The paper addresses a method for spreading messages in social networks through an initial acceleration by Spreading Groups. These groups start the spread which eventually reaches a larger portion of the network. The use of spreading groups creates a final flow which resembles the spread through the nodes with the highest level of influence (opinion leaders). While harnessing opinion leaders to spread messages is generally costly, the formation of spreading groups is merely a technical issue, and can be done by computerized bots. The paper presents an information flow model and inspects the model through a dataset of Nasdaq-related tweets.