Abstract A common management intervention to support declining wild pollinators is ‘pollinator planting’. However, despite years of inclusion in conservation initiatives, global pollinator declines continue. Using the agent‐based model BEE‐STEWARD with two example species, Bombus terrestris (L. 1758) and B. pascuorum (Scopoli, 1763) (Hymenoptera: Apidae), we explore when during the year bumblebee resource demand is the highest and how that relates to seasonal changes in colony composition (numbers of eggs, larvae, pupae and adults). We then examine the impact of two‐week periods of resource scarcity across the year. Finally, we explore how enhancement with early spring‐blooming herbaceous species or trees changes colony survival and queen production. In the United Kingdom, there is a previously under‐appreciated ‘hungry gap’ for bumblebees in March–April, before the peak flight period, driven by the demands of larvae for protein and thermoregulation in the colony, rather than the number of adult bees. A 2‐week gap in forage availability during this period drives a 50%–87% drop in the production of daughter queens. Adding early‐blooming species in the model had significant, positive, long‐term effects on colony survival probability and daughter queen production. Pollinator‐planting initiatives in both national and international conservation schemes need to include plants that flower up to 1 month before the adults of target social pollinator species are apparent in the field, during the period that larvae dominate the colony. This approach is likely to increase colony survival and queen production, contributing towards halting and reversing global pollinator decline.