<p>This article introduces a novel class of estimators aimed at estimating population variance in the presence of random non-response. It focuses on combining different estimators based on varying strategies, and this combination is optimized using an optimal constant value. While previous works have made significant contributions, the approach presented here is unique in its incorporation of ratio and regression estimators in a hybrid form, providing a new perspective on the estimation process. The key idea is that the hybrid form of these estimators yields a range of values for the constant being considered, enhancing the flexibility and effectiveness of the estimators. Notably, the estimators proposed in the article encompass those suggested by Singh and Joarder (Metrika 98:241–249, 1998), Ahmeda et al. (Int J Inf Manage Sci 16(2):73–82, 2005), and the ratio structure from Bhushan and Pandey (J Stat Comput Simul 91(18):3814–3827, 2021). The proposed class of estimators outperforms those introduced in these previous works. Furthermore, the article highlights the power of the optimal estimation procedure in providing greater efficiency when compared to existing methods, supported by simulation studies and numerical results that confirm the superiority of the proposed approach. Overall, the article presents a significant advancement in the field of statistical estimation by offering a more efficient and flexible way to estimate population variance in the context of random non-response.</p>