In 1969, Stanley Milgram performed a neat trick.
He had 1 to 15 actors stare up at a window of a high office building while he counted how many others stopped to look.
He found the more actors he had, the more people would stop and glance up at the window.
This experiment was repeated in 2011 with modern tracking technology. The more actors who glanced up, the more passers by would follow.
However, when the crowds were thickest and people were faster moving (busier), this effect was significantly reduced.
Which matters because when we try to build communities, we often have nobody to begin with and target where most people are. Which are typically the most crowded places with the busiest people.
Instead you want to find the sparsely populated places where people pay attention to the messages they receive. That might not be e-mail, Twitter, LinkedIn, and Facebook. It might be showing up at events in person, introducing yourself, and finding places where people are open to receiving messages from strangers.