Designing an active acoustic enhancement system (AAES) can be a challenging task due to the multichannel coupling of an active system with a passive room. To avoid manual tuning, algorithmic optimisation of the digital signal processing (DSP) unit is particularly valuable. State-of-the-art AAES optimisation operates on the system open loop, i.e., the measured transfer paths of the system microphones, DSP and loudspeakers. This can be suitable for equalisation (EQ) of the feedback loop itself, but it does not account for how the system is perceived in the audience area. This study proposes a closed-loop optimisation pipeline for AAESs which allows the propagation into, within, and out of the electroacoustic system to be accounted for. Two use cases are presented to demonstrate the benefits of optimising an AAES based on its closed loop. It is shown that EQ optimisation offers a nominal improvement over the open-loop method for colouration reduction. Additionally, the closed-loop method can be applied to matching the late reverberation of a target room in the audience, which was not possible with previous work.