It was a small enough community that I was the only advertiser (great!) but it also netted me about 50 visits, which was less than I'd hoped. I still found it very helpful in validating my ideas and getting feedback from a targeted audience.
I've been on the top of a sub reddit (music). Got 7k unique for the day. 500 uniques the following day.
Within 3 days, traffic was back to where it was. I've read this is typical behaviour for getting reddit, digged, HN, TechCrunched, Mashabled.
This concerns me because with the advent of social networsk, we're teaching the next generation to wait for content to be pushed to them. So it's unlikely they'll go/explore the main web site. The maxim seems to be:
If my social network isn't sharing it, it must not be good. So I'll wait for it to come to me. Why leave the walled garden.
And if they do leave the walled garden, and follow the link, it's only temporary. Users will simply go back back to their learned response of waiting in the wall garden for content.
This empowers social networks, since they get to serve advertising off your content, but it kills off the main revenue model (ads) for the rest of the web.
http://blog.trackjumper.com/2010/08/first-1000-visitors-part...
It got about 1800 visits. It's a quick overview of some useful (and not so useful) directory-type traffic drivers.