groups. This exercise may be slightly more complex, but could be achieved more systematically if (say) expected productivity rates for each product- group, by job category, were available. Hence, assuming no required change in the make-up of the outputs, if we wish to increase our overall staffing levels to 1,000, then we would need to hire 100 employees in addition to the above 120 In this case, the allocation of the extra 100 employees across the three job categories would be achieved on approximations of the existing proportions across the three job groups. This would give us an allocation of 11 (ie, 100/900 (or approximately 11 per cent) of the 100) to the management group; 22 i.e 200/900 (or approximately 22 per cent) of the 100) to the engineering group and 67 (i.e. 600/900 (or approximately 67 per cent) to the group of technicians In reality, the decisions on these allocations require more minute considerations and are based on productivity rates. Further, these allocations should take into consideration potential (and anticipated) changes in the volumes as well as the make-up of the outputs (or services), in the case of a service organization. As mentioned above, these could be further refined and fine-tuned by taking into consideration other internal forces (such as expected redundancy rates in each position in different locations); various market forces affecting different products/services and even the power-distribution across the different job categories (i.e. intra-organizational dynamics). This clearly requires consideration of other matrix-based coefficients representing different constraints.