First panel: WASP-12’s RV dataset, folded with respect to the period of WASP-12 b (transits occur at t = 0). The best-fitting “WASP-12 b + tidal bulge” model is plotted in black. Second panel: tidal RV signal, obtained by subtracting the RV model with K_tide = 0. The 6 m s⁻¹ modulation due to the star’s tidal bulge is detected, illustrating the sensitivity of the dataset to low-amplitude signals. Third panel: RV residuals. All error bars show formal uncertainties, before inflation with jitter terms (Table 1). The residuals appear randomly distributed around zero, with underestimated uncertainties but no obvious trends.