We introduce a model of strategic experimentation on social networks where forwardlooking agents learn from their own and neighbors’ successes. In equilibrium, private discovery is followed by social diffusion. Social learning crowds out own experimentation, so total information decreases with...