PurposeThe purpose of this paper is to develop new model formulation for reducing the workload in pre‐batching at a manufacturer of flavors and fragrances, by optimally assigning ingredients to different storage types, taking into account past usage of ingredients and several restrictions about volumes and number of storage units.Design/methodology/approachTwo models were developed, using mathematical programming, accommodating either fixed or variable shelf settings in vertical carousel storage, along with other varying storage types. The models were validated, yielding a sizeable reduction in workload, and run under varying scenarios of storage additions to reduce workload even further.FindingsFor different storage addition scenarios, application of the new model formulation yielded a reduction of 22 percent in workload. In addition, aside from space savings, approximately 40 percent of refill order total waiting time was reduced due to improved allocation.Research limitations/implicationsThe authors' analysis was limited by considering direct cost savings only. Yet, reduced workloads in pre‐batching may also yield indirect cost savings, such as reduced quality costs, inventory costs, and investment savings by not having to extend the pre‐batching area. Such considerations may be addressed in future research, provided data on indirect savings are available.Practical implicationsWhile intuitive storage allocation overloads the pre‐batching department and increases space utilization, suggesting an incorrect perception that more personnel and space are needed, the proposed approach provides a better alternative through optimized allocations.Originality/valueThe present paper adds to the literature on carousel storage location by explicitly addressing the storage sizing issue, as well as workload balancing.