A few months ago, the first mobile expression of a news concierge materialized in the form of NOD – News On Demand. NOD is an iPhone app that tailors news to the user’s available time and attention by providing three important stories daily, in two sizes, and allowing to find them from a catch up calendar.
It’s about separating signal from noise to a quite extreme extent (wtf, 3 news per day only??!) and I’ve received many questions – at presentations, in meet-ups and conferences – about the selection process of these 3 stories. Why only 3? And why these and not others? But the question that people are really asking me is: define “important”.
So, I tried to explain what I talk about when I talk about “important” news:
Interesting: is the story telling you something insightful, necessary or fascinating about the world we live in today?
Matter: will the story also matter tomorrow? Is this the beginning of something that you may hear from in the next few days (think: a trial start, a cop killing leading to protests)? Is this the conclusion of an event that will make history?
Pressing: is the event very likely to be in conversations with friends or at work today, or in a few days maybe? Then you may want to be caught up on it.
Original: is there an odd and unique angle to an apparently “light” news story? Taking the opportunity of George Clooney’s wedding to learn about the great work of human rights lawyer Amal Alamuddin makes a celebrity story more important.
Relevant: our users send us a lot of signals through their behavior. We try to act on them : for example, our users give a lot of “nods” to stories on climate change, so when Chile is flooded or a typhoon of uninque size hits the Philippines, it seems more relevant than a man past-retiring age saying he’ll retire.
Trending: is the story making headlines in a wide array of outlets, including major publications, aggregators like Google News and social media like Twitter and Facebook? Will our users wonder why everyone’s making a fuss about this?
Accurate: too often, stories are broken and then turn out to be falsely stated or completely trumped up. Fake kidnappings, murder charges that are dropped, people tested for Ebola who make headlines for 2 days cause they’re found to be healthy after all… So I’d rather wait 24 hours. For breaking news and breaking corrections, there are other outlets.
New: is the story about something we didn’t know yet? A lot of “NEWs” is merely an update in a longer story arch or a quote ping-pong (think Hillary’s email-gate, negotiation processes and plane crash theories).
Timely: obviously, all the stories we curate need to be fresh. If we put out an edition every 24 hours, the events have to have been reported on in the same time-frame.
There will be instances where at least one of these criteria may apply to something that doesn’t show up in the daily NOD edition. This is not enough, though: stories need to score, if not all, as many of those criteria as possible. Sometimes, I push the boundaries a bit: I do this on purpose to trigger user feedback and make NOD better (please write when you disagree!)
Of course, this is anything but perfect. Being human, we shouldn’t aim for perfection anyway, but for continuous improvement. So, I want to know: how do you define important news?
As a 2013 Knight fellow, I spent the past 10 month working on a way to make news consumption more time efficient. I placed myself in the vast “slow news movement” (hence this site), as I focused my work on allowing people to catch up at convenient times, rather than helping them spend less time reading the news (say hi to Cir.ca, Summly!)
The result of my explorations became a web-based prototype, The Newstapes. And a memorable talk in the Bing Auditorium at Stanford, on July 12, where I introduce the concept behind all this. Meet the News Concierge!
I am now working on taking this project to the next step. Any thoughts, interest, ideas are most welcome :)
How can we catch up on news in a time efficient matter? The question is at the heart of my project as a Knight fellow, at Stanford. In order to try to solve this equation, I launched the Newstapes experiment about two month ago.
What is it ? The Newstapes is a Tumblr blog where I try to help people catch up or understand a story with 1 to 3 links to useful articles, videos or infographics. Depending on how much time a person has, he/she can choose to read only the first piece, or digg deeper with one or two other pieces. The user I’m aiming for is a casual news reader, not a hardcore news junkie, with no prior knowledge of those topics.
How does it work ? Everyday, I pick a story that is making headlines – lately, the bankruptcy of Cyprus, JPMorgan’s Senate hearings, the 2013 Oscars or the elections in Kenya. It has to be on the front page of major news outlets, potentially unfolding over several days and complex enough to need a brief roundup if you haven’t been living in sync with the 24 hour news cycle.
Then, I read and search for the best summary of the situation (story level 1), a good explainer or detailed version of the story (story level 2), and a third piece that will provide more context or shed the light on an intriguing perspective on this given story. The three links are displayed in this order, and I display the time it takes to read or watch each piece. Occasionally, I’ll throw in a great long piece that is a great explainer of an older topic by itself – for serenedipity’s sake.
What did I learn so far ?
- Displaying time : it’s a tricky subject. People tell me they want to know beforehand that they’ll find out “everything they need to know on topic X in 5 minutes”. Others, tell me they’re put off by this, saying “even 5 minutes is way more than I want to spend on this topic”. I tried to solve this problem by picking very short stories of level 1 (30 seconds to 2 minutes) and to display the time for the following stories as incremental (“+3 minutes”).
- Content format : it’s a tricky subject. When I started, a lot of complaints I heard were : “you only pick text, what about video, radio, infographics ?!”. Once I tried to mix things up a little bit more, people complained again : “I don’t like videos, they force me to watch this thing without knowing where it’ll take me”. Take away : there’s no way of pleasing everyone. I’m going with the best explainers and summaries for now. Whatever the format.
- Catch up worthy stories : one of the reasons I’m so interested in making news consumption more time efficient is that I hope to, eventually, create more time in people’s schedules for the good rewarding long reads that make readers and journalists happy. But the Newstapes that attracted the most readers to date were : 1. the 2013 Oscars ; 2. the death of Hugo Chavez ; 3. Queen Elizabeth’s visit to the hospital. And I’ll have to keep this in mind in the future.
Why the Newstapes ? People born before 1990 might remember the mixtapes, these great collections of songs a friend or a DJ would put together on a cassette in order to share it with other people. It was a good way of telling stories musically and discovering new sounds, recommended and curated by some DJ you liked. And I’m trying to do the same thing with the news, as I’m arranging little collections of news stories for friends and other random curious people. In both cases, it’s about discovery and social interaction.
Where do I go from here? With this experiment I’m finding out how people react to quantified news. It also helps me define what is worth catching up on. Eventually, it may help me redefine the concept of “daily news” itself : instead of going for the all you can eat buffet of stories, what would be the *one* story you would want/need to explore, on any given day?
A few days before Christmas, my co-fellow Andrew Donohue teased me, asking on Facebook “with everything that’s been going on, I could really use MCB’s innovation when I return. Do you think you can have it finished by Jan. 3?“.
I decided to take him up on that challenge and work out manually a prototype of the service I hope to offer to people in his situation. On January 4th, I shared with him (and my social media friends) a Google Doc, inspired by a friend’s mention of Jason Calacanis debut with Launch ticker. It listed one major event per day since his Winter break had started and a link to catch-up on the story.
- Read paper copy of USA Today, on January 2nd. Goal : get an overview of the topics in the news, in order to identify things that are unfolding and that one might need context for, if they missed the story that set events into motion. Stories on the fiscal cliff, Hillary Clinton’s health and North Korea stood out after an hour of reading.
- Check out Twitter and Google News to see what’s trending. Twitter was mostly about sports, following the Rose Bowl and Sugar Bowl. On Google News I confirmed that Hillary Clinton was a widely covered, as were stories on Hugo Chavez in Venezuela and Syria.
- Check out “pictures of the week” features of several online news outlets. Objective: find out about events that happened during the elapsed time but aren’t in the headlines anymore. This is how I decided to add different moments of the days after the Sandy Hook mass shooting, such as Obama’s speech and the NRA’s news conference. I also realized with these images that protests in Egypt and the adoption of their news constitution had been a major event during the past month.
- Rewind major events of each week with the perfect “10 things you need to know today” feature of The Week. Goal : find what I missed. Here I narrowed my stories on Syria down to the talks between Brahimi and Assad, discovered the terrible snow storms in the North east of the US, was reminded of the resignation of the Italian Prime Minister and the death of a first Gulf war commander.
- Search for useful link to catch-up on the story. Depending on the topic, it was either the article of that day, a timeline or a dynamic infographic that proved to be most useful.
Here are the things I learned with this “Andy” experiment, both through my own experience and users feedback:
- when narrowing the media offering down to ONE article per day, the user needs to know precisely the criteria used to do the selection
- Google Docs is bad for gathering data on the number of visits
- the catch up digest has to mix context for unfolding stories as well as brief reports of one shot events that still matter in the future
- it would make sense to be more explicit about what the user is going to find in the catch up link
- calendar-based catching up works for general news, but there’s demand for topic-centered catching up
- major news outlets are still useful to understand the global agenda.
Winter break has started at Stanford. For me, it began with a week of vacation in Hawaii. I decided to put this time off to good use and do some work on my Knight project. Or, actually, to have other people do some work for me while I was going to enjoy some Pina Coladas on the beach.
Dreading the perspective of diving trough piles of unread emails and missing out on great stories, to catch-up on a week of news – as I was prepared to follow information only casually – I put a message out on Facebook. It read :
“Friends! I want to run an experiment and need your help: I won’t read any news this week, as I’m travelling. Please share in the comments below your 1-3 favorite articles or stories of the week, so I can catch-up on the greatest things next week-end. This curation will also help me push forward my Stanford project – so many many thanks in advance. And make it GOOD :) “
My hope was to gather about 30 to 50 links from a small participating part of my 400+ friends, some of which would be redundant and hence give me a sense of importance in volume, some of which would be very random and hence make sure I got a random and surprising harvest.
What I received was indeed surprising. But not quite in the sense I would have expected. Here’s what I gathered thanks to four devoted/playful friends:
Twelve facts about guns and mass shootings in the United States, Washington Post (4’48”)
The worst CEOs of 2012, Bloomberg Businessweek (2’07”)
110 Predictions for the next 110 years, Popular Mechanics (6’55”)
Coursera to offer social dance, The Stanford Flipside (35″)
Time to stop the massacres, Jeff Sachs (9’40”)
Looking for America, New York Times (2’13”)
Politicizing tragedy and the Aurora Theater shooting, Mother Jones (1’26”)
A response to Jack Stuef’s Buzzfeed article about me, The Oatmeal (10’08”)
The secrets of the Internet’s most beloved viral marketer, Buzzfeed (3’22”)
Poll: which party is Santa in?, Politico (11″)
Was it the Newtown mass shooting that slightly threw off the experiment? I didn’t feel like I could jump right into the many op-eds that were shared and started by looking for a timeline of events (2’41”) that I had just discovered through Facebook hear-say and emotional reactions.
41 minutes and 31 seconds later, I knew a bunch of stats on guns and mass shootings in the United States, I had learned that some law passed in Australia in 1996 contributed to a zero-mass shooting environment down under and I had heard of some clash between a comic artist and Buzzfeed. I had learned that the CEO of Zynga had beaten those of Facebook and Groupon to the list of worst CEOs in America and realized that it didn’t matter much since in fifty years, we’ll all be needing help of robots to programm our own brains….
It may say a lot about my caracter but somehow, I wasn’t sure I could consider myself fully informed on the major events and good surprises of the last week with this bouquet of news curated by my friends. Especially when some dated back to… last July. So I went off to scan my inbox for good stuff to read in the various newsletters I subscribe to (it to about 2 hours to open dozens of tabs of things to read).
Then, I was intrigued to learn that my instinct was to go for local news first (16″) while on a day to day basis, it seems like candy you munch on after dinner. I also enjoyed the “I am Adam Lanza’s Mother” (5’04”) a friend had recommended me by email. The piece felt more interesting that all the op-eds I had swallowed earlier. But I couldn’t have understood it without some context a.k.a the initial timeline of events. Also, I could not resist indulging a not that important yet meaningful piece to me, on GigaOm : It’s not Twitter, it’s just the way the news works now (4’08”).
Finally, there are still many of favorites waiting on my Twitter feed to be read and understood.
Still, here are a couple of take-aways from this experiment. Things I’ll keep in mind while further researching ways to slow down the news:
- Your friends can be the best and the worst of curators.
- Relying only on them (social curation) may feed your fear of missing out on good stuff.
- Local news rise in importance when offline for a while.
- Op-eds without context (=headlines, timelines) fall into the catefory of “nice things to have”, not “vital stuff”.
- Personal interest > funny story.
- Funneling the reader from short contextualizing stories/headlines might be a good idea.