Many newspaper journalists have been working to shed their print-centric mindset as their newsrooms move to fully integrate print and online news-gathering and publishing. Staff-produced and audience-contributed videos and slide shows have become common on U.S. and European newspaper websites, along with blogs, interactive graphics, searchable databases and Twitter streams, as newspapers seek to engage audiences/readers so they keep coming to, and spending more time on, their sites.
For the past couple of years, I've had the opportunity to witness the development of a new technology that is aimed at helping solve this challenge. It's called Apture. Two news organizations, washingtonpost.com and propublica.org, recently began using it some. Examples are here and here, and last week the company began making it available free to bloggers. (Disclosure: Although I have no financial stake in the company, I have been a volunteer adviser since the founders sought out members of the Knight Fellowship at Stanford in 2006, of which I was one.)
At its basic, Apture adds links to stories. Other than a small distinguishing icon, an Apture link looks like any other. But when a reader hovers on the icon/phrase, a small window opens with additional, editorially selected content. The reader can explore that material -- text, videos, photos, documents -- without ever leaving the story page.
Some tech bloggers reviewing Apture have compared it directly to some linking tools already in use. But there is a fundamental distinction: Apture deployment involves human editorial judgment. Co-founder Tristan Harris describes it as semi-automated enrichment of a web page that allows editors to add additional material to a story without manual embedding of URLs. (The software also allows editors to add video or other multimedia to a story.) Apture pulls from an array of free information sources, including Wikipedia, YouTube, Googe Maps, Hulu, Revver, Imeem and Flickr as well as a websites' own archives. PDFs of documents can be linked via Scribd, an especially useful feature for investigative stories.
Apture is launching just as U.S. newspapers are making steep costs cuts and the company faces a challenge in persuading those publishers that the editorial time needed to use Apture is worthwhile. Harris says because ads can be served within the link windows, Apture lets publishers increase the amount of advertising on their most popular pages. Apture operates on a subscription model that gives it a share of revenue from those ads. Harris said the firm also has had conversations with other news websites, including outside of the U.S.
The idea that became Apture began taking shape a couple of years ago because Harris, then a senior computer science major at Stanford, and some friends were frustrated with the online news experience. It felt, well, cumbersome. Flat. Harris, who ultimately dropped out of a graduate program to start Apture, and his colleagues are pretty passionate about the importance of journalism in fostering a vibrant society.
Recently, Harris and I discussed Apture in email (his answers after the screenshots).
What's your short description of Apture?
"It's a rich presentation environment for news, based on editors hand-selecting good stuff around the web and bringing it into this interactive, dynamic experience. We built a platform that lets publishers take their editorial values to the next level. We try to make finding good, relevant media one click away (so it's an efficient process), but still leave the editor in charge of selecting what's going to be included. That's the job of good news staffs anyway: to sift through the vast swaths of information and pick out the good stuff."
Why would four young computer science students want to get involved in journalism, anyway?
"First, we really aren't traditional computer scientists. We are not huge geeks. We were more interested in building applications that truly made a dent in the lives of everyday people who use the web. It sounds naive, but if you take a step back, the web is truly the most amazing medium we've ever had to tell stories. ... but in the year 2005, compared to what the web could have been, it felt broken. We had huge opportunities to deliver new ways to communicate, but the web was still so flat, so based on the past: We were basically publishing flat pieces of paper, connected by links to more flat pieces of paper. Really, why?"
As Harris and his friends set out to see if they could figure out a solution to that "flatness", they began by studying big online news sites.
"They were run by publishers, not technology specialists. That's when we realized there was an opportunity to innovate. We could put our heads together and use our technical backgrounds to conceive and build something innovative that these organizations could not, and that was a powerful idea to us. But we were very careful. It's too easy for computer scientists to fall into the trap of what they CAN do, versus what they SHOULD do. And they should always focus on what's good for the user. So, we met with many journalists, starting with the Knight Fellows at Stanford. We didn't know about news or media or the politics of the industry. We were 21. So we sat down with them and picked their brains and asked for feedback at every step. We knew that if we didn't do that, we'd get stuck thinking like computer scientists. And you see how they solve problems. You can see it reflected in the ways technical teams at these big news sites usually "contextualize the news" -- the features they think are important that don't take story-telling or editorial taste into account."
"They automatically highlight key terms on a page and link to content. When readers see it, they click the link, but instead of getting more information about the thing they clicked on (what the reader wants), they're taken to a page with the last 10 new articles that mentioned the clicked phrase. Computer scientists love that stuff because it's obvious. It's the first thing that comes to your mind, if you think in terms of of computation -- show me stuff related to this, search this word against this body of text. But what the reader wants is: Give me relevant information about what I just clicked, right now. They don't want search results. ...They don't want to be taken away from what they're reading. That's unintuitive. So that's exactly what we went after."
The point of doing this?
"The goal was to get people to care. To get people to invest more in a story, to bring more meaning and empathy into this box pointing light at your eyes. And then we said, you know that's not enough. We can make it even better. People want to know more about what they just clicked, not just a quick bio or a video or a photo. They want the video and the photo and the bio. So we let them use that linking/publishing system to discover and bring in several more items. And then we built a way for readers to explore all of it, without leaving the page."
How is Apture different?
"Everyone else in this space does it the automatic way. An algorithm finds random phrases it thinks are relevant, and an algorithm links those phrases to content it thinks is relevant. Algorithms don't know how to tell stories. This is why news organizations still exist. ... No one else has built a rich experience for viewing multiple pieces of third-party media like this without leaving the page. It's about hand-selecting webs of fascinating media that draw the reader into the story. Not just single items, but lots of interesting content, all accessible at zero cost."
For the website owner, what's involved in using Apture?
"You pop in one line of code (the Apture script) onto your website to enable Apture. The code appears on every web page on your site, but is invisible to the user. Visit any page on that site, hit a shortcut key and the Apture editor dashboard opens. You login with a username/password. Then you choose the phrases on the page you want to link up. The selection causes a window to open up that returns Wikipedia articles, images, videos from free sources like YouTube/GoogleVideo/Revver/Metacafe/Veoh, etc. -- and, if you're a publisher, material from your own text and multimedia archives. I click on any search result to preview it and if it's what I want, I link it to my page (to the selected phrase) with one click. That's three clicks total to take something on the page, select a search result and click link."
UPDATE 8 July 08: A video interview with Apture's Tristan Harris on Revver.