EnVision was selected as ESA’s 5th M-class mission, targeting a launch in the early 2030s. The mission is a partnership between ESA and NASA, where NASA provides the Synthetic Aperture Radar payload. The scientific objective of EnVision is to provide a holistic view of the planet from its inner core to its upper atmosphere. The mission phase B1 started in December 2021 to complete trade-offs, consolidate requirements, interfaces and system specifications. Phase B1 will be concluded with the Mission Adoption Review planned in fall 2023, followed by Mission Adoption in 2024. To meet its science objectives, the EnVision mission needs to return a significant volume of science data to Earth, with a large distance-to-Earth dynamic range (from 0.3 to 1.7 AU), from a low Venus polar orbit, in the hot Venus environment (exacerbated by the operation of highly dissipative units), while operating three spectrometers in an almost cryogenic level environment. This needs to be achieved within constraints on the spacecraft mass as well as Agency programmatic boundaries. Achieving the science objectives under these multiple constraints without oversizing the spacecraft calls for a careful planning of science operations, making the science planning strategy a critical driver in the design of the whole mission, against which the spacecraft and ground segment are then sized. The payload reference operations scenario simulation demonstrates that all identified surface targets can be imaged with VenSAR, with a performance fully compliant with the science requirements. The first two cycles allow imaging once 80% of the identified Regions of Interest (RoIs) at 30 m resolution. The following two cycles are mostly devoted to 2nd observations of these areas for stereo-topography mapping and the two last cycles to 3rd observations of the “activity” type. Dual polarization and high resolution SAR observations can be performed at any longitude at least once across the 6 cycles. Our strategy is to obtain the widest range of data types that enables us to put the highest resolution datasets into regional and global context. Similarly, understanding atmospheric processes requires a combination of global-scale mapping with targeted observations resolving smaller-scale processes.