The potential of GNSS-R techniques to estimate land surface parameters such as Soil Moisture (SM) is experimentally studied using 2014-2017 global data from the UK TechDemoSat-1 (TDS-1) mission. The approach is based on the analysis of the sensitivity to Soil Moisture of different observables extracted from the Delay Doppler Maps (DDM) computed by the SGR-ReSI instrument using the L1 (1575.42 MHz) left-hand circularly-polarized (LHCP) reflected signals emitted by navigation satellites. After quality-filtering the data (spacecraft not `in eclipse' and the direct signal not present in the DDM), topography filtering the data, correction for background noise, antenna pattern gain, different distance propagation factors, and vegetation attenuation, the sensitivity of different GNSS-R observables to SM and its dependence with incidence angle is analyzed.