Introduction

1In many industries, purchasing managers are measured on costs or cost savings. However, “many of the measurements now in place don’t measure the things that count in competition” (Monczka and Morgan, 1994). Most academic articles discussing savings still use a theoretical rather than an empirical approach, and focus on Total Cost of Ownership (TCO) or soft savings (for example, Ellram [2003]). Unfortunately, only a few articles have analysed the evaluation of hard savings (for example, Monczka et al. [1979]), although it is well recognised that hard savings are an important component of evaluating a purchasing department’s performance. In fact, without evaluating hard savings, it is impossible to evaluate soft savings.

2In this chapter, “hard savings” refers to monetary savings directly impacting the bottom line and due specifically to the actions of the purchasing department (e.g., reduction in purchase price in the current year over the previous year). Conversely, “soft savings” refers to savings that could eventually be converted into monetary savings, but which are usually not, since they are more difficult to measure (Keen, 1997). For example, purchasing decisions could impact favourably the actual financial results of other departments (e.g., purchasing a higher quality raw material allows less rework and the generation of complete orders), but it is difficult to measure the specific impact of that decision on the bottom line.

3At first glance, hard savings seem easy to measure. However, in attempting to do so, organisations could get dragged down into a system that is too detailed, or ignores important aspects of the measurement. Here too, academic literature does not offer much help, since no agreement on a clear definition of cost savings has been reached, whether a few decades (Croell, 1980) or a few years ago (John, 2002). It is quite surprising that so few recent papers have explored the complexity of measurement systems. The technical side of measuring costs and the interest in publishing on issues that appear to be much more strategic might have decreased the level of prestige associated with the topic - and the resulting academic interest.

4This chapter provides a real-life example of how an organisation, aiming for a comprehensive cost savings measurement system in purchasing, has created high levels of complexity. In fact, at one point, it was almost impossible to manage. The methodology used to support our objective is a case study in a very well-known Canadian organisation, often looked at by others as a leader in cost savings measurement. We have followed the progression of this organisation over a 1-year period. The measurement system in place, mainly based on hard savings, was already extensive. Then, we witnessed its growth into a more complex system, followed by its rationalisation. Part of the simplification was carried out following our own observations and suggestions to management, at their own request.

5This chapter first positions our research within the literature. Second, the methodology and the presentation of the organisation introduce the case study. Third, changes made to the system are described and analysed, based on a model. Fourth, the five main lessons based on the actual system are discussed; since this analysis presents pitfalls to be avoided, it is useful not only to academics, but to managers as well. Fifth, there is a conclusion. Finally, limitations of the research are discussed.

Positioning the research within the literature

6Between 1880 and 1980, measurement systems were mostly composed of financial measures (Ghalayini and Noble, 1996). With the increased internationalisation, there was additional pressure to reach short-term objectives and a tendency to evaluate and compare the results of each department (Ghalayini and Noble, 1996; Easton et al., 2002); consequently, non-financial measures were needed. This development resulted in: (1) preserving an equilibrium in the system (John, 2002; Evans, 2004); and (2) linking organisational to departmental objectives (Ghalayini and Noble, 1996; Atkinson et al., 1997). These outputs impacted positively on the results (Matthews, 1998), the processes, the stakeholders (Ghalayini and Noble, 1996; Atkinson et al., 1997), and the evaluation of the efficiency of each function (Ghalayini and Noble, 1996; Trent and Monczka, 1998), including the purchasing function.

7There is no doubt that measurement systems have indeed interested many researchers, including those in purchasing. Lardenoije et al. (2005) indicate that this interest dates back many decades: for instance, in 1931, the National Association of Purchasing Agents (NAPA) organised a conference on this theme. According to the same authors, it is mostly in the 1970s that researchers really looked more closely at savings measurement systems in purchasing. This flow of research has resulted in a literature on aspects such as the specificity of purchasing performance (Van Weele, 1984), the suitability of performance measurement systems to embrace purchasing performance (Wouters et al., 2005), and the impact of the assessment of performance on the decision-making process (Dumond, 1991).

8According to Van Weele (2005), there are three main reasons why purchasing performance is an important issue for purchasing managers too: (1) it improves communication, by translating purchasing actions into something more tangible for other departments, an idea also supported by Castellano et al. (2004); (2) it enhances decision-making for the function and for the organisation, since this allows to take the appropriate corrective actions for the future (Van Weele, 1984; Matthews, 1998); and (3) it increases buyers’ motivation by showing them the tangible results of their efforts. As Dumond (1991) shows in her experimental study, a performance measurement system has indeed a significant influence on the decision process buyers use.

9Performance measurement is part of a system (Kaydos, 1999). Clearly, savings measurement systems - in purchasing or for an organisation as a whole - are one of the ways to measure performance. The first step is made of the disaggregation of corporate objectives into sub-objectives, which become associated with specific departments responsible for the optimisation of the results (Nilsson and Olve, 2001). These objectives are then be translated into budget indicators (or “financial summaries”). The next step is to determine the frequency of data collection. For each objective, the follow-up frequency (weekly, monthly, annually) will be based on a combination of both the cycle of the activity which performance is being measured and on the resources required for data collection. This step is important, since it makes the follow-up of each activity possible, and it helps to determine if the objective is to be met or if actions are required. Basically, performance measurement becomes a communication exercise aimed at reinforcing appropriate behaviours or at adopting better practices (Bayerlein and Gailley, 2005). An appropriate information system makes the follow-up easier, for data collection, storage and analysis purposes (Neely, 1999).

10The value added by a purchasing department mainly results from the evaluation of the savings generated (Van Weele, 2005), the obvious one being cost reduction (Van Weele, 1984). However, the supply function impacts a broad range of costs and savings. Consequently, no system is appropriate for all organisations (Pooler, 1973), since any system has to be compatible with the environment an organisation operates in. It must also reflect the development stage of the purchasing function (Van Weele, 2005). Therefore, it is no surprise that there are so many types of performance measurement systems.

11There are two other important problems often surfacing when evaluating and measuring purchasing performance: (1) formal objectives and performance standards for purchasing activities are not set clearly, and (2) an inaccurate measurement system is used. To address the first problem, it would be possible to evaluate purchasing performance by comparing the alignment of the supply function objectives with the strategic goals of an organisation (Evans, 2003). Strategic qualitative criteria, such as the development of supplier/buyer relationships, could also be used (John, 2002). The second problem, which deals with accuracy, is even more complex; not only is there no agreed-upon definition of what really constitutes performance, but there is a problem of scope: purchasing performance results from many activities which, due to their intangible nature, are difficult to evaluate. Although benchmarking is a common way to evaluate performance in supply, it can be difficult to compare an organisation’s performance with others’, since measurement methods and criteria are not universally recognised and applied (Easton et al., 2002).

12However, a savings measurement system must deliver more than strategic alignment: it has to be well understood and operational. It is clear that the measurement of savings is a major preoccupation for managers. In fact, 95 % of organisations evaluate the cost savings generated by the purchasing department (Evans, 2003). Unfortunately, “the basic problem is in defining just what constitutes a cost saving” (Croell, 1980). The general point of view is that cost savings include two categories: hard savings and soft savings (Keen, 1997). A third category, cost avoidance, is often discussed though.

13Hard savings is by nature quantitative, since “tangible” information is required for its measurement. A cost reduction is measured by comparing the negotiated price with a price referential (Monczka et al., 1979), or with the last price paid. Hard savings are related to all buyers’ actions that directly impact the bottom line, such as reductions in price, in the workforce, or in transaction costs. Since most organisations evaluate purchasing performance based on the price paid, hard savings are of utmost importance in actual measurement saving systems (Verma and Pullman, 1998). It is useful to know that in such systems, “workforce” refers to the human resources directly involved in purchasing, and not the reduction of the number of employees in other departments, following negotiations with suppliers (e.g., the supplier now proceeding with quality control, which used to be performed in-house).

14Soft savings refers to savings created by outcomes based on “qualitative” criteria; their impact on the bottom line might be indirect, for instance when benefits are spread across various departments. For example, a saving could result from a packaging modification due to a reduction in the time required at the warehouse; this change could increase workforce productivity, but not necessarily reduce the workforce per se. Since there is more subjectivity involved, the complexity of measuring soft savings accurately is greater than for hard savings. However, most soft savings can be converted into hard savings, by using industrial standards or other measures presented in the literature (Keen, 1997).

15Cost avoidance is the reduction or the elimination of a future cost. For example, following a supplier’s successful bid for a raw material at a cost higher than the previous year’s, a purchaser is nevertheless able to get a price reduction; this is often considered a cost avoidance (subsequent negotiation by the purchasing department), rather than a hard saving (bid that would have been at a price lower than the previous year’s). Since in a case like this, there is also a saving, professionals do not always make a clear distinction between cost avoidance and hard savings (Monczka et al., 1979). However, some authors consider that cost avoidance is not a saving at all: “… whilst it can involve significant procurement activity to negate inflationary pressure for instance, it does not contribute to the financial accounts.” (CIPS, 2002). However, most researchers agree to measure cost avoidance, but not to integrate it into savings measurement (Monczka, 1974; Dobson and Shorrock, 1980; Beidelman, 1987; Morgan, 1990).

16From the discussion above, it is clear that evaluating supply performance is not as easy as it seems; this supports what Van Weele (1984) already wrote in 1984 Even nowadays, the purchasing function remains one of the most difficult to evaluate (Easton et al., 2002). According to Purchasing, 37 % of respondents are dissatisfied with the actual performance criteria used in their organisation (Morgan, 2000). Purchasers often consider that performance criteria are non-representative of their work.

17Although many interesting articles have been written about the strategic implications of measuring purchasing performance and the difficulty in measuring it (for example, see a 30-year-old article by Ammer [1974], and, more recently Morgan [2000], and Jacoby [2005]), there is a major void on the operational side. For instance, how does a measurement system evolve within an organisation and why does it evolve the way it does? This research provides answers to this question and illustrates why it is so difficult to implement measurement systems; it also reviews the challenges stemming from the variety and the complexity of the measurement criteria.

The methodology selected: the case study of a renowned organisation

18Our research adopted a case study approach for three key reasons, namely to describe: (1) an actual system and its intricacies appropriately; (2) the changes having taken place over time and the reasons thereof; and (3) the dynamics of a measurement system.

19According to Yin (1994), a case study is appropriate “when the focus is on a contemporary phenomenon within some real-life context”. This research is a descriptive-exploratory analysis, which is suited to this kind of chronological structure (Yin, 1994). A case study is not simply a methodological decision, but the selection of one object on which the study focuses (Stake, 2000). A sample of one organisation allows researchers to concentrate their efforts on a unique situation compared to the general situation or other specific cases (David, 2004). It also provides them with the opportunity to evaluate not only the actual situation, but the impact of the changes made to the system over time.

20Internal validity has been built through an explanation-building approach, which consists of analysing “… the case study data by building an explanation about the case” (Yin, 1994). Consequently, the methodology selected did not aim at measuring causality and did not allow quantitative analyses. The construct validity was supported by the use of multiple sources of evidence (interviews, documentation, etc.), by a chain of evidence during data collection, and by the managers having reviewed the case study, such as suggested by Yin (1994).

21Since the researchers became very familiar with the savings measurement system, the Director asked for their comments, mostly before major changes were considered. The fact that many of the researchers’ ideas were implemented in subsequent systems has no impact on validity, particularly external validity (both researchers observed the same characteristics for the system, and had access to the Director when required, particularly to confirm major observations). Also, the fact that this research was about the actual evolution of a savings measurement system, based on decisions made by managers (and not by the researchers) alleviates greatly any methodological caveat. Although this was at no point in time a consulting assignment, it is normal to provide advice in exchange for the research opportunity.

22The in-depth investigation took place in a well-known Canadian organisation (subsequently called ABC), mostly during 2005. The organisation operates in a relatively stable environment, but it offers products and services that require innovative tools and production methods.2 Purchasing is centralised in one department of 125 people. In 2003, the organisation acquired for 1.1 billion dollars of products and services, which represented a 29 % increase over the previous year. Purchases are often based on contracts that can be extended for up to 5 years, an approach which allows the organisation to maintain long-term relationships with some of its suppliers.

23The organisation has an excellent reputation and is often used as a benchmark by other organisations, for its savings measurement system. The supply managers have worked hard at improving this system. But even an organisation used frequently as a benchmark hides telling stories… As will be developed in the next section, many purchasers as well as managers of other departments were getting dissatisfied with the savings measurement system which had been in place since 1997. It was time for a change. In reality, not only one, but two major changes were made to the system in just over a year Information for this case analysis was gathered through: (1) in-depth interviews with the managers involved in the changes made to the measurement system in 2004 and 2005; (2) regular follow-up during 1 year on the evolution and the impact of the changes made; and (3) internal documents related to the computation of savings and to the measurement process.

24The initial interviews, based on a detailed interview guide, were made with the two individuals (one Director and one Assistant Director) in charge of the modifications to the system. Initially, our intent was to interview about 10 purchasers as well, mostly to determine their perception of the system and which suggestions they had to make to improve the system. However, as evidenced by the modifications made to the savings measurement system, the pressure was on the Director to come up with changes quickly. Therefore, both the Director and the Assistant Director went ahead with interviewing purchasers about the savings measurement system. Consequently, this data, which became secondary - rather than primary - data for the researchers, was nevertheless taken into consideration for making the initial changes to the system. When we asked if we could still interview some purchasers, we were told that the savings measurement system had become such a sensitive issue with purchasers, that it was preferable not to do so.

25However, in addition to the secondary interview feedback, two other data collection instruments proved very valuable: the regular follow-up and discussions (in person or on the phone) with the Director or the Assistant Director, and the access to all internal documents about the system. In fact, since the Director had made 10 presentations to other Directors and Vice-Presidents so far, all the supporting documents and the slides were made available to us, so that we could follow all the important avenues considered for assessing the initial system and the potential changes.

26Detailed notes were taken during each discussion, and these were analysed shortly thereafter. Since we had a relatively easy access to the managers involved in the process, we could clarify any aspect subsequently. Our focus was clearly not on the amount of the savings per se, but on the changes to the process, on managerial perceptions and on all information shedding some light on aspects such as the perceived appropriateness of the system.

27One of the Director’s main objectives for having accepted to collaborate in this research was the intent to use the researchers’ questions and comments, in order to develop a more suitable system. In fact, over time, two of the researchers became so knowledgeable about the system, that their expertise provided valuable insights to the Director. Some of the positive impacts that this research had on the savings measurement system were:

1. A visual representation of the initial system, and then of the modified one (see next section for further details). At the time, that figure was used by the Director for at least one of his presentations to middle and top management. The intent was to continue to develop similar figures as changes were made to the system, so that purchasers could use them too.
2. Questions or comments that demonstrated some of the weaknesses of the first two systems.
3. Specific suggestions on ways to improve the second savings measurement system; as a result, many of the suggestions were incorporated in the subsequent system.

28It could be said that the researchers became progressively more involved in the changes, albeit indirectly, through their questions and comments. This is actually a very valuable characteristic for a research project like this one, since the analysis can be done on an on-going basis. In addition, the fact that two of the researchers became so familiar with the system provides even more accuracy for the system description and implications, as if the researchers would be a third party (observer). However, it is important to mention that for the Director and the Assistant Director, our input was only that, i.e., input. They were always the ones making all decisions about the changes to the system.

Analysis of the dynamics of the savings measurement system

29Measurement systems do evolve over time. Based on the ABC case, and our experience with other organisations, as well as on the literature, we believe that the three-phase model developed here could also be representative of savings measurement systems in other organisations (see Figure 1).

Hard savings were measured first

30At ABC, the first measurement savings system was initiated in 1997. It took into account only hard savings, as illustrated in Figure 1. This is clearly representative of phase I in our model, since it is easier to agree upon and measure hard savings. At that time, managers believed that measuring hard savings was a more conservative and a more appropriate approach, since the measurement of savings was a new approach in this organisation. Actually, the Vice-President and the Directors were concerned that the implementation of the system could be compromised by other departments. As can be expected, the savings generated usually decrease over time, despite constant or even increased efforts. In fact, at one point, it might be impossible to get a better price, unless buyers feel justified to acquire lower quality products (Dumond, 1991) or suppliers assume a loss, which transfers the problem to the supply chain (Anderson et al., 2003). Often, the emphasis is definitely on price. However, “if price alone was going to be the yardstick by which to measure purchasing and its suppliers, things would never improve. Sourcing decisions must be predicated on value, not price.” (Raia, 1989).

Figure 1. Complexity of measurement systems

31At ABC, the first system was composed of four stages. In stage 1, named “validation of the purchasing strategy”, only savings generated by one of the so-called “accepted purchasing strategies” (for example: negotiation, promoting healthy competition, seizing an opportunity to take advantage of market trends) could be considered. This approach is typical of methods used by managers to align savings measurement systems to strategic corporate decision making, as recommended by Evans (2003). Stage 2 was made of two steps:

• The first step consisted of selecting the most appropriate comparison basis. Savings from repeat purchases could be measured against the historical price of a product, of a group of products, or of a group of suppliers. A new purchase could be compared with the initial supplier price, the market price, or the client’s evaluation.
• The second step was the selection of the appropriate measure. Since, there were no formal rules to select the best method among the nine savings measurement methods then recognised, we created a decision tree. Those nine methods, named M1-M9, are described in Table 1. The decision tree is similar to Figure 2, which includes not only M1-M9 (phase I of the savings measurement system), but also M10-M13 (phase II of the system). Since both decision trees are similar, we are presenting only the most complete one (see Figure 2).

32In stage 3, purchasing personnel compiled the savings twice a year. Lastly, stage 4 was the follow-up of savings. Over time, it became clear that personnel at ABC tended to use only a few of the measurement methods. In some cases, the choice was easy and justified. However, in other situations, the savings could be due to a combination of methods. There were many instances where buyers became frustrated, since their efforts and strategies generated savings, although the system did not take into consideration what they thought were savings. In fact, what they termed “quantitative savings” (actually, hard savings) was clearly insufficient to describe their accomplishments.

33In 2004, cost avoidance of over 10 million dollars (not considered by the initial system) partially prompted the change to the existing measurement system. At the same time, other departments wanted to obtain more precise information through their budgets, since top management was asking more questions about their “savings budget”. In fact, what was happening is that most hard savings had already been taken into consideration over the years. So, where could additional savings come from, if not mostly from soft savings? However, this integration of both types of savings had to be done properly.

Adding soft savings to the picture

34A revision to the system, corresponding to phase II of our model, finally took place. In phase II, and as illustrated in Figure 1, less emphasis is placed on the measurement of hard savings, and more efforts are put into the computation of soft savings. Surprisingly, despite the 2004 event creating significant dissatisfaction, there remained a reluctance by purchasers and managers to move away from something they were familiar with (or so they thought). A new system was implemented at the end of 2004 and early 2005.3 The Vice-President decided to maintain the prerequisite that all savings needed to be supported by one of the “accepted purchasing strategies”. Four measures were added, (see methods M10-M13, in Table 1 and in Figure 2), for a total of 13 measures. As discussed in the methodology, this figure proved to be very valuable subsequently to the organisation.

Table 1. Measurements in the 1997-2004 system, and in the 2004-2005 system*

Figure 2. Decision tree before the rationalisation process (stage II)

The importance of the communication plan

35In order to improve internal clients’ satisfaction, the purchasing managers determined not only the impact of savings on the actual budget, but also the expected impact on subsequent budgets. This was viewed as trying to communicate more accurate and timely information. Departmental budgets now included three components: hard savings (mostly measured through methods M1-M13), avoided costs (not considered by one of the methods, although the amounts were computed), and cost increases. Although these changes might appear minor, they were not: (1) all managers had to agree as to what the three components included; (2) in addition, many changes to the computer programs were made, competing for time allocation with changes in other areas of the business.

36The organisation also developed financial summaries to demonstrate the benefits that the purchasing department was bringing to other departments. The main objective of these was to provide information in addition to what was communicated through the savings measurement system. For example, instead of measuring only the savings over the previous purchase (already done by methods M1-M9 in phase I), a longer time frame was also considered to demonstrate the long-term benefits for clients to use the purchasing department’s services. This also motivated buyers, since they now saw better the impact their work had on cumulative savings. Another financial summary presented avoided costs, but only when the savings had not been recorded through any of the other measures. This approach is in accordance with Monczka et al. (1979), who mention that avoided costs should be measured, to emphasise the importance of resisting cost increases. The plan also proposed to present other qualitative and quantitative benefits such as volume rebates, reduction of lead times, and inventory reductions.

37Comparison bases also underwent changes. The organisation modified the basis to determine the historical price. Instead of a base year (i.e., 1997, when the system was first developed - see Table 1), it was now the last purchase price or the price paid in the previous year which was used. In the case of a new product, the comparison basis used to be the client’s estimate; under specific circumstances, buyers were now allowed to modify these estimates, so that they were reasonable. As for soft savings such as inventory reduction due to “better purchasing”, the only accepted measure would now be the initial price or the market price.

The savings measurement system becomes too complex

38Despite the fact that much time had been spent over the years (particularly in 2003 and 2004) thinking about potential improvements to the system, phase II of the measurement system lasted only a year. The modified system was simply too complicated for purchasers and users. The first group did not see the necessity of such a detailed system, despite the fact that their complaints about not taking all savings into consideration had been listened to. The second group was actually not looking for a complex system, but rather for a few key numbers.

39Actually, there comes a time when too many - and often too complex - measures are used to measure cost savings. In phase III, the process of measuring performance consumes more resources than the marginal benefits it brings, and a rationalisation of the system is required. The difficulties in developing or acquiring the competencies required to put these advanced systems in place can also be a barrier to their adoption (Evans, 2003).

40At ABC, the table was set for phase III of the model: simplification. Another round of changes took place at the end of 2005, since the system was too complex and needed to be rationalised in order to be more effective (achieving goals), but also more efficient (too many resources used). For this phase, a much-improved communication plan oriented on what the users wanted was put in place. With a simpler system, it was easier to explain how the system worked.

The focus is now on the individuals the system was for in the first place

41Why were the changes made in phase III selected rather than other changes? Considering that the main objective was to simplify the system both for the purchasers and the users, each method (M1-M13) was examined carefully for a combination of effectiveness, frequency of use, value added, efficiency and clarity. Other methods were also discussed. The changes also took into consideration that top management and other users were now more convinced about purchasing’s contribution and efforts. Therefore, some information previously reported could be grouped or simply abandoned. A reduction in the number of measures is clearly warranted in phase III, since a lot of thought has already been put into what is really useful and at what cost - or information overload - it comes.

42Here are a few representative examples of the changes made. First, the researchers proposed to eliminate the clients’ estimate as a comparison basis, and to replace it by the market price and the initial supplier price. Standards created in-house can indeed be subjective, and can underestimate the real potential savings (Mendleson, 1969; Matthews, 1998; Leenders et al., 2005). Managers can also voluntary overestimate the potential savings in order to look better to headquarters. Secondly, some associations between the comparison basis and the measures have been improved. For a renewal, the new price used to be compared with the price paid in the base year (1997), but the new system proposed to use the last price paid. An improvement made to that new system was to use the initial supplier price to measure avoided cost. As researchers, we also proposed to managers to use the market price or the initial supplier price to measure hard savings resulting from a renewal.

43As illustrated by the few examples above, process rationalisation occurred mostly in the measurement methods. Globally, three measures were eliminated, six measures were grouped into two, one was added, and only four were not modified (see Table 1). The system, which had 13 measures, now has 7. The additional measure (M14) was developed specifically to measure avoided costs. Two measures could not really be associated to any comparison basis, and they were therefore deleted.

44Based on the experience of this very successful organisation with its savings measurement system, valuable lessons can be learned. At first glance, those lessons might appear quite obvious; however, the reader should be careful to think that way, because:

1. Although the extant literature makes it possible to deduce some of the lessons, it is always easier to think so after having read that lesson. If it was that easy, other articles would have been written with a similar content before ours.
2. Despite the fact that the lessons might appear to be “general”, they are nonetheless based on the experience of managers and personnel in a well-known organisation. Managers in the supply department confirmed that if they would have had these suggestions, understood how they integrated one with another, and implemented all of them, it would have made their job much easier. Although some lessons might appear to be general, they can still be useful and precise.

45The next section thus focuses on the practical implications, often termed “contribution for managers”. However, we believe that these aspects also carry theoretical implications, and can therefore be useful to academics as well.

Five practical lessons, working best when integrated

Lesson 1. Determine what will be measured and what it will contribute to managers

46A savings measurement system is not simply the outcome of deciding, for example, to compare past and present prices. It entails a clear analysis and understanding of what will be measured and how it will be appropriately measured. At ABC, only in phase III was there a clear understanding of what managers really wanted: a simple system reporting key numbers. Effectiveness could finally come with improved efficiency.

47The categories of savings presented to managers through financial summaries are selected according to the objectives pursued. They are also reflected in the measurements, since each measurement can only take one category of savings into consideration (see M1-M14 in Table 1). The format of the information presented can be quantitative or qualitative, and show monetary value or percentages.

48Although the measurement method influences the selection of an appropriate comparison basis, the latter also influences the accuracy of the measurements used, but to a lesser extent. Savings are always relative to what one compares them to. So, using the wrong comparison bases can be overly conservative or result in over-estimating the savings. Establishing these relationships helps managers to elaborate and improve the system.

49Managers should create a system based on their own experience and knowledge, on the strategic role of purchasing in the organisation, and on the main objectives pursued with the measurement system. The interrelations among the variables are a good approach to initiate the creation of a system. However, this approach will not help avoid the pitfall of overloading the system.

Lesson 2. Avoid all-encompassing and complex systems

50“[…] It is possible to argue that in many companies we suffer not from too few measurements but from too many, and those that we have are frequently of the wrong sort.” (Syson, 1995). A system that is too complex unnecessarily burdens purchasers for the measurement and the follow-up of savings, but also lengthens the training time required.

51At ABC, one purchase could impact more than one category of savings (hard savings, avoided costs, etc.). Since it was not unusual for one purchase to use up to 5 of the 13 measures, it also often implied more than one comparison basis (client’s estimate, initial supplier price, etc.) to measure a saving. A well-documented system and some tools such as a decision-making tree (see Figure 2) can help see incoherence, facilitate training and determine which measure is the most appropriate. However, the complexity will often vary depending on the development stage of the system. When a system becomes too complex, often following an approach by trial and error, there is a desire to move to a simpler system that will emphasise effectiveness, while being much more efficient (phase III).

Lesson 3. Measure hard savings first

52If managers have never measured savings, it would be difficult to create a savings measurement system that would integrate both hard and soft savings. In fact, to demonstrate the benefits of the system, it would be wise to begin by measuring only hard savings. At ABC, clients’ apprehensions about the representativeness and the subjectivity of the system, pressures from the controller to be conservative, and purchasers’ reluctance to change were all considerations related to people that pushed the supply managers to keep the system simple by considering only hard savings. There is no reason to believe that those “normal” human reactions would be very different elsewhere.

53Once the system has been launched and initial worries taken care of, a calm period begins, characterised by a greater acceptance of the system. However, systems are dynamic. Therefore, some forces impact them, and changes are bound to occur. At ABC, internal clients wanted more detailed information, and purchasers were disappointed that savings were decreasing, despite their continuous efforts. The challenge was to incorporate other categories of savings and to create a better communication system, while increasing the representativeness and the accuracy of the system.

Lesson 4. Determine how to integrate soft savings properly to a measuring system initially based on hard savings only

54As the system evolves, one of the most important steps is definitely the integration of soft savings into the system (partly done in phase II at ABC, mostly through financial summaries, rather than through the specific M10-M13 measures). Despite the well-known difficulty of this integration, little information is available to managers on the subject. Without a clear understanding of what constitutes a soft saving, it is suggested to incorporate all soft savings into one category.

55Measuring “qualitative” savings can be quite complex. One way to do so is to transform them into quantitative savings. Due to the additional efforts required, this approach should be used only for major savings. At ABC, purchasers also did a list of other less important soft savings; this demonstrated that other savings were also achieved.

56In phase I, the challenges of the managers at ABC were mainly related to people. Subsequently, it was the technical aspects of the measurement and of the communication system that required a lot of attention. This is typical of phase II. Recognising these technical challenges and the fact that the new system was too complex, the managers made the appropriate changes, which resulted in phase III. It seems that it can take time to move from phase I to phase II. However, phase II does not appear to be sustainable, mostly due to the complexity of the system. The pressure to change will be more present than in phase I.

57Overall, an organisation has to find a balance between the conservative approach of considering only hard savings (phase I), and a too precise system that could ultimately include every possible soft saving (phase II). Since a system is dynamic, an approach that was valid at some point may not be - or look - as valid later.

Lesson 5. Use a solid communication plan

58Clarifying users’ objectives in the communication plan is necessary. Users are often more concerned about short-term objectives, but long-term objectives should also be integrated to support purchasing’s efforts in making its contribution known. Once these objectives are established, they need to be communicated properly. Otherwise, users might not understand how the system works, and might even lose interest in the system itself. However, organisations need to be careful when deciding which information should be shared, since providing too much information will confuse clients instead of improving system utilisation. A solid communication plan is even required when the measurement system moves from phase II to phase III.

59The larger the number of measurements, the more complicated it is to explain the differences among them. The same is true if too many types of financial summaries are presented to managers, since what these individuals really want are a few key reliable numbers. The purchasing managers at ABC had created additional summaries to fulfill internal clients’ desires for a better communication. Although users appreciated the efforts made, they (and even purchasers) had difficulty understanding some of them. In retrospect, much of the communication plan was reactive, when it should have been part of the overall implementation plan for the savings measurement system.

Conclusion

60The measurement of purchasing’s contribution to the organisational performance is an important issue for both practitioners and researchers. In this chapter, we have presented and analysed the dynamics and the complexity of an actual (and likely representative) measurement savings system. The research results have been generated through a case study, a methodology found to be appropriate for the objectives of this type of investigation. Interviews were conducted with those individuals directly involved in the conception and follow-up of the performance measurement system. Although this was a research and not a consulting assignment, we were asked for our input, due to our intimate understanding of the system. Most of our suggestions were actually incorporated into the new system.

61For many organisations, measuring hard savings is less challenging than doing so for soft savings. The ABC case emphasises the concept of a performance system. Performance measurement is about communication, in order to reinforce some behaviours. The follow-up frequency must be related to the operations cycle, so that there is only a small-time gap between action and results evaluation. This case shows that there is a trade-off between the frequency of the measurement and the resources required: at a certain point, efficiency becomes a major preoccupation, although effectiveness remains the main target. In fact, having too many measures or/and financial summaries is not as useful as selecting the right ones for each stakeholder. Therefore, a purchasing manager should develop reports and systems that are well targeted to the audience.

62The real challenge in this chapter was to analyse and present the technical aspects, without going into too many details, while still presenting helpful conclusions. Through a diachronic model presented in Figure 1, we have proposed an inductive vision of how such a system evolves over time. The figure illustrates the limits stemming from the growing complexity of the savings measurement system and, how the system was finally redefined in order to be more effective. No conclusive answers can be given regarding how a savings measurement system must evolve, but we have pointed out some important lessons making it possible to understand better: (1) how an actual measurement savings system works; (2) what the challenges are; (3) some of the important principles that can make a system operational, despite its complexity.

Recommendations for further research

63A single case study was the methodology selected, since the information required was based on very precise details. No comparison with other organisations has been made. So, to determine if the results could be generalised, further research would be required. Also, since the methodology used was qualitative, no statistical evidence could result from this research. However, the objective of the research was not to measure any aspect of the system, but to present qualitative information about the dynamics of a system, and the implications thereof. It would have been useful to discuss directly with the purchasers, since they were the ones involved daily in the system. Although this places a limitation on the research, it also opens the door for future research.

64There are additional issues which could be addressed in future research. First, soft savings can be transformed into hard savings, but also into avoided costs. Research to determine how organisations can actually transform “qualitative” into quantitative savings could encourage managers to measure more savings. Second, a study related to the limits of hard savings would fill a gap in the literature (and help many supply practitioners faced with demands from top management). Lastly, it would be interesting to follow a few products over many years to observe at what point and under which conditions hard savings reach a plateau. Then, an analysis of the organisational and market contexts could provide some clarification as to how it happened, and what comes next, for instance, efforts for additional savings, other types of savings, or allocation of the resources elsewhere.

65Purchasing managers should not try to have a perfect savings measurement system that takes care of all situations and pleases all, otherwise it will simply drown the objectives. The same is true for this chapter. By analysing the technical aspects and the evolution of an actual complex system, we have focused on a constant preoccupation for managers. It is not always by following other researchers or taking the easy path that research really progresses. Research can indeed be messy and useful.