We present in a new way of analyzing direct quantum dynamics simulations based on a Mulliken-type population analysis. This provides a straightforward interpretation of the wavepacket in much the same way as semiclassical trajectories are usually analyzed The result can be seen as a coupled set of quantum trajectories. We apply this to the study of the photochemistry of a 12-atom model cyanine to explore possibilities for intelligent optimal control The work presented here builds on previous semiclassical dynamics simulations [Hunt, P. A.; Robb, M. A J Am Chem. Soc. 2005 127, 5720] Those calculations suggested that, by controlling the distribution of momentum components in the initial wavepacket, it should be possible to drive the system to a specific region of the conical intersection seam and ultimately control the product distribution. This was confirmed experimentally by optimal control methods [Dietzek, B; Bruggemann, B, Pascher, T; Yartsev, A.J Ant. Chem. Soc 2007 129, 13014] This paper aims to demonstrate this in a quantum dynamics context and give further insight into the conditions required for control Our results show that directly addressing the trans-cis torsional modes is not efficient. Instead, one needs to decrease the momentum in the skeletal deformation coordinates to prompt radiationless decay near the minimum conical intersection at large twist angles.