Dense suspensions consisting of many solid particles suspended in a fluid are ubiquitous in nature and everyday life. Past studies have found remarkable success in modelling the bulk properties of these suspensions via a combination of stresses that capture effects such as particle contact and fluid mediated interactions. Experimentally, however, it remains difficult to resolve the observed flow behavior into contributions from these stress components, especially for more complex shear profiles that better represent practical processes. In the first part of this thesis, we introduce an experimental method that enables us to overcome this limitation for large amplitude oscillatory shear (LAOS), a shear protocol commonly used for material characterization. By systematically conducting a series of shear reversal and cessation experiments, we show that the total stress can be resolved into its constituent components \textit{at each point} within the shear cycle. Such a decomposition can enable more precise characterization of material properties and serves to more rigorously test predictions from theory and simulation. Taking inspiration from such contact-based models, the second part of the thesis introduces new approaches to externally tune the suspension bulk properties by modifying the force networks formed through particle contact. The first approach leverages on the ability of very dense suspensions to solidify when a sufficiently large stress is applied, a phenomenon termed shear jamming. We demonstrate that acoustic perturbation enables us to tune the suspension memory encoded in the contact networks prior to shear jamming. Due to the antagonistic nature of these contact networks, shear jammed states with dramatically different material properties can be obtained as a result of the tuning. The second approach leverages on the fast response time of the suspension force networks to acoustic perturbation to generate viscosity metamaterials. The instantaneous viscosity of these metamaterials rapidly oscillates with time, allowing access to states with markedly different flow properties which can be exploited using time-dependent shear protocols to generate unconventional material properties such as a negative average viscosity. The last chapter of this thesis presents preliminary results for using activity to tune the flow properties of a dense suspension of Quincke rotor in 3D. Imaging and rheology experiments reveal large, novel changes to the suspension bulk properties that have not been observed in less dense active suspensions. Collectively, the work presented in this thesis provides novel strategies for experimentally understanding and tuning the bulk properties of dense suspensions under shear.