BusyTonight? Friday, February 09, 2007
I've mentioned that I use upcoming.org as well as meetup.com to keep an eye on what various chums are planning on getting up to in future as a prompt/tipoff so that I too can wander along to interesting/fun events. Of course, the trouble with this is having to rely upon them to be socially conscious/responsible enough to post their future plans for "my benefit".
BusyTonight is a service that is trying to overcome this problem as well as trying to deal with the fact that the market for event listings is hopelessly fragmented
Presently only operating in the US, it is a search engine for events. It is algorithmically driven via a Web crawl. The index of event listings is generated "automatically" - the software aims to identify pages with events and extract the event listings accurately.
Hence it seeks to eliminate the need for good natured folks to post their plans to a service and the same service! As such Busytonight is not dependent on having a critical mass of motivated users to type in event data - the content only needs to published on the Web.
They claim to be closing in on one million unique, future events with plans to expand to the rest of the English speaking world in Q1 of next year, and begin including non-English content in Q2.
They intend to generate revenues from a combination of pay-per click advertising and commissions on ticket sales. There's also the potential to sell the harvested data to a variety of other sites that offer tangential businesses.
Only founded in April 2005, there are only five people working full-time on BusyTonight including Joshua C Lerner (CEO), Richard Mintz (COO) and Matt Kangas (CTO).
Clearly there are a number of challenges here
- These are not recommendations from friends that you trust or can meet up with at the events. So the quality of the events is uncertain.
- Depending on how successfully the crawler works, you could easily end up with misclassifications of place, date or content that means you miss things or waste time looking at irrelevant stuff on listings it suggests.
- The sheer volume of stuff it finds will take some wading through. Upcoming lists for London are big enough. Whilst I could filter/search for specific events or topics, that overlooks the serendipitous events you stumble across which look interesting
- Timeliness of the crawling. If events can be posted on any site and at any time, will the spiders pass by sufficiently frequently to spot last minute entries on sites? Whilst other automated crawlers have the same issue, a human posting to a service overcomes this (of course, this can be one of the sites busytonight crawls over).
- Accuracy of tags could be frustrating. I tried to search for "comedy" in San francisco as a test. The results were odd as you can see from the image below, which included CityFlight - Black and African American History …
I very much like the concept though, representing a very useful vertical search - most of us go out and not just to eat/drink.
For future enhancements it would be great if they could add to it
- services like google's alert provides to notify me via email/rss that something matching my criteria has appeared.
- the ability for people to "ping" the service to notify it of events announcements, rather than await the crawlers
- the ability to "tag" their site as an event site in order that the crawlers visit more often