I was the editor of the Mirror’s data blog Ampp3d for a year. These are some things I learned trying to make data journalism mainstream.
1. Interactives make an enormous amount of sense as a doorway to a big spreadsheet. They allow for that great kicker for a story – personalisation. A general story about crime figures not catching your attention? How about knife attacks on your route home from work – you’re more interested. Interactives are a natural tool for data journalists.
2. People love real-time. Even fake real time. Though it’s only the big numbers it really works for. You know that urban legend – if Bill Gates dropped a $50 dollar bill it wouldn’t be worth his time to bend down and pick it up, because he’d already have made twice that in interest in the 10 seconds it would take him.
One of our biggest hits was basically a visualisation of that, with a different millionaire.
3. Polls and quizzes are great and there is no reason you can’t use a question and answer format to deal with complex and weighty topics.
This story – Does the Home Office Think You’re Gay? – was more new formats than data journalism. But quizzes are another good way to find yourself in the dataset, in the news story. Even to build empathy by putting you into the mindset of the participant – in this case a gay refugee. For a nominally small story about a gay asylum seeker this actually beat out Kim Kardashian to make the most top 5 most read on the Mirror’s site.
4. I think you should sit next to your developer. Not least so you can draw things on bits of A4 and generally keep on top of what they are doing. Just saying “oh make a tool to go with the ONS dataset” has from experience, turned out not to work. So to contribute to this debate – yes – sit next to them.
5. Think about UX and keep the message simple. One of the reasons it was interesting to work with my old boss Martin Belam was that he brought in a UX background that is rare in journalism. One general tip. Keep things devastatingly simple. This was our most shared graph ever. And let me tell you we did a lot of graphs.
6. Deal with the mobile screen. It’s long. Writing for mobile not desktop is increasingly becoming the norm. But it’s more than just putting things in one column, you can really make the format work for you.
Another UX / screen design thing Martin did once or twice which I really loved is making a kind of comic strip that runs vertically down the page and so reveals as you scroll down it. EG. This story about the effect Tory education minister Michael Gove had on teacher morale.
It makes space part of the story.
We did it sometimes with images where you redo a photo or graph several times with more notes on it each time. Again the idea is it reveals the narrative as you scroll. Yes, Buzzfeed’s lists basically work like this, but this takes it further.
7. Of course you should annotate graphs. You’re a journalist – you’re telling a story. Use the graph to tell a story. Thanks
9. Developing can be extremely creative. The games on our sister site UsVsTh3m plus some of the widgets we managed to produce on Ampp3d really opened my eyes to how creative coding and developing can be. After all it’s fun to control the whole page, rather than just the 600 words in the middle of it.
That plus a few macroeconomic factors affecting the journalism industry are the main reasons I decided to make the switch into becoming a developer when Ampp3d was shut down in June 2015.
[Thanks David Stevenson for the beautiful graphs above, and more, and Federica Cocco and Sophie Warnes.]
David Foster Wallace is one of my all-time favourite writers. Though sometimes he can be a gruelling read as this recent clickhole post featuring an incredibly bleak quote from him nicely illustrates.
His thoughts about modern journalism are a little depressing too…
Anyway. I’ve finally got round to The Pale King, his unfinished last book about the IRS, the American tax service. I knew this would have interesting insights about data overload, boredom and attention. It really does. Here’s one. I think it sums up one of journalism’s big problems in this modern era – being boring.
“Consider from the Service’s perspective, the advantages of the dull, the arcane, the mind-numbingly complex.
“The IRS was one of the first government agencies to learn that such qualities help insulate themselves against public protest and political opposition, and that abstruse dullness is actually a much more effective shield than secrecy.
“For the great disadvantage of secrecy is that it’s interesting. People are drawn to secrets; they can’t help it.”
[p.85 in the Penguin paperback edition]
You get the gist. There are two problems:
1. Finding the story in the first place in a tsunami of excruciating detail.
2. And if you do find a story, telling it in a way that doesn’t drown the reader in facts and jargon and make them click off your story and onto a witty story about Tinder or a bear. Good luck.
The rest of the quote, since it’s all so good:
“Had the Service tried to cover up or hide its conflicts and convulsions, some enterprising journalist(s) could have done an expose that drew a lot of attention and interest and scandalous fuss. But this is not at all what happened. What happened was that much of the high-level policy debate played out for two years in full public view […]
“Even in the financial press, there was hardly any coverage; can you guess why? If not, consider the fact that just about every last transcript, record, study, white paper, code amendment, revenue ruling, and procedural memo has been available for public perusal since date of issue. No FOIA filing even required. But not one journalist seems to have ever checked them out, and with good reason: This is stuff is solid rock. The eyes roll up white by the third or fourth. You just have no idea.”
I was interested to see that The New York Times’ data journalism blog the Upshot recently published a piece called ‘Death to Data Journalism‘. This is a key sentence:
You’ll notice that at least a couple of us find the phrase “data journalism” to be a bit strange.
Data journalism, ultimately, has the same aim as “quote journalism” and “anecdote journalism.” They all aspire to be “fact journalism” or, more eloquently, journalism.
So regardless of the source – journalism is journalism. It’s what I said in this interview I did with journalism.co.uk about Ampp3d, the data journalism site for the Mirror that I edited.
Editor Anna Leach told Journalism.co.uk that after a year of working on Ampp3d, she sees data and numbers as just another source for news stories.
“I don’t think there is something that intrinsically different in data journalism from ‘normal’ journalism and I think it’s a little bit set to one side,” she said.
“A number is evidence in the same way a quotation is.”
The dichotomy between data journalism and normal journalism is false in many ways. The division probably reflects different skill sets within the newsroom rather than qualitatively different stories.
Sure – I haven’t thought of a way this could be used to clarify or illustrate a news story yet – but, in the meantime…
This is a bit tidier than some of my other efforts.
I copied a lot of the bounce-off-the-edges code from this great jfiddle document.
Bugs include the fact that the results text sometimes gets stuck part way down the page, and that the reload button reloads the whole page – not just a part of the script. Which is all it needs. Something for next time…
It’s been great to work with Conrad at Ampp3d. And one of the many things I’ve learned from him is to make sure I align everything that can be aligned on Illustrator. He does not like it when things are not aligned. That is what this game is about. Miss you Conrad.
The end result was the ampp3d coffee machine story.
Our designer Dave drew it to make it look good, added some internet game tricks he’d learnt from UsVsTh3m and our developer added the CSS so it all fitted behind the design. So it ended up looking like this.
To see it work properly – check it out on the Ampp3d site here.
Otherwise I’ve just bound addition functions to the coffee buttons, and multiplication functions to the lifetime bill, and collected them all in the variable runningcost.
About a million years ago – three years ago – I went to New York and I begged Jessica Coen the editor of Jezebel to have a coffee with me. She obliged. Which was brilliant because Jezebel and Gawker Media are the best in the blogging game, and of all the blogs, I especially love Jezebel.
I asked her some questions like “what IS a good blog”, filed the answers away, and left it all on my hard drive for four long years. I found it again.
The bit I remember from the conversation – and have often referred back to is her point about Seinfeld. It’s choosing to blog about something that people notice, but isn’t enough of a “news story” to break into the news cycle. 80% of Buzzfeed articles consist of this.
I also asked her whether there was a distinction between journalists and bloggers. Jeez. It was 2011, forgive me.
Here’s a few  pointers from the best.
ME: What are your non-blog inspirations? I remember a guy from Gawker saying something about writing what people were saying at parties..
JESSICA: I remember Nick Denton saying just after I joined – like in 2004 or something – “On the site” [which was just Gawker at the time but I think it really does inform all blogs period] “what we write about is what people are talking about when they’re not working.” It’s bigger now, but when it was just a small media site – it was ‘what are journalists talking about at the bar?’ That kind of thing.
I think that sensibility kind of informs everything we do: which is what are people talking about that nobody is necessarily writing about or saying? And I still think that’s at the core of what we do.
Obviously a huge part of what we do now is aggregate and discuss and further stories that are already out there. But I think that the best pieces on any blog are almost Seinfeld, you know Seinfeld? The best parts of his humour are always – what’s the deal with xx?
And that kind of thinking – like when you’re talking with someone and you ask “what’s the deal with that thing on the roof over there? [ points at a roof] Seriously I don’t get what the deal is with that thing on the roof?” And then if somebody writes about it I’m like “oh my god – yes”.
So it’s pulling something out of that everyday environment and saying – “actually look at this properly”
Yes, or if you go to dinner multiple times with different groups of people and the same conversation keeps coming up – what are people arguing about? What are people caring about that isn’t just straight news? Or if it is a straight news story – what’s really screaming at people? So I think it’s not just what’s news-worthy, but what is it people want to discuss.
What are the qualities of a great blog post?
The writing. You have to pull the reader in. With writing online particularly. Any platform online is so frantic, frenetic and crazy and loud with pop-up windows and blinking banner ads and god knows how many other windows, tabs and programmes the reader has open. So in order to hold their attention as a writer, you have to write bigger. More voice, more personality, more attitude, more of everything. You have to fight for their attention a bit more.
Especially on blogs, I think at this point that people who regularly read blogs know what to expect in terms of how a post is laid out: here’s the intro, here’s the block quote, here’s the kicker – whatever.
So you have to make sure that whatever you’re doing with that grabs their attention enough so they don’t just zone out and not really read, and that means that the writing has to be really big.
Then the peg has to be really clear. I’d say your first three sentences have to be really killer. Not unlike a newspaper article, but it’s not quite the same as the inverted triangle theory of writing for newspapers because you have to have a compelling hook to bring people in. If you think of great blog posts they’re usually really well written and really funny, maybe not like slapstick funny but entertaining. I’d say that’s a big part of it.
Be mindful of length. Sometimes less is more. I don’t mean like 3 words long but 750 words might be better than 1500 words. And I think people’s eyes really glaze over and go totally dead when you are writing too long on the web and they’re not even reading after the seven hundred words.
So a great blog post. What’s your point? It should be really clear, the reader should know what they’re reading right away. So it’s a balance between the headline and the first paragraph, but I think that the way the web is working now, your headline has to be really clear.
Unfortunately there’s not enough room for witty headlines like there used to be. The writer/creative in me hates that – it drives me nuts not being able to be as witty in headlines -but still there’s a way to play with it a bit.
The headline has to be clear because people are going to know what they want to read by the time they’ve finished that headline and if you’ve got any chance to get them after that, it’s those first three sentences.
I asked Jessica about her favourite blog that she’d written to date. She mentioned one, I think it was this one, and then talked about it as a blog.
It was not a important piece, it was not like journalism. It was to me actually kind of what blogs are… they exist for us to say what we want to say. Yes we can put journalism on there but they’re also like “hey this happened, here’s a story for you”.
I’ve put plenty of stuff up which is like “okay – here’s a silly story for you” and it does really really well.
Because that’s what makes blogs blogs, that’s what sets us apart. We can sit and summarise all the news like everyone on the Huffington Post site and we can sit and write analytical pieces like Salon and Slate – we do that too – but we also just write stuff. I guess this goes back to what we’re saying earlier that there should be a certain element of conversation to your writing. Inherently conversation is just what you’re talking about with your girl friends and guy friends. I find that often really does strike a chord.
This solved my Christmas card problem last year.
And in a frenzy of trying things, here is Timeline.js. A Timeline of making Thanksgiving dinner last month. Much more lush than I expected. Very simple to use, and all built out of a google spreadsheet.
You don’t have to pretend to be interested in this series of events but the tool is cool. The slideshow thing is sexy, though arguably detracts from the getting an overview of time. For example, it’s great for a narrative but perhaps not so good for a