Many tasks are too big for a single person or firm and can only be performed with the help of the crowd. Such tasks include data collection projects (e.g., Ushahidi maps), data processing projects (e.g., GalaxyZoo), and wide-area search tasks (e.g., the Red Balloon Challenge). The success of these tasks relies on having a large number people working on them. However, recruiting a large number of people is a difficult task itself, which often cannot be performed by a single person/firm… In order to succeed in crowdsourcing such tasks, recruitment of people should also be crowdsourced.
Consider an old example of crowdsourcing — search for a criminal. A bounty encourages to capture/report the wanted person. However, a bounty provides no inventives for referrals. A modern-day version of such search tasks do not have to rely on posting the photo of the target on every street pole. Instead, information about the task can spread through referrals on social networks, ideally leading fast to the people who have the right information. The Red Balloon Challenge demonstrated that such search can be very successful.
The tools are there: social networks make it easy to share information; keeping track of referrals is also not hard. The question is how to encourage people to share information given these tools. There will not be a golden bullet solution but rather a better understanding of what mechanisms can be used and which are more appropriate for a given type of project (e.g., fundraising, citizen science, natural disaster relief).
In the context of selling goods, referral programs have been studied in marketing for decades. Multi-level marketing deals with referral chains. There are companies whose business focusing specifically on sales referral management referralcandy and curebit.