Unknown neutron-capture reaction rates remain a significant source of uncertainty in state-of-the-art $r$-process nucleosynthesis reaction network calculations. As the $r$-process involves highly neutron-rich nuclei for which direct ($n,\gamma$) cross-section measurements are virtually impossible, indirect methods are called for to constrain ($n,\gamma$) cross sections used as input for the $r$-process nuclear network. Here we discuss the newly developed beta-Oslo method, which is capable of providing experimental input for calculating ($n,\gamma$) rates of neutron-rich nuclei. The beta-Oslo method represents a first step towards constraining neutron-capture rates of importance to the $r$-process.