Inventory control - Inventory Optimization Software

Inventory control (definition & insights)

Home » Knowledgebase » Here
By Joannes Vermorel, June 2013

All the processes that support the supply, the storage and the accessibility of items in order to ensure the availability of those items while minimizing inventory costs. In practice, inventory control covers diverse aspects including the management of inventory, recording both quantities and locations of items, but also the optimization of the supply.

Management vs Optimization

Inventory control is a broad domain that can be split in two major areas:
  • The management of inventory, which is almost impossible to dissociate from the inventory management software in most computer-based setups. When managing the inventory, the goal is to sustain a high productivity for all inventory operations.
  • The optimization of inventory, where the costs, such as carrying costs and stock-out costs have to be minimized while facing an uncertain future demand. When optimizing the inventory, the goal is maximize the financial output of the inventory for the company.
    While physically, there is only “one” inventory, those two areas reflect radically different problems, which are better addressed separately.

Management of the inventory

In its modern form, the management of inventory is quasi-indistinguishable from the inventory management software that supports it. Indeed, the software takes care of preserving an electronic representation of the inventory which is constantly used to instantly address routine questions that would otherwise require a very lengthy physical inspection of the inventory itself. Ex: how much units do we have left for the product X?

In order to maintain accurate electronic records of the inventory, all inventory operations need to be accounted for in the software. In practice, data entries are vastly accelerated by the use of barcodes or RFID (radio-frequency identification). In the most modern environments(1), physical operations performed on the inventory itself are robotized in which case the inventory is truly digitally managed from end-to-end.

Unlike the accounting system that focuses to provide an accurate financial reflection of the assets managed by the company, the inventory management system is action-oriented to support the company operating its inventory. The first goal of the system is productivity, that is, to perform all inventory operations whith the least amount of time or effort. The second goal of the system is provide a sustainable accuracy of the electronic representation of the physical inventory.

Optimization of the inventory

The inventory represents an anticipation of the future demand, and a financial trade-off between conflicting costs. Too much inventory, and the carrying costs skyrockets; and too little inventory, and there is nothing to serve anymore which incurs stock-out costs.

Unlike inventory management, the inventory optimization focuses on taking the best decisions that governs the inventory such as:
  • Deciding when and how much to reorder (see also reorder points)
  • Deciding where to store an item in the facility
  • Deciding which item need to be counted and when (see also phantom inventory)
  • ...

Once the decision is taken, it is carried through the inventory management system; however, the management system is not necessarily in charge of taking such decisions or even of producing suggestions to be manually validated by an operator.

The primary challenge of the inventory optimization problem is the uncertainty associated to the future demand. Indeed, as the future demand is unknown, most of the inventory optimization techniques rely on statistics to forecast the demand. The optimized decisions are computed as the ones that minimize the expected future costs.

Sometimes, combinatorial problems further complicate the optimization. For example, a retailer may wish to pass replenishment orders that take advantage of all weight and space available in the truck performing the delivery in order to minimize shipment costs. In practice, this means choosing the right mix of heavy items and bulky items.

Comparison of the two visions

As stated here above, managing or optimizing the inventory are distinct problems. The table below outlines the primary divergence between the two viewpoints.

Management Optimization
Essence of software Feature-driven. More features typically mean higher productivity because the software provides more support for less frequent situations. Performance-driven. The software is assessed based on the financial performance of the decisions computed by the software.
Organizational impact High. Most inventory processes of the company are directly structured by the software itself. Low. Inventory processes pre-exist, the system merely proposes alternative decisions.
Operational availability Real-time. If the software is unavailable, the company literally cannot operate its inventory anymore. Offline. Most inventory decisions are taken only once - sometime twice - a day, and decisions can be generated in batch.
Computational load Low. The software only needs to reflect physical movements of the inventory as they unfold which is very slow compared to the processing power available on a modern computer. High. The software needs to perform simulations, or equivalents, frequently reprocessing the entire history many times to perform the optimizations.
Cost of change High. As all processes are structured around the software. As the software represents the ''state'' of inventory, it is rather impractical to have coexisting systems as inventory records quickly diverge. Low. Multiple systems can coexist as long as an ''applicable scope'' is defined for each system. It is possible to gradually migrate from one system to the next.

Lokad gotcha

Historically, ERPs have emerged as monolithic solutions to address both the inventory management and the inventory optimization problems. However, as detailed in the previous section, the ingredients to make good inventory management software are very different from those required to make good inventory optimization software. As a result, we observe that companies that adopt a monolithic design nearly all suffer from either poor management or poor optimization - the latter being the most frequent case.

The problem is further accentuated by the pace of change within the software industry. Indeed, inventory management softwares are, by nature, very sticky software: once adopted, the cost of change is so great, that we routinely observe that it can take up to a decade to fully transition toward an alternative solution for large companies. While the delay is shorter for smaller companies, multi-year transitions are frequent. This means that many companies operate management software one or two decades old - losing the benefits that would have been brought the latter better solutions presently available on the market. However, as the cost of change is high, there is little to be done here.

In contrast, the optimization part comes with much lower friction which regards to the cost of change. Indeed, it is usually possible to have multiple systems, each one generating this own set of proposals (ex: the list of items to be reordered); and then defining a process to establish which scope of authority should be given to each system.

(1) For example, Kiva Systems produces order fulfillment systems that use mobile robots.