Upcoming events in Keene, NH

In the fall of 2003, the Many-to-Many blog announced Andy Baio's upcoming.org, a collaborative event calendar. Ray Ozzie noticed, and asked a bunch of questions about calendar formats and mechanisms for syndicating and aggregating events. I weighed in with some examples of the XHTML microcontent techniques that I was (and still am) exploring. But eighteen months later it's clear that we were just philosophers debating how many angels can dance on the head of a pin. None of our arcane technical discussions really mattered because upcoming.org hasn't gone viral the way del.icio.us did. This isn't just my own observation. Recently someone in Edinburgh asked: "Why is upcoming.org not more popular?"

It surprises me that it isn't, because upcoming.org does lots of things right. Anyone can start a new metro, add venues, and post events. You can say you're planning to attend an event, and you can see who else is thinking about attending, which is a powerful way to build buzz and help people decide whether to attend. And of course you can subscribe to your upcoming events as an RSS feed, or export them in iCal format.

Now, mind you, plenty of people have signed up for upcoming.org. In Massachusetts, for example, the Boston metro has 319 users. But when I drilled down, I found that seven of the first ten had no scheduled events. This mirrors my own experience: I've had a hard time getting to critical mass with upcoming.org in my own town, with my own friends and acquaintances.

And yet there's a pressing need for a solution like upcoming.org. In my town, for example, there's no authoritative online source of information about events. The Keene Sentinel has a calendar, for which I've even synthesized my own RSS feed. And Keene State College has a calendar which, when I get around to it, I'll splice into my local events feed. Ditto the high school. But add all these together and you're still missing all kinds of stuff.

Yesterday, on a walking tour of downtown, I took photos of event sheets posted on walls and shop windows. If you compare the resulting photo gallery1 to the various online calendars you'll find some overlap, but not much. The vibrant diversity that you see in those photos, and that you see written on the walls and shop windows of every town, has yet to find a way to express itself online.

What does it mean for upcoming.org to "go viral"? In my town of 23,000 I think it means the service has tens of active users and hundreds of casual users, has most of the events on the various official online calendars, and has most of the grassroots stuff too.

How would upcoming.org, or an equivalent service, get there? There's never a surefire recipe, but here are some things that might help.

Folksonomy. Upcoming.org has a fixed schema:

That's a reasonable foundation, but a complementary tagging scheme a la del.icio.us would give people a way to build on it, and to federate their event data with other tag-aware services.

APIs. Currently there aren't any. In this exchange between Erik Benson and Andy Baio, Erik proposes an API for pumping events into the system and Andy responds that the requirement for manual data entry is a useful crap filter. I can see his point, but I think it's a bad tradeoff. Coupled with the lack of an API to pull venues and events out of the system, which creates data lock-in, I'd have a hard time evangelizing upcoming.org to local organizations. With import and export APIs, though, I'd show the newspaper and the schools how they could federate with the service, reap its collaborative benefits, and (if necessary) maintain their own views of the raw event stream, filtered by tags or by trusted users.

Calendaring on the web has been simmering for what seems like forever. A few more ingredients, another stir of the pot, and upcoming.org might yet show the way.

Update: Looks like the pot may get stirred sooner rather than later. Excellent!

1 Made with JAlbum, by the way, a really handy tool for building a statically-served photo gallery that can run as a slideshow.

Former URL: http://weblog.infoworld.com/udell/2005/03/21.html#a1198