In high power impulse magnetron sputtering (HiPIMS) operation, there are basically two goals: a high ionized flux fraction of the sputtered target material and a high deposition rate. In this work, it is demonstrated that the former always comes at the cost of the latter. This makes a choice necessary, referred to as the HiPIMS compromise. It is here proposed that this compromise is most easily made by varying the discharge current amplitude, which opens up for optimization of additionally four external process parameters: the pulse length, the working gas pressure, the magnetic field strength, and the degree of magnetic unbalance to achieve the optimum combination of the ionized flux fraction and the deposition rate. As a figure of merit, useful for comparing different discharges, ( 1 - beta t ) is identified, which is the fraction of ionized sputtered material that escapes back-attraction toward the cathode target. It is shown that a discharge with a higher value of ( 1 - beta t ) always can be arranged to give better combinations of ionization and deposition rate than a discharge with a lower ( 1 - beta t ). Maximization of ( 1 - beta t ) is carried out empirically, based on data from two discharges with Ti targets in Ar working gas. These discharges were first modeled in order to convert measured plasma parameters to values of ( 1 - beta t ). The combined effects of varying the different process parameters were then analyzed using a process flow chart model. The effect of varying the degree of unbalance in the studied range was small. For the remaining three parameters, it is found that optimum is achieved by minimizing the magnetic field strength, minimizing the working gas pressure, and minimizing the pulse length as far as compatible with the requirement to ignite and maintain a stable discharge.