<![CDATA[The diffusion of innovation begins with the adoption of innovation to increase the productivity of corn farming so that a study of the innovation diffusion process is needed. The double track innovation is an innovation on how to grow corn which has been proven to be able to increase corn productivity by increasing the number of tillers in one hectare. This research was conducted from March to June 2023 with the objectives of 1) analyzing the main actors in spreading double track innovation among corn farmers participating in the TJPS program, 2) analyzing the relationship between socio-economic factors and the process of spreading double track innovation. The research location in Manusak Village, East Kupang District, Kupang Regency was chosen deliberately because it is the first and widest location for implementing the 'double track' innovation in the CSR program. The survey method was used in collecting data on 83 respondents. To answer the first objective, the data were analyzed using the analysis of the structure of the communication network with the sociogram approach, and the second objective was analyzed using the Spearman Rank Correlation Test (rs). The research results are: 1). As the main actor in the diffusion of double track innovation in corn cultivation in the TJPS program, the farmer with the first serial number, Mr. Foris, became a source of information for 24 farmers with a percentage of 29%. 2). Socio-economic factors that have a real relationship with the flow of innovation diffusion are the age of the farmer, the education of the farmer and the dependents of the farmer's family, on the other hand the area of land has no real relationship. The implication of this research is that in increasing the diffusion of innovations in farming communities it is hoped that they will utilize local-level change agents or key actors so that innovation spreads to all farmers, increase the role of millennials in the process of adoption and diffusion of innovations and apply effective agricultural extension models.]]>
