Evidence-based program implementation in criminal justice and community settings is complex and rarely encounters an ideal set of conditions. Implementation failure may be more common than success, and the field has only started to seriously consider how to improve dissemination and implementation. Research-practitioner partnerships are a potential mechanism for achieving implementation success. Using an implementation science framework, we examine the impact of researcher involvement and activity on implementation outcomes in a sample of 49 community-based multi-agency initiatives to improve public safety. Our findings suggest that projects focused more heavily on data and analysis from the outset achieved better implementation outcomes across multiple domains, including fidelity, penetration and dosage, complexity, execution, and sustainability. Similar improvements in outcomes were not associated with more robust data collection, access, or analytic capacity.