The ongoing research on developing metal additives to conventional explosives is a complex multifaceted problem that we can address through material engineering. Fuel-oxidizer composite particles produced by arrested reactive milling (ARM) can react much faster than other additives since the metal fuel and oxidizer are intimately mixed on the nanoscale. In the milling process the composite’s characteristics from size, shape, internal porosity to the degree of mixing of metal fuel and oxidizer, can be varied for a wide variety of compositions. We have developed a high-throughput method where we suspend a small quantity of particles from a prepared batch (~50 x 10-12 kg) in transparent polymer wells mounted on glass and introduce a powerful shock (up to 30 GPa) while we observe the particles from beneath using a microscope. The response of these particles is then followed using fast pyrometry and high-speed burst videography. We can see which particles ignited during shock transit and follow the growing reaction front as a function of time with nanosecond resolution. To find milling conditions that produce the best particles for shock sensitivity, we develop particle structure-shock response relationships. However, in the shock experiments, since the optical access to particles is limited to their projections all we see is the size and shape of the particles. By sectioning hundreds of composite particles and imaging internal features using electron microscopy, we identify nanostructure descriptors as a function of particle size. Since both the dynamic testing and nanostructure determination cannot be performed simultaneously, we have developed statistical methods to relate nanostructure to performance by routinely characterizing hundreds of particles in a day. This approach of using material characterization to interpret shock compression-driven reactivity will be showcased using Al-CuO compositions to isolate the effect of individual powder attributes.