The gig economy and rise of crowd-based translation platforms brought the need for a new type of localization processes where one size does not easily fit all. In a crowd containing thousands of translators with diverse skills, backgrounds, productivity and availability, the management workflows must often respond to conflicting requirements, stemming from the different types of translators in the pool. This talk offers a comprehensive typology of crowd translators and the associated behavior patterns. It demonstrates how such a deep understanding of translator types and behavior patterns can be used to build efficient localization processes in a gig localization model.