Eskom Procurement Book 2015

TOTAL COST OF OWNERSHIP (TCO)

6.4 TOTAL COST ELEMENTS Regardless of where a company applies total cost models, these models all attempt to capture data beyond unit price. Like forecasting models, total cost models almost always have some degree of unreliability. This section explains how the categories of information that populate total cost models contribute to this inaccuracy. The data that populate a total cost model can usually be segmented into three categories. How the data are allocated across these categories affects the reliability of the total cost model. The first category, and the one that presents the highest degree of reliability, includes actual data. Unit price, insurance on an overseas shipment, and tariffs are examples of actual cost data when dealing with international suppliers. Few, if any, total cost models will include only actual data. The ones that do include are probably excluding some cost elements or data. The second category includes approximations or averages. The main characteristic here is the data are based on figures derived from internal sources. The challenge with many cost elements is that the cost of identifying the true cost of something could outweigh the value of the actual data. Would, for example, a detailed study be conducted to identify the true cost every time a late delivery occurred? The benefits of identifying the true cost of non-conformance may be outweighed by the cost in time and resources required to undertake the study. Overcoming this issue can result in the identification of various costs categories or accounts with standard or average charges applied whenever there is a need to allocate a charge. The following illustrates the use of averages. In a recent year, 50 late supplier deliveries cost a buying company an estimated R500 000 in total non- conformance costs. The average standard charge in the total cost system for a late delivery, therefore, becomes R10 000 (R500 000/50) per occurrence. Other cost categories could apply standard or average charges per hour. Correcting a minor defect in a supplier shipment might require 12 hours of labour. The model can include standard labour rates per hour. The usual warnings about using averages apply. A wide dispersion of true costs around this average creates a concern that the average charges may under- or overstate the cost. The least reliable total cost data are based on assumptions. Assumptions come from external sources that form the basis for applying total costs. Let’s say a study by researchers at a university concluded that it costs R150 every time a buyer issues an order. Therefore, every time a buyer issues a material order to a supplier, a R150 charge is applied to the total cost model for that item. But does the R150 charge really apply to what happens at every organisation? What if one company relies extensively on Electronic Data Interchange (EDI) to reduce ordering costs while others in the study do not? It is important not

6.4.1 TOTAL COST DATA

137 CHAPTER 6