New research, initially shared on the ESS Open Archive, investigates how the width of updrafts and model resolution modulate the vertical growth of convective clouds. The study highlights that the initiation and growth of deep convection are highly sensitive to mid-level relative humidity, low-level instability, and the width of the cloud-base updraft. Importantly, narrower convective cells are found to be more susceptible to mid-level relative humidity, while wider cells respond predominantly to atmospheric instability. The study utilized data from the LES ARM Symbiotic Simulation and Observation (LASSO) database and the CACTI field campaign in central Argentina, employing simulations across various horizontal grid spacings (2.5 km, 500 m, and 100 m). A key finding indicates that kilometer-scale models, despite their widespread use, may not fully resolve updraft dynamics, leading to updrafts that are 2-3 times wider than those observed in higher-resolution Large-Eddy Simulations (LES). This under-resolution affects the accuracy of predicting cloud depth, particularly the sensitivity of narrow cells to dry mid-tropospheric air entrainment. As model resolution increases, the sensitivity of narrow cells to mid-level relative humidity intensifies, underscoring the critical role of moisture entrainment from the surrounding drier air in limiting cloud depth. These results suggest that current kilometer-scale simulations may not accurately capture the nuanced importance of environmental factors influencing vertical cloud growth. The research, therefore, emphasizes the urgent need for enhanced treatments of sub-grid variability within climate models to more accurately represent key convective processes. This advancement is vital for improving the fidelity of weather prediction and climate models globally, with direct implications for understanding and forecasting atmospheric phenomena in regions like India, which are significantly impacted by convective weather systems.