A couple ideas for improving the news experience for readers

This past week of thinking through improving the news experience was fairly productive.

Last week I set the goal of creating a virtuous cycle between reader insights and the larger news cycle. In this post, I’m going to delve deeper into this high level and embody it in what I’ll tentatively call “Evolution”. In communication with Michael Bernstein, I also came up with a fresh new idea that I will also describe here. I had wanted to create some mockups to express these ideas, but alas that will have to wait until later this week! Apologies for the length of this post, but this is really useful for me to articulate here.

Evolution

Motivation. Our news experience is still quite primitive, like medieval medicine. Stories are generally organized independently, regardless of the issue they address. But even if they are organized by issue, they tend to be just a list. How can we learn about how the details of a story has developed, as well as how public sentiment has evolved, with such a fragmented presentation of what are often multi-story issues? But its not just stories that are fragmented. We are currently missing an opportunity for inspiring more reflective thought. Current commenting environments are usually a no-mans-land filled by hyperpartisan trolls, driving away those who may want to have more constructive interactions. Perhaps we can inspire some of those constructive people who have been marginalized to spend their time thinking through the news story if there is demonstrated connection to the higher level news cycle for an issue.

What? Evolution enables an interactive news experience around an issue rather than just a story. Consider the current debt limit issue in the states. Over the past months, many stories and editorials have been written on the issue as it has developed. Evolution presents a view of those stories, but adds functionality that helps show how reader opinions and insights have evolved through the history of the issue. Evolution uses an interactive timeline to convey stories and summaries of reader opinions and insights. These summaries derive from the comment sections of those respective stories. If relevant, subsequent stories by the news organization can be explicitly linked to and “inspired” by reader reflections, helping to show the virtuous cycle in action. Select leaderboard widgets can also be designed that recognize commenters who have added influential comments that inspired the followup articles (or who have helping create the summaries). This can help encourage people that making constructive comments can pay off. Other custom news experience items can be integrated as well at this level. For example, the NYTimes budget game Amanda Cox discussed is a natural fit alongside the story/comment summary timeline for the recent budget discussions.

Where do we start? The magic in Evolution will be in how those comment-board summaries are created and what exactly they try to summarize (two possibilities: clusters of opinions in the form of position statements, or specific future investigatory paths that will help clear up reader confusion). The first guiding principle is that the burden to create these summaries should fall on the readers themselves. It is not feasible for overburdened editors in the news room to be the responsible ones here. Build a system and community where readers can construct succinct high quality summaries that can be easily digested by and responded to by those in the news room. Wikipedia and related projects show that this general concept is feasible, even if Evolution does not end up looking like a wiki.

Reflect is one place to start for creating summaries. It shows one possible first step in identifying value in individual comments and moving toward a synthesis of them. Near the end of the Wikimedia Strategic Planning Initiative of 2009-2010, moderators used Reflect internally to summarize the hundreds of comments they were getting input from and they found the interface to be quite useful for moving from individual comments to full discussion summaries (see some of the comments of the head facilitator).

The next step after Reflect is to figure out how to crowdsource that step from comment-level summaries to discussion-level summaries. If we can do that, we can enable Evolution. One aspect that I’m contemplating is whether we can incentivise people to help summarize discussions by introducing an artificial scarcity of voice: If you want to comment, you need to earn your comment by doing some high-quality work. Imagine that every user gets a certain number of characters that they can use when commenting. So I might have 500 characters initially that I can use up commenting on a story, or split them for two shorter comments on a couple stories. However, I could earn more voice by summarizing other people’s comments, or by being recognized as contributing an insightful comment. This is somewhat based on Slashdot’s karma system, but simpler. Aside from getting useful summarization labor performed, people might be more concientious about how they spend their voice.

Strangeface

Michael Bernstein, who submitted Penpal News in the moznewslab, contacted me to brainstorm Penpal News and we finally got to chat yesterday. Penpal News is a really interesting idea – create a service that matches two kids from around the world to discuss a news item, with special application in to the classroom. Simple but powerful. I’m disappointed that Michael didn’t get selected to participate in this round because I think this is a completely fresh idea that opens up some possibilities we haven’t seen before. And truly goes “beyond comment threads”.

Anyway, as we were talking, I realized that at the core of Michael’s idea is a really compelling generalization: what if we created a service (lets codename it “Strangeface”) that matches pairs of people to videochat about a particular news item or issue or whatever. You come to a story about the debt limit, hit a button and are matched with someone else who has also indicated they want to chat.

Think Chatroulette, but with a reputation system behind it to cut down on unpleasant encounters, perhaps based on Couchsurfing.

Why would we want to do this? Simple: we’ve been trying to have effective large discussions for a long time now, but they don’t scale very well. Perhaps this is the wrong approach to take for facilitating discussion around news items. Maybe people would rather just have deeper synchronous interaction about the issue with one other person with whom they may or may not agree.

There are many bells and wistles that might improve this kind of service. For example, the match making algorithm could try to find two people who have some common ground, but differ on others. Scheduling will also be an issue – you might specify when you are interested in chatting with someone if no one is currently available or you would rather chat later. There would have to be some kind of rating of the quality of the discussion partner afterwards in order to learn who are good faith participants in the system.

I also just remembered that this idea has been implemented to some extent with Formidable Opponent, put together by some University of Washington undergrads.

(apologies for this mockup — first time I tried using Omnigraffle on the iPad)