First Allocation Algorithm - Part II: Optimization

Once a robust prediction of demand is available, the next step is to optimize the allocation of stock to the different stores. 

The optimization process is intended to maximize the probability of sale while meeting required restrictions and minimizing the “cost” associated with the number of products to send (e.g. If for a given product sending 1 unit to 1 store has the same sale probability as sending 2 to 2 different stores the first option is preferred).

The optimization engine allocates units of each “Product-Size (SKU)” in order of value in sending them to the stores. This is done by taking into account the probability of sale, logistics costs and value of keeping it in the warehouse while applying different types of local restrictions.

Restrictions: The optimization process takes into account specific inputs that are important for the business. Read more about Restrictions

Prepacks: For a particular product, the allocation of prepacks can be prioritized before sending single units. Used when products are received grouped in packs in the warehouse, and not in single packages.

Min. Display: Minimum number of units that must be displayed in a store at product-size level. This can differ from one store to another.

Assortment: List of which products will be introduced in which stores. It defines the scope of first allocation.