but across the whole newsroom, really talking to people about how if we were really going to make a difference with the way that we presented data, and to distinguish ourselves from everybody else, that what we wanted to do was to become much more comfortable as an entire newsroom, working with data visualization, and presenting data in sort of not just novel ways, because I think anyone can make something novel, but like, in ways that actually made a difference. And so we've created this resource called visual vocabulary at the FT. And this was like a learning resource for the rest of the newsroom to sort of say, to steal someone else's phrase that there's a bit of a grammar to graphics, that there are different ways of presenting data, depending on the type of data that you're presenting. And so in the years before COVID, we started to sort of get editors to come up to me and ask about Sankey diagrams. Now, believe me that but that's not normally how newsrooms talk. Right? They don't come over, you know, and so we'd spent a lot of time building and understanding of different types of charts and data so that we could move away from just the sort of regular line on pie charts and bar charts, you know, we haven't moved away from them, but we've kind of extended beyond them, you know?
Rosemary Pennington
So I'm a journalism professor and I teach multimedia. And often I'm trying to get my students to think of presenting data in visual ways. And I wonder what advice you have for the journalist, right? So you are coming from a statistics and data visualization background, but what advice do you have for people who are not that well versed in that sort of that realm? For how they might engage with this? Because it seems like it's growing increasingly important in newsrooms?
Alan Smith
Yeah, I mean, I have to go back to what John was saying about narrative purpose. I think that's the way to sell it to journalists, right? That these are narrative assets. These are things that can hold a story in their own way. They can live within bigger stories, but I kind of like to think of charts in some ways, just as micro stories, they have to be self sufficient. They have to be on their own. And you know, for journalists, here's a great technique to write a story. Start with the charts, put a set of charts on a page, and then right around them, write, like and don't put any old charts on page, put the charts on the page, the ones that I said, that surprised you that connect in a certain sequence that answer certain questions, put them on page, write the titles for those charts, and then write around them, right. And I think what you end up with if you do that is just a much tighter piece of journalism, right? And when it comes to things like how do I visualize this data, you can use that narrative intent to guide you. And that's how we kind of created the visual vocabulary to allow people to think in that term. So if you're writing about a surge, inward investment, that's a hint in that word right there that you're looking at change over time, right? This surge, right. Like, as you know, if something fits to the top of the rankings, well, maybe pick a rankings chart, right. So it's about the narrative purpose of the chart. And I think that that worked really, really well so that by the time COVID arrived, we had a set of journalists already quite comfortable doing this, you know, John Burn Murdoch could have been doing it for a while, but also just any number of ft reporters started to become much more accustomed to thinking about charts not as, here are some numbers. But here are some patterns in numbers that are important.
John Bailer
So when when you're looking at some of the graphics before they get rolled out, whether it's on a web page, or whether it's in print, do you have this iteration where it's presented to sort of a team of reporters who go, oh, that doesn't work for me, or, you know, this would be more effective if you do that kind of what type of iterative processes there. And that's also recognizing that I know that there's tight constraints often on what you're having to do.
Alan Smith
So timing will affect a lot about how you operate and execute. I like to think that in anything where there is at least some leeway, I mean, because we do some big projects, and some sort of smaller projects. And then obviously, breaking news, they're all kind of different on different parts of the spectrum in terms of time. But we do a lot of socializing, chart design and showing internally within the team and also working directly with reporters. I mean, I think this is another thing that's really interesting is that if you go back historically, in the newsroom, in the newspaper industry, reporters have generally never talked to graphics desks, a lot of them were banned from doing so right like it because it was an editor's relationship with the graphics desk. That's the way that the newsroom workflow would work. And so what we've been trying to do is to unpick that and say, Well, look, let's pair up reporters who might be covering particular beads with visual and data journalists early in their process. So we get involved when they're asking a question, right? Like, or, you know, at the point of origin for the story, rather than I've written this story, what can we do with it right, like, because that limits your options, much more. So we socialize a lot. And actually, interestingly enough, when I run data visualization training courses sometimes for the Royal Statistical Society in the UK, and whenever I run these courses, the ones that are the bits of the course that are most popular, is when I give people some pens and a flip chart and tell them to stand around a flip chart in little groups, and design some charts together. Because not many of them, when they look back on it, realize that it's quite a solitary activity, right, chart making people kind of people hunched in front of Excel, or whatever they're using, doing this thing on their own. And the moment you actually tuck it around a flip chart with someone, and there's three or four people chipping in some opinions and pulling it around and whatever, then socializing the chart designers is a really, really useful process.
John Bailer
I love this image of embedding journalists and journalists, you know, this is sort of the embedding the data in graphical journalists with other journalists and doing the reporting. I know some of these different outlets for the work that you do. And you know, this idea of static versus dynamic or fixed versus interactive. And with a within print, you have some limitations. But when you start thinking about the kind of stories that you want to tell within a web context, you have a lot more freedom. And with the interactive, you can also personalize. So there's some other aspects of it. Can you talk a little bit about that? The things that you think about when you're presenting information between some of those decisions?
Alan Smith
Yeah, definitely. I mean, I think actually, when I joined the FT, it was really interesting. By the time I joined the FT from ONS, ONS had already got rid of its print portfolio, like a long time ago, I think it was probably a you know, it published PDFs online, perhaps, but like the actual print product had gone. And so for me, that was a very clear focus on online. And of course, as soon as all of this new technology started emerging, at the start of the century, I was kind of into JavaScript and writing SVG code and go, Oh, wow, you can hover over this and pop over this and do that. And it was all wonderful. And coming back to the FT, it's really interesting, actually, because the FT still has a print operation. And I love the fact that we still do print sometimes, you know, as a sort of big team project will find its way into print. And you talk about constraints. One constraint that print doesn't have is it doesn't have to work on a three inch mobile phone screen. Right? So I love those days in print, where we can go big and just take over the whole because the FT is a pretty big printed newspaper, right? It's a broadsheet. It's big. And so on something like the UK is Budget Day, you'll see a double page spread of charts taking over the entire paper, right. So I still love that. And I think that's got a real place in the sort of data visualizers kind of toolkit with online, it is interesting, because we do need to make sure that we're sort of serving mobile first content, right? Because increasing audiences, especially younger audiences, are increasingly mobile, increasingly social. And that does create problems for interaction design, right like that. You can't suddenly just publish a scatterplot and tell people to just tap the dots, right? Because on a mobile phone screen, you've got real accessibility challenges with doing that. So I think you need to think very carefully about how you use interaction and I think there's also a very fundamental point about that when you make something interactive, what you're doing is you're asking the reader to lean forward and take control. Right? Like, you're actually it's a sort of lean forward movement where you're saying, this story is so nuanced, I need you to see different views of it. So you need to lean forward and take control. And to some extent, that's counter to the journalistic method, right? The journalistic method is I'm going to tell you a story, sit back, right. And I'm going to tell it to you. And so I think, interactivity is a great option for us to have on an online, but there's that old quote from Spider Man always lingers in my head, sort of with great power comes great responsibility, you know, you need to make sure you use it in a way that that matters. And I think Archie at the New York Times, several years ago, made this great general point, he said, if you ask the reader to do anything other than scroll, something spectacular has to happen, right? Like I mean, you've got to be doing something that makes it worth the while. And I think the days have just, you know, hover over this, and you can see the number. It's way beyond that right now. Right? Like I think static, well designed, static graphics are still the lifeblood of data visualization for that reason.
Rosemary Pennington
Well, that's all the time we have for this episode of Stats and Stories. Alan, thank you so much for being here today.
Alan Smith Thank you so much for having me.
Rosemary Pennington Stats and Stories is a partnership between Miami University's Department of Statistics and media, journalism and film, and the American Statistical Association. You can follow us on Twitter @StatsandStories, Apple podcast, or other places where you find podcasts. If you'd like to share your thoughts on the program, Send your email to statsandstories@miamioh.edu or check us out at statsandstories.net, and be sure to listen for future editions of Stats and Stories, where we discuss the statistics behind the stories and the stories behind the statistics.