We learned last year how Ditto Labs, the leading visual search company, reads photos on Twitter, Tumblr, and Instagram to discover authentic consumer insights and to find influencers. The math wizards at Ditto are now calculating how product adoption tends to cascade across social networks. For example do liquor brands tend to cluster among people connected online while car brands are less socially influenced? With social photo insights marketers can predict which people are most likely to purchase a product based on what is revealed in their friends' photos! Data scientists call this flocking behavior homophily. Understanding these network effects will forever change how marketers identify influencers and the fastest, most efficient path to purchase.