KNC 2010: FollowIndy tries to marry aggregation and geography
[EDITOR'S NOTE: We're highlighting a few of the entries in this year's Knight News Challenge, which just closed Tuesday night. Did you know of an entry worth looking at? Email Mac or leave a brief comment on this post. —Josh]
Former Indianapolis Star software developer Chris Vannoy brings something unusual to his News Challenge application: a fully functional site already built on nights and weekends.
FollowIndy is a hyperlocal aggregator, tapping into the vast web of information published through Twitter, Flickr, news sites, and blogs. Its value is in the limits of its geography: The site only targets news and information relevant to Indianapolis. “Unlike a lot of aggregators that sort of cast a wide net, the idea is to get a very small net that’s aiming for a specific area,” Vannoy said. “It’s about getting a full picture of what’s going on in Indianapolis and then providing some context around what people are talking about.”
Once the sources are pulled into FollowIndy, content is automatically tagged and aggregated, which makes it possible to aggregate all material related to arson, apartment complexes, or Peyton Manning.
Aggregating both professional and personal feeds means Vannoy has data to track how stories are pushed by each — if mainstream media is pushing a story that’s then being picked up by personal users, or vice versa. That’s similar to the Media Cloud project of our friends down the street here at Harvard. For instance, here is a visualization of mentions of the word “flu” in the sources FollowIndy tracks. Notice how mentions spike after The Indianapolis Star mentions is around 24 seconds in:
So if FollowIndy is already up and running, what does Vannoy need $100,000 of Knight money for? Vannoy wants to expand the network of sources it tracks (it doesn’t yet include local blogs, for instance), and it needs a variety of infrastructure improvements, such as better autotagging of content. But he believes FollowIndy is a model that can be duplicated in other markets, as he writes in his application:
By focusing on a single geographic area, you can go deep: pulling in blogs…alternative news weeklies, business journals, television stations and Twitter to try to grab every last speck of news. It’s that volume of data that suddenly makes things interesting and makes some things possible. Small-scale trends by geography, what a geographic area is linking to and the like become not just possible, but relatively easy.
I like the idea behind FollowIndy, and it looks really useful if you’re mainly after “official” news outlets. (I’m impressed with the auto-tagging.) But like a lot of aggregators without a human hand-feeding things into the hopper, it feels a little like a wheel without a hamster. (Also, er, what’s the business model?)
Speaking of hamsters. We’ve got a great geographically-based aggregator in Boston: Adam Gaffin, the guy behind Universal Hub. He highlights the best stuff from local blogs and tweets, opines on Globe/Herald/community-news coverage, maps crime, publishes the groans of frustrated T riders (and Sox fans!) in real time, follows the courts, and sometimes breaks news. It’s an aggregator with a fair amount of content, but it has a voice and a personality of its own–Adam’s sort of a metro columnist who writes 10-15 times a day.
I’m involved in a project to get something similar off the ground in a much more rural area, the Catskills in upstate New York. It’s called the Watershed Post. We’re very lucky to have Adam on board as our web developer, and we’re hoping (but not waiting!) for a little Knight News Challenge love as well. Slings and arrows most welcome:
http://generalapp.newschallenge.org/SNC/ViewItem.aspx?pguid=6aee8166-fb7c-4a2e-8581-fa6f6ff036dd&itemguid=c9e49e07-c18c-4597-a41b-fd7b5394e760
First, Mac, thanks for the nice writeup.
Lissa: Regarding your Knight idea, being a native West Virginian, with similar geography and lack of news coverage issues, I think you might have something there (even if I’m not the biggest Drupal fan in the world … to each their own).
On to a couple of your comments:
After running FollowIndy for nearly 10 months now on my own dime, I can definitely see some of the issues you bring up. Until very recently, I haven’t had a lot of server resources available to expand out to many non-traditional news sources, but as I sit and watch the Colts game, I’m starting to work on that now.
One other thing I’m keenly interested in trying is using the tagging mechanism not necessarily as a simple way to navigate around FollowIndy itself, but as a way to follow threads of news in the community through a related stories API.
Write a blog post, get a block in your sidebar with links to related, local content. Publish a news article, and get the same.
As for revenue models, I want to create something sustainable here, which to me means covering the server costs and maybe earning enough money for an employee or two. I have some ideas, but some of that is dependent on how the grant process turns out.
Most of the ideas are tied to the earlier-mentioned related stories API, however. Revenue could either be generated from a commercial licensing or membership fee, or (and this idea just occurred to me this evening) a kind of AdSense for content, where the system pays you for sending traffic out of your site and charges the publisher for links coming in, all on a bid basis (again, a new thought, still noodling with it).
Thanks very much for the feedback!
“Write a blog post, get a block in your sidebar with links to related, local content. Publish a news article, and get the same.”
That is very cool indeed. Cool enough to pay for? That’s the $60,000 question. *grin*
Good stuff, Chris–I’m amazed that your site is a one-man band, and it’s a bold move to build first and fund later. Thanks for your comments, as well.