This is a little old, but I think it bears re-examining.
Viral marketing campaigns, like real viruses have what is called a reproduction or transmission rate. That is what is the average number of times a person will pass on a specific message. If the average person who sees your viral creative shows it to one other person, your transmission rate is 1. If that rate is lower than 1, the network of people who’ve seen your creative won’t grow, it will atrophy. If that number is over one the number of people exposed to your ad will grow. So clearly it is very important to persuade visitors to your message to share it with their friends, but most viral campaigns do not have a transmission rate greater than one. You can target prolific social media users (who are likely to have very high transmission rates) with your viral seeding tactics, but achieving very high rates is difficult.
On the other hand if you increase the “seed” size (that is the number of people who see the creative at launch) you can ensure that even if the transmission rate is low a large number of people will still see the ad.
This is called a Big Seed. Duncan J. Watts, Jonah Peretti, and Michael Frumin wrote a paper on it (the original PDF seems to have disappeared, but Google still has a cache) and Watts and Peretti wrote a summary for the Harvard Business Review. Interesting reads and a topic any serious viral marketer must understand. Go read.
In a more practical context, social sites like Digg, Reddit or Delicious are perfect launching pads for Big Seed efforts. And of course not only do you get your message in front of tens of thousands of seed viewers, but the makeup of these audiences typically skews very heavily towards social media connoisseur. This means that not only do you get the benefit of a big seed, but the first wave of “infections” will have a higher-than-average transmission rate.