Decay and dispersal of the tilted Bipolar Magnetic Regions (BMRs) on the solar surface are observed to produce the large-scale poloidal field, which acts as the seed for the toroidal field and, thus, the next sunspot cycle. However, various properties of BMR, namely, the tilt, time delay between successive emergences, location, and flux, all have irregular variations. Previous studies show that these variations can lead to changes in the polar field. In this study, we first demonstrate that our 3D kinematic dynamo model, STABLE, reproduces the robust feature of the surface flux transport (SFT) model, namely the variation of the generated dipole moment with the latitude of the BMR position. Using STABLE in both SFT and dynamo modes, we perform simulations by varying the individual properties of BMR and keeping their distributions the same in all the cycles as inspired by the observations. We find that randomness due to the distribution in either the time delay or the BMR latitude produces negligible variation in the polar field and the solar cycle. However, randomness due to BMR flux distribution produces substantial effects, while the scatter in the tilt around Joy’s law produces the largest variation. Our comparative analyses suggest that the scatter of BMR tilt around Joy’s law is the major cause of variation in the solar cycle. Furthermore, our simulations also show that the magnetic field-dependent time delay of BMR emergence produces more realistic features of the magnetic cycle, consistent with observation.