Gulia and Wiemer (2019; hereafter, GW2019) proposed a near‐real‐time monitoring system to discriminate between foreshocks and aftershocks. Our analysis (Dascher‐Cousineau et al., 2020; hereinater, DC2020) tested the sensitivity of the proposed Foreshock Traffic‐Light System output to parameter choices left to expert judgment for the 2019 Ridgecrest 7.1 and 2020 Puerto Rico 6.4 earthquake sequences. In the accompanying comment, Gulia and Wiemer (2021) suggest that at least six different methodological deviations lead to different pseudoprospective warning levels, particularly for the Ridgecrest aftershock sequence which they had separately evaluated. Here, we show that for four of the six claimed deviations, we conformed to the criteria outlined in GW2019. Two true deviations from the defined procedure are clarified and justified here. We conclude as we did originally, by emphasizing the influence of expert judgment on the outcome in the analysis.