Abstract: A system and method for improving the performance of a viral marketing program develops a plurality of trials of a viral marketing program having a business objective with each trial having a plurality of attributes at least one of which differs from an attribute of the other of the plurality of trials, exposes a first plurality of users to a first trial and a second plurality of users to a second trial, collects data indicative of the users response to the trial to which they were exposed that is reflective of the likelihood of the trial obtaining the business objective, compares the data to identify a trial more likely to achieve the business objective than a non-identified trial and continues to utilize the identified trial in the viral marketing program while ceasing use of a non-identified trial.