Comparison of the variance A(K) of the lensing profile reconstructed for SPT-SZ and Planck binned in K = |k′ − k| (or equivalently L = |l′ − l|). The figure displays the ratios between the combined A_SPT+Planck(K) and the single datasets A_i(K), and the sum of both ratios. It shows that the Planck dataset provides the lowest error on the lensing potential at large scales (K < 50.9 or L < 1830), while the SPT-SZ dataset performs better at small scales (K > 145.0 or L > 5208). The sum is the product between the combined A_SPT+Planck term and the inverse variance sum l/A_SPT + 1/A_Planck- It is below unity for K > 18.9 or L > 680 demonstrating the significant gain of the combination with respect to individual measurements.