The current issue and full text archive of this journal is available at www.emeraldinsight.com/0957-4093.

htm

IJLM 18,2

Logistics collaboration in supply chains: practice vs. theory

Erik Sandberg
Department of Management and Engineering, Institute of Technology, Linkoping University, Linkoping, Sweden
Abstract
Purpose This paper aims to present results from a survey that investigates the situation in real world supply chains concerning logistics collaboration. Design/methodology/approach Based on supply chain management literature, a questionnaire was developed covering important topics from the literature. The questionnaire was sent to the logistics manager at Swedish manufacturing companies and a net response rate of 37.8 percent (177/482) was achieved. Apart from purely descriptive statistics, factor analysis, cluster analysis, analysis of variance (ANOVA), and cross tabulations with x 2 tests were used for the analysis. Findings From the results, three major conclusions are drawn. First, there is a clear relationship between the intensity of the collaboration and the positive effects experienced from the collaboration. Second, the results indicate that top management is an important driver for higher intensity collaboration. Third, there are serious differences between supply chain management (SCM) theory and practice, due to the absence of strategic elements in the collaboration and the different ways in which supplier and customer collaborations are managed. Research limitations/implications Related to SCM theory, the lack of strategic elements in the collaboration might prevent or decrease the expected positive effects of the collaboration. It is therefore important to increase the strategic level component in the collaboration. Top management involvement is likely to be important to achieve this and hence more research on the topic is suggested. Originality/value The study presents a broad overview concerning logistics collaboration that covers many of the most important supply chain management issues. In addition, their relationship is discussed. Keywords Supply chain management, Cooperative marketing Paper type Research paper

274

The International Journal of Logistics Management Vol. 18 No. 2, 2007 pp. 274-293 q Emerald Group Publishing Limited 0957-4093 DOI 10.1108/09574090710816977

Introduction During the 1980s and 1990s a new trend towards integration and collaboration instead of so-called arms-length agreements between suppliers and customers has been recognised by researchers as well as business practitioners. Actors participating in the same supply chain identify tradeoffs with their adjacent customers and suppliers and recognise the importance of integration in the chain in order to focus on what is offered to the end customer in terms of cost and service. Internal excellence is not enough anymore; there is also a need for external excellence in the whole supply chain. This is the philosophy underpinning supply chain management (SCM), which has received enormous attention in research journals as well as in industry. (Christopher, 1998; Lambert and Cooper, 2000) True SCM-based collaboration among supply chain players can have signicant benets. According to researchers as well as consultants, massive reduction of costs and improved service are possible (Sandberg, 2005). Within the eld of logistics, best practice companies have applied collaborative approaches based on the SCM philosophy

and have achieved extraordinary results. As an example, Wal-Marts collaborative planning, forecasting and replenishment (CPFR) collaboration with their suppliers is a well-known success story. SCM literature, which often takes its starting point such collaborations represented by Wal-Mart, is fairly unanimous: SCM initiatives in supply chains are positive and are expected to be benecial for all parties involved. These discussions and conclusions are however often on a conceptual level and are seldom based upon more rigorous theory (Bechtel and Jayaram, 1997) or empirical material (Lee and Whang, 2000; Stank et al., 2001) and are therefore often supercial. As a consequence of this, the literature becomes fuzzy about what actually occurs when companies collaborate and what specic effects are achieved. Therefore, more research, especially empirical, is needed in order to verify existing literature. Furthermore, it could be questioned how much SCM exists in todays supply chains. Can SCM be considered to be implemented in most supply chains, or should SCM be considered as the exception rather than the rule? From a logistics perspective the question of what SCM and collaboration really means for the individual company is an issue. How is collaboration performed and what actions are undertaken? What parts of the very broad spectrum of actions suggested by literature have been adopted? For what reasons is collaboration applied and what barriers to collaboration exist? What are the consequences of collaboration in terms of effects? To answer these questions an examination of the situation in existing supply chains concerning logistics collaboration was undertaken. In November 2004 a survey was mailed to 482 logistics managers at Swedish manufacturing companies[1] with more than 100 MSEK (, 12M Euro) and 100 employees. Based on the SCM literature, the survey covered a number of important collaboration issues such as the content of the collaboration (i.e. what is actually done in logistics collaboration), driving forces, the barriers experienced and the effects. The term collaboration was dened in the survey as a relationship characterised by openness and trust where risks, rewards and costs are shared between the parties. Furthermore, a basic prerequisite is that all involved parties should be able to inuence the design of the collaboration. Typical collaborative logistics activities are joint planning and information sharing. This paper aims to present some of the main results from the study. In this study, the perspective of a focal company situated in a supply chain is taken. The emphasis is on how the focal company acts and behaves compared to what is recommended in the SCM literature. Figure 1 shows the focal company and its main
Triadic collaboration where both suppliers and customers are involved

Logistics collaboration in supply chains 275

Supplier

Focal company Dyadic collaboration with supplier

Customer Dyadic collaboration with customer

Source: Sandberg (2005, p. 6)

Figure 1. The chosen perspective and the three types of collaboration investigated in this study

IJLM 18,2

276

collaboration possibilities in the supply chain. The focal company can be involved in three types of collaboration, each of which were investigated; a triadic collaboration where both the supplier as well as the customer to the focal company are involved, or dyadic collaboration with either a supplier or a customer. As pointed out in the Figure, only so-called primary members of the supply chain (Lambert and Cooper, 2000) are considered. This leaves for example collaboration between the focal company and a third party logistics provider outside the scope of this study. Methodology This study can be classied as a descriptive survey. A descriptive survey aims to describe what situations, events, attitudes, or opinions are occurring in a population (Pinsonneault and Kraemer, 1993, p. 80). Hypotheses exploring connections between variables can be investigated. However, in contrast to an explanatory study, no causal hypotheses are set up and tested:
Analysis stimulated by descriptive questions is meant to ascertain facts, not to test theory. The hypo is not causal but simply that common perceptions of the facts are or are not at odds with reality (Pinsonneault and Kraemer, 1993, p. 80).

In the study no clear, determined hypotheses were tested. However, since the questions in the questionnaire were all generated from the SCM literature, they can all be considered as theoretically underpinned and therefore it is the SCM literature that is indirectly tested. The questionnaire contained in total 30 questions on 10 pages. After a few general questions about the company, the respondents were asked to pick the most important relationship they had with either a supplier and/or a customer that they considered as collaboration. If they had any triadic collaboration, they were asked to choose this collaboration before dyadic ones. In the rest of the questionnaire they were asked questions about their chosen specic collaboration. The term collaboration and the purpose of the study, etc. were specied in the questionnaire as well as in a cover letter. The questionnaire and a cover letter were mailed to the 482 logistics managers in November 2004. Of the answers received, 177 questionnaires were considered useable and were coded in SPSS, which means a net response rate of 37.8 percent (177/468). In her doctoral thesis, Forslund (2004) makes a literature review on response rates in other similar research projects. Those response rates ranged between 5 and 58 percent. In comparison, it can be concluded that the response rate received in this thesis is satisfactory. Furthermore, 37.8 percent is far more than the minimum suggested by Forza (2002), which is 20 percent for mailed questionnaires. Three actions were undertaken to investigate possible systematic differences (Malhotra and Grover, 1998; Forza, 2002) among respondents and non-respondents. First, differences in the number of employees were investigated between respondents and non-respondents. As a second action, 41 logistics managers that had not answered the survey were briey interviewed about their reasons for not answering. Above all, lack of time was the most common reason. No statistical differences concerning non-response could be found. Finally, as a third action to investigate systematic differences between respondents and non-respondents, early respondents (i.e. those respondents that answered the questionnaire within two weeks) were compared with late respondents (those respondents that answered after two weeks).

The data analysis of the questionnaire was made in SPSS 11.5. Apart from purely descriptive statistics such as mean values, standard deviation, and frequencies also factor analysis, cluster analysis, Analysis of Variance (ANOVA), and cross tabulations with x 2 tests were used. Results Because of the chosen industry, most of the respondents characterised themselves primarily as a supplier or manufacturer (14.7 percent and 83.1 percent, respectively). Furthermore, a majority of the respondents belonged to a business group (94.4 percent). The mean number of employees of the companies was 490 and mean annual turnover was SEK 1344 million (, 100 million Euro). The characteristics of the products manufactured by the respondents were characterised by two ve-point Likert scales; one ranging from single products to volume products, and the second ranging from customer specic products to standard products. The mean values were 3.84 and 2.66, respectively. The respondents were asked to dene which of the following three types of collaboration they had in mind when they answered the questionnaire (Figure 2). The distribution shows that the majority of all companies are not engaged in triadic collaboration, which is considered to be the minimum scope for SCM by many authors (Mentzer et al., 2001). Of the respondents, 17 (9.6 percent) answered that they had no relation that they considered as collaboration. Process According to the SCM literature, the transition towards a process view is without doubt one of the most important characteristics of SCM. A description and denition of the collaboration in terms of a process will lead to a better understanding of what activities are involved in the collaboration and how they are related to each other. This becomes especially important in interorganisational collaboration when the process stretches over company borders, since it is often difcult to understand and to get a good insight into other actors internal activities (Melan, 1993; Willoch, 1994). The use of processes in the collaboration with adjacent suppliers and/or customers is however very low. Only 24 percent of the respondents have together with their partner/partners dened and described their collaboration in terms of a process. This gure must be considered as surprisingly low given the massive amount of literature
Triadic collaboration 44 (24.9%)

Logistics collaboration in supply chains 277

Supplier Supplier collaboration 60 (33.9%) Source: Sandberg (2005, p. 90)

Focal company Customer collaboration 56 (31.6%)

Customer

Figure 2. Distribution of different types of collaboration chosen by the respondents

IJLM 18,2

278

advocating a process approach over the past two decades; SCM literature and logistics business concepts such as CPFR as well as other management concepts such as business process reengineering are all based on a process-oriented view of the rm. The use of process related measurements in the collaboration was investigated with three different types; total logistics costs for the collaboration and different kinds of lead times and services between the actors. The mean values and standard deviations of the answers, given on ve-point Likert scales, can be seen in Table I[2]. In general, lead times and service measurements are used considerably more than measurements related to the total logistics costs. This also agrees with the internal use of measurements at the respondents companies where service related measurements are applied far more. Other studies report similar ndings (Aronsson, 2003) and can probably be explained by the difculties of measuring logistics costs. The results of this study also show a relationship between having a dened process for the collaboration and the measurement of it; ANOVA analysis shows that respondents with a dened and documented process in their collaboration also measure it more. This is valid for measurements of costs, lead times as well as service. Probably as a consequence of a low degree of process denition, the respondents have not applied a predened way of how to share costs and savings in the collaboration to any great extent (a mean value of 2.54 on a ve-point Likert scale). This result is also in line with the results presented by Spekman et al. (1998), who found that risks and rewards were shared equally between the partners to a very low extent. These ndings indicate a great potential for improvements in this area for future collaborations. The respondents were also asked to estimate to what extent[3] they had documented their processes in the areas of procurement, production, and distribution internally at their company. Also the documentation of their whole internal process stretching through the entire company was asked for. The mean values (standard deviation in parenthesis) are shown in Figure 3.

Type of measurement Table I. The use of process related measurements in the collaboration Total logistics costs Lead times between the companies Service between the companies Source: Sandberg (2005, p. 71)

Mean 1.97 3.16 3.28

Standard deviation 1.19 1.38 1.35

Procurement 3.86 (1.05)

Production 4.02 (0.95) Whole Company 3.48 (1.07)

Distribution 3.72 (1.07)

Figure 3. The degree of internal process documentation at the focal company

Source: Sandberg (2005)

As can be seen in the gure, the respondents have documented to a relatively high extent their internal processes in all parts of their companies. This can be interpreted as a consequence of the widely spread adoption of ISO 9000 certications. Planning of logistics activities A process approach to collaboration puts a natural focus on the coordination and integration of the activities involved (Melan, 1993; Willoch, 1994). Aronsson (2000) suggests that the question of organisational responsibility for the different activities should be considered later when the process already is optimized. From this it follows that the division of the organizational planning responsibility between companies could be changed from a traditional view towards other, better integrated, solutions. Dening and describing a process could make these possible options more visible for the supply chain members. Examples of how the planning responsibilities can change and be treated in different ways can be seen in collaborative-based business concepts such as efcient consumer response (ECR), vendor managed inventory (VMI), and CPFR. Inuenced by the SCM philosophy, these three concepts aim to improve inventory management and other logistics issues in the supply chain. In the case of ECR, the participating actors plan the collaboration design jointly. However, at the operating level no change in planning responsibility for logistics activities may be seen. Instead when considering VMI, this change of responsibility can be seen clearly since the main idea behind VMI is the recognition that the vendor is better positioned in the supply chain to plan and manage the replenishment of the customers warehouse. Finally, when considering CPFR, the participating actors do the planning of some activities jointly. Thus, the three concepts represent three different opportunities for handling the planning responsibility of activities involved in the collaboration (Sandberg, 2005). In the questionnaire, the means by which supply chain activities were planned was categorized in the way as shown in Figure 4[4].
Do not agree 1 2 3 4 Completely agree 5 Mean

Logistics collaboration in supply chains 279

Joint operational planning N=157

3.31

Delegated planning N=150

3.00

Joint strategic planning N=142 0% 20% 40% 60% 80%

2.20

100%

Source: Sandberg (2005, p. 73)

Figure 4. How planning of logistics activities is performed in the collaborative arrangement

IJLM 18,2

280

Figure 4 indicates that the most common way to plan logistics activities within a collaborative arrangement is joint planning on an operational level. This can also be considered as the least advanced and less demanding type of planning. The degree of joint strategic planning can be considered as low. Of the respondents, 66 percent answered one or two to the question concerning the extent to which joint strategic planning of supply chain activities is performed. Interestingly, the degree of joint operational planning is not positively related to the degree of joint strategic planning and vice versa[5]. Thus, more strategic planning does not seem to facilitate or lead to more operational planning or vice versa. Of the respondents, 46 percent have answered four or ve to the planning type where one actor plans for all parties including himself. This relatively high percentage is probably due to the fact that many of the collaborations are inspired by the idea of VMI. Even if not asked for specically in the questionnaire, many of the respondents have answered on follow up questions that their collaboration is a typical VMI-collaboration, etc. Information sharing Information sharing among the supply chain members is an important prerequisite for collaboration (Lee and Whang, 2000; Yu et al., 2001; Xu and Dong, 2004) and has a great impact on the performance in the supply chain (Barratt, 2004). This study shows that the type of information shared within collaborations and the frequency of sharing differ considerably. Table II shows mean values (in days) of the sharing frequency of a certain type of information[6]. The type of information that is shared by most respondents at least once a month is forecasts, which is shared by 94 percent of the respondents. Forecasts are relatively easy to share and are also reported as the most commonly shared type of information in other survey-based studies, see, e.g. Selldin (2002). Of the respondents, 58.8 percent share at least two types of information at least once a week, which in this study has been considered as the lower limit to describe the information sharing as intensive. For information sharing to be successful is not only a question about sharing the right information with a sufcient frequency, but also how
Percentage of respondents sharing the type once a month or Mean (in more often days) 71 76 94 43 81 37 22 79 66 9.09 8.97 12.23 13.37 8.67 16.76 18.17 11.84 6.85

Information type Production planning Inventory levels Forecasts Sales information (point of sales data) Error messages, etc. Product campaigns Price levels and pricing Future deliveries, etc. Conrmations, track and trace, etc.

Number of respondents 111 118 149 68 126 58 35 124 104

Standard deviation 7.37 7.52 7.49 8.09 7.16 6.09 4.91 7.75 6.20

Table II. The mean values for frequency of sharing of different types of information

Source: Sandberg (2005, p. 74)

the transferred information is used and implemented by the receiving company (Lee and Whang, 2000). An important factor to facilitate the implementation and usage of the information is to share processed data instead of non-processed, i.e. share data that is more specically developed for the receivers needs. Such data is often more valuable and will have a greater impact on planning efciency and performance in the supply chain (Xu and Dong, 2004). The degree of adjustment for the specic receiver of the shared information was investigated with a ve-point Likert scale, ranging from the information is not adjusted/processed to the information is very adjusted/processed. The mean value (standard deviation in parenthesis) for the answers was 3.66 (1.12). This indicates a high degree of adjustment of the information that is shared, which in turn improves the possibilities for the receiver to use the information properly. The development within the IT and technology sector over the last decades has had a great impact on information sharing in supply chains and is seen as an enabler for more advanced collaboration. Despite the strong support in the literature, only 50.6 percent of the respondents consider EDI and/or internet-based EDI to be the main means of communication in the collaboration. Cross tabulation and x 2 test shows that respondents with EDI or internet-based EDI in their collaboration also belong to the 58.8 percent of the respondents that share at least two types of information at least once a week. This result was expected due to the better possibilities to share information in a more efcient manner with EDI and other internet-based alternatives. The use of more advanced communication methods does not however seem to inuence the extent to which the shared information is processed and adjusted for the receiver. The results imply that the use of EDI and internet-based alternatives facilitates the transferring of information, but not the content of the information, i.e. to what extent the information is processed and adjusted for the receiver. This is also supported when considering what types of information are shared signicantly more frequently in collaborations with EDI or internet-based EDI. Above all, it is operational information types such as error messages and inventory levels that are shared more frequently by EDI users. The need for adjustments of these types of information is probably lower than for instance, forecasts and other information types that are more difcult for the receiver to interpret. Supply chain orientation SCM and collaboration can encompass a broad range of activities for companies in a supply chain. However, apart from suggesting what the actors actually should do, most authors also stress (even if they seldom discuss it more extensively) the importance of undertaking the actions with the right intentions, referring to trust, win-win thinking, common goals, etc. Mentzer et al. (2001) refer to this management philosophy as supply chain orientation (SCO), and dene it as the recognition by an organisation of the systemic, strategic implications of the tactical activities involved in managing the various ows in a supply chain (Mentzer et al., 2001, p. 11). Experienced problems with these soft variables such as trust, organizational compatibility, commitment, vision, key processes, etc. were in the questionnaire investigated with statements on a ve-point Likert scale ranging from do not agree to completely agree. In general, the answers indicate a low degree of problems, which

Logistics collaboration in supply chains 281

IJLM 18,2

282

means a good SCO. The respondents also felt that they had good possibilities to inuence the design of the collaboration relatively to the other participants. Another important prerequisite for a SCO is top management support (Mentzer et al., 2001). This was investigated with two statements about the managements involvement during the initial phase of the collaboration and the involvement during the ongoing collaboration. The mean values[7] and distribution of respondents can be seen in Figure 5. Logistics activities The survey also explored the different logistics activities in which collaboration took place. The understanding of the functional scope of SCM today is very broad and covers all the traditional intra-business functions within the company (Mentzer et al., 2001). From a logistics perspective this means a broad range of areas, where numerous collaboration possibilities exist. The respondents were asked to specify on a ve-point Likert scale[8] to what extent their collaboration was performed in the logistics areas shown in the Table III. As can be expected, strategic planning is the least common area to collaborate in. Several analyses were performed to investigate if there were any differences in actions undertaken in the collaboration depending on in which logistics area the collaboration took place. Without success, factor analysis as well as ANOVA analysis was used to nd possible correlations between the actions undertaken and certain logistics areas.
Do not agree 1 Top management involvement in initial phase N=157 2 3 4 Completely agree 5 Mean

3.31

Figure 5. The degree of top management support from the focal company in the collaboration

Top management involvement in ongoing collaboration N=158 0% Source: Sandberg (2005, p. 77) 20% 40% 60% 80% 100%

3.08

Logistics area Table III. Mean values investigating to what extent the collaboration is performed in a certain logistics area Production planning Forecasting Inventory management and replenishment Transportation planning Strategic planning Source: Sandberg (2005, p. 78)

Mean 2.77 3.88 3.44 3.43 2.25

Standard deviation 1.29 1.07 1.27 1.24 1.19

The result implies that the logistics area in which collaboration is performed, and the actions undertaken are independent of each other, i.e. no specic actions are undertaken in specic logistics areas. Driving forces From a theoretical point of view, collaboration among supply chain members results in many positive effects. On a general level lowered total costs, improved service and shorter lead times are often mentioned. Also more intangible effects, such as the wish to strengthen the companys market position and increase its competitiveness, can be seen as driving forces. In the questionnaire the driving forces for the collaboration were categorized and measured by two main factors: cost and service. Figure 6 shows the mean values and distribution of the respondents for the results[9]. The results indicate that both cost and service are considered as important reasons for setting up collaboration. Of the respondents, 72.6 percent have answered four or ve to the question of how important cost related factors were. The same gure for service related factors was 95.3 percent. The ndings are in line with previous survey-based research made by Spekman et al. (1998), who found that the most important reasons to engage in SCM collaboration could be found both on issues related to cost reduction as well as service. The most important reasons for this, according to their study, were increased end-customer satisfaction, improved prots, satisfy supplier/customer requests and reduction of overall operating costs. Barriers As mentioned in the introduction, research indicates that surprisingly little collaboration among companies can be seen in supply chains. Despite the many obvious advantages, collaboration is still something unusual. Furthermore, existing collaborations are faced with shifting results of their collaborative efforts (Smaros, 2003). A reason for the absence of logistics collaboration and positive effects could be that barriers for collaboration have not been successfully tackled. Two main categories of barriers can be identied in SCM literature; those related to technology and those related to human beings (Sandberg, 2005). As described above, the respondents had a rather good SCO and experienced a low degree of problems related to trust, personal chemistry, etc. When it comes to technology related problems, this variable had the second highest mean value after the variable
Do not agree 1 Cost related factors N=146 2 3 4 Completely agree 5 Mean

Logistics collaboration in supply chains 283

3.87

Service related factors N=151 0% Source: Sandberg (2005, p. 83) 20% 40% 60% 80% 100%

4.43

Figure 6. Mean values for variables measuring the driving forces for the collaboration

IJLM 18,2

investigating problems with different logistics competences between the participating actors. The results are in accordance with previous research (Hoffman and Mehra, 2000), and show that technology related problems still occur and cannot be ignored. Effects Nine statements on a ve-point Likert scale ranging from disagree to completely agree investigated to what extent the respondents had experienced specic effects of the collaboration. Figure 7 shows mean values and distribution of answers for the nine statements. As can be seen in the gure, more positive effects are in general experienced on service related issues than cost related ones, even if differences are small. Apart from positive effects directly connected to either cost reduction or service improvements, this study also shows the more intangible effects of collaboration, i.e. increased competitiveness, clearer division of responsibility between partners, and more measurements and follow ups. With factor analysis it was concluded that these variables were correlated to the cost effects as well as the service effects.
Do not agree 1 Respondent has achieved lower costs N=150 Partner has achieved lower costs N=138 Respondents service towards partner has been improved N=155 Partners service towards respondent has been improved N=152 Respondents service towards other actors has been improved N=154 Improved lead times between companies N=156 Respondent has been more competitive N=146 Clearer division of responsibility between partners N=149 2 3 4 Completely agree 5 Mean 3.55

284

3.59

4.11

3.88

3.72

3.88

3.82

3.52

Figure 7. Mean values of variables investigating experienced effects of the collaboration

More measurements and follow ups N=153 0% Source: Sandberg (2005, p. 85) 20% 40% 60% 80% 100%

3.41

In cases with dyadic collaborations, this study also investigated the impact the collaboration had on relations with other companies on the other side of the respondents company. Respondents with a supplier collaboration were asked questions about its impact on relations with their customers and vice versa. These effects are shown in Figure 8. From the gure it can be concluded that improved service towards other members of the supply chain is the most experienced effect. This is valid for both types of dyadic collaborations, i.e. independent of whether the collaboration is upstream or downstream in the supply chain. Three major contributions Below some of the most interesting results on a more aggregated level are summarized into three major contributions. First, there is a clear relationship between the intensity of the collaboration and the positive effects experienced from the collaboration. Second, the results indicate that top management is an important driver for a higher intensity of the collaboration. Third, there are major differences between SCM theory and practice. Intensity drives positive effects The ndings from the survey show that: . there is a strong relationship between the process orientation, the degree of joint operational planning, and intensity of information sharing; and . that the intensity of these aspects are positively related to the experienced effects of the collaboration. Several analyses showed that the respondents degree of process orientation, intensity of information sharing and degree of joint planning are closely related to each other. On a general level it can be stated that more of one thing also means more of the other two, indicating a positive spiral of intensity (Sandberg, 2005). The results are in accordance with SCM literature, i.e. having a clear process approach can facilitate and be the starting point for the participating actors to increase
Do not agree 1 Increased collaboration N=109 2 3 4 Completely agree 5 Mean

Logistics collaboration in supply chains 285

3.09

Increased and faster information sharing N=109 Improved service by the respondent towards other actors N=112 0% Source: Sandberg (2005, p. 87) 20% 40% 60% 80% 100%

3.13

3.63

Figure 8. Experienced effects on relations with actors on the other side of the respondents company

IJLM 18,2

286

information sharing and the degree of joint planning since the advantages will be more obvious. However, this study only investigates statistical relations and what comes rst of the three aspects in the collaboration is therefore not investigated. Most probably, it would not be possible to make a strictly causal order of the aspects. For example, as stated above a collaboration could start with a clear process approach and cause more intensive information sharing due to an increased understanding of the need for sharing more information. The opposite order is however also possible; an intensive information sharing could force the actors involved to specify their collaboration in terms of a process in order to be able to use the shared information in a more efcient way. Furthermore, a need for an intensive information sharing could justify implementation of EDI between the partners, which in turn could demand a process denition to enable its proper use. In the following section the effects experienced and the variables measuring process orientation, information sharing, and joint planning are discussed. Process orientation and effects. Analysis indicated a relationship between the degree of process approach adopted and the experienced effects of the collaboration. To start with, respondents with a dened process for their collaboration experience a signicantly clearer division of responsibilities between themselves and their partner/partners. As discussed above, the description of a process will increase the awareness of the activities involved and contribute to a better understanding of the collaboration. Thus, a clearer division of responsibility between the participating actors can be a consequence of a clear process approach. It can also, together with proper measurements, be an important driver for improvements in the collaboration. A clear description of responsibility facilitates and encourages actions for improvements. Respondents with a dened process are also likely to measure the effects of the collaboration more, both in terms of costs and lead times as well as service. The use of measurements is in turn positively related to the effects experienced in the collaboration; respondents who measure more also experience better effects of what they measure. Information sharing. The many promising effects of information sharing in supply chains are another fundamental feature of the SCM literature. On a general level this can be conrmed in this study; respondents reporting more positive effects also share information more frequently. However, due to differences in intensity of information sharing among the types of collaboration[10] and a too low-response rate, the results cannot be considered as independent from the type of collaboration that is performed. Joint planning. The explanation as to why a higher rate of joint planning improves the effects of collaboration can also be explained by reduction of uncertainty. Joint planning can be considered as a kind of information sharing since it demands communication between the actors. For example, in the SCM-based concept of CPFR, this communication with a constant negotiation connects and integrates the actors with each other, which in turn leads to problem solving and improved effects (Barratt and Oliveira, 2001). This study can conrm the relationship between a high level of joint operational planning and more positive effects from the collaboration. It is above all service related effects, lead time improvements, and other more intangible effects that are improved by joint operational planning. The expected positive relation between a high degree of joint strategic planning of supply chain activities and better effects cannot be conrmed by this study. The lack of a relationship could be due to the fact that the

study also shows a weak relationship between a higher level of strategic planning and increased intensity in the collaboration (concerning actions undertaken in the collaboration). Thus, a reason for the lack of relationship could be that it is the intensity that causes the positive effects, and a high degree of joint strategic planning does not contribute to this. The process approach and intensity in information sharing and joint operational planning is however not easily achieved and is not only related to positive effects of the collaboration. In some cases it also leads to more problems experienced in the collaboration. For example, respondents with a dened process experience more problems with the variable different goals of the participating actors. This should however not be interpreted as if documentation of processes makes the collaboration more difcult. Instead, this result should be interpreted as respondents who work with processes in their collaborations become more aware of the problems and thus have a better possibility to deal with them and minimize them. Actors collaborating without processes may not even be aware of some of the problems. Furthermore, respondents with more intensive information sharing (i.e. sharing at least two types of information with their partner at least once a week) also experience signicantly more technology related problems. The reason for this could be that these respondents also use EDI and internet-based EDI more often (this relation was discussed in chapter 5.2.3 above). However, when considering the mean values[11] between respondents with more complicated technology (respondents using EDI and/or internet-based alternatives) and respondents using traditional means of communication (telephone, e-mail, etc.) no such signicant difference can be found. Thus, technology related problems are experienced to the same extent by those respondents who only use traditional communication means. Top management is an important driver Several analyses indicate that the involvement of top management in the ongoing collaboration seems to be positively related to the intensity of what is done in the collaboration. Top management is more involved in collaborations that have more frequent information sharing and joint operational planning of the supply chain activities. A reason for this could be that involvement by top management gives the logistics department the authority to carry out the collaboration and bring it to a more intensive level. The involvement of top management means an increased focus on the collaboration and is regarded as a necessary prerequisite to enable SCM-based collaboration to function (Ireland and Bruce, 2000). For the same reason a higher degree of involvement from top management should also increase the degree of joint strategic planning between the actors. No such relation can however be found in this study; a higher degree of top management involvement in the ongoing collaboration is not positively related to the degree of joint strategic planning of supply chain activities. A reason for this result could be that, it is the intensity of the collaboration on an operational level that causes the positive effects of the collaboration, rather than, if the collaboration includes joint planning on a strategic level. The results can be interpreted as top management engaging in operational issues since these are connected to the intensity of the collaboration. It is the intensity that in turn causes the positive effects of the collaboration, rather than joint planning on a strategic level.

Logistics collaboration in supply chains 287

IJLM 18,2

288

This study also shows that top management involvement in collaboration with a supplier is important for increasing collaborative efforts on the other side of the company, i.e. the customer side. This relationship is also valid for the opposite situation, i.e. the more top management attention in customer collaboration, the more collaborative efforts can be found on the focal companys supplier side. It could be argued that a high degree of commitment in the collaboration makes the top management aware of the possibilities for the other side of the company. Thus, thanks to the top management involvement in the dyadic collaboration on one side of the company, relations on the other side of the company could be positively inuenced. Andraski (1998) claims that external as well as internal collaboration will only become reality if driven by effective leadership (Andraski, 1998, p. 11). Top management must understand what the collaboration means (Ireland and Bruce, 2000), and have an active role and be involved in the collaboration. Note that the word driven used by Andraski above indicates that support is not enough. In this study, the word involvement has been used in order to distinguish from top management support. SCM in theory and practice differs Based on the survey results, it can be concluded that the respondents collaborations to a high extent are concerned with operational issues and that the degree of strategic activity is very low. This result conrms previous research (Spekman et al., 1998) that SCM as it is discussed in conceptual articles is a rare occurrence. Collaboration based only on operational issues cannot be regarded as completely integrated SCM; it is only when strategic issues are involved that a shift towards real SCM-based collaboration can be possible. The low level of strategic issues in the collaboration could also be the reason why the respondents SCO seems to be rather good among the respondents and independent from the effects experienced. The operational content in existing collaborations does not challenge the respondents SCO since such collaboration can be managed without demanding a change of mindset among the participating actors. Except for the absence of strategic issues in the collaboration, the results also indicates serious differences in attitude and behavior depending on whether the collaboration partner is a supplier or customer, despite the fact that no differences concerning background variables such as, e.g. company size and characteristics of the products can be found. The most important differences are shown in Table IV.
Supplier collaborations More inuence on the shape of the collaboration Less problems with the sharing of costs and rewards between the parties Involves more strategic planning Customer collaborations More top management involvement More intensive information sharing and use of EDI and internet-based alternatives More technology related problems More improvements in service towards collaboration partner Moe increased competitiveness

Table IV. Differences between respondents with supplier and customer collaborations

Source: Sandberg (2005)

In comparison to supplier collaborations, customer collaborations seem to be more important for the focal company and are in general more advanced. They are getting more attention from top management and contain more intensive information sharing. Related to this, (probably as a consequence thereof) EDI and internet-based communication means are used to a higher extent. This is also mirrored in the barriers experienced in the collaboration; respondents collaborating with their customers also report signicantly more technology related problems. Despite the lower top management involvement, respondents with a supplier collaboration consider their collaboration to involve more strategic planning than their colleagues with customer collaborations (even if the rate is still low). Note that, this is the respondents own opinion and that the strategic planning is made within their own company; no differences can be found when considering the degree of joint strategic planning in the collaborations. This could be due to the fact that these respondents also feel that they have signicantly more inuence on the design of their collaborations. The more the company can inuence and manage the collaboration, the more strategic it becomes for the company. When considering the effects of the collaboration the difference between upstream and downstream relations becomes even clearer. Respondents with customer collaborations experience to a signicantly higher extent that their service towards their partner has been improved and that they have been more competitive as a consequence of their collaboration. These results are in line with Mattsson (2002), who discusses differences in attitudes between suppliers and customers. The differences can be explained by the fact that respondents with customer collaborations often nd themselves in a weaker negotiation position than their colleagues with supplier collaborations. They feel that they need to struggle to keep their customers and therefore the service towards their customers becomes important. The collaboration can also increase their competitiveness due to the closer ties with the customer, which in turn strengthens the opportunities to keep the customer (Mattsson, 2002). According to the conceptual SCM literature the differences shown in Table IV should not exist. SCM philosophy advocates a focus on end customer demand and that the whole supply chain should be managed as one single entity. Differences in behavior, depending on if the partner is a supplier or a customer, tend to be out of the scope for most SCM literature. This pattern can have serious consequences and risk the opportunity for creating a true SCO among the actors in the supply chain. Discussion The description of logistics collaboration given in this paper shows that it is a more intensive collaboration on an operational level that contributes to the achievement of better results, and that top management involvement is an important driver for: . increasing the intensity of the existing collaboration; and . increase collaborative efforts with other members of the supply chain. The study also shows that the companies are still concerned with operational issues and that their collaborations are seldom brought to a strategic level. In addition to this, the results indicate that there are serious differences in attitude and behavior between supplier and customer collaborations. The results are shown in Figure 9.

Logistics collaboration in supply chains 289

IJLM 18,2

Supplier

Focal Company

Customer

Top management is an important driver

290

Collaboration

Collaboration

There are differences between customer and supplier collaborations

Figure 9. The main results from the study

The more intensive collaboration, the better effects are experienced Source: Based on Sandberg (2005)

In their study, Spekman et al. (1998) concluded the following:

In summary, we have implied that business has yet to crack the code; supply chain partners still do not share a common vision or react to the same set of metrics. If this is true, opportunities have been lost and many challenges remain. For a number of rms, talk is cheap and supply chain management is still only part of todays jargon. A number of rms are sacricing cost effectiveness, revenue enhancement, and customer satisfaction because they are unable to work effectively across the rms that comprise their supply chains (Spekman et al., 1998, p. 648).

Related to the results presented in this study, the conclusion made by Spekman et al. (1998) eight years ago is still valid. Indeed, SCM as described in theory still seems to be more of an utopia than reality, and many opportunities for cost as well as service improvements have been lost. Compared to theory, the most striking difference is the absence of a strategic dimension to the collaborations, something that is seen as a prerequisite for success by most SCM theorists (Cooper and Ellram, 1993; Lambert and Cooper, 2000; Mentzer et al., 2001). Even if collaboration on an operational level is good, the logic behind the SCM theory says that including a strategic level will have an even greater positive impact on the supply chain performance. Thus, it could be questioned whether all the promised supply chain improvements will be realized unless the collaboration is taken to a strategic level. This does not however mean that the operational level of the collaboration should be ignored or underestimated. Successful collaboration means working on both levels (Moss Kanter, 1994), and the results from this study clearly conrms that the operational collaboration that is performed is positive in terms of effects achieved. The results imply that one of the main questions of concern for future research becomes how strategic elements can be brought into the collaboration. Even though

some work on how strategic partnership in the supply chain should be created and managed exist (Christopher and Juttner, 2000), more research in this area has to be done. Referring to the results of this study, top management involvement is an important driver for successful implementation of SCM. This issue has been discussed and stressed by many authors within the SCM eld (Andraski, 1998; Ireland and Bruce, 2000) and in strategic alliance literature (Moss Kanter, 1994). However, despite its importance, the role of top management has not been investigated in-depth. Statements like: top management support is a prerequisite for successful SCM-based collaboration are often made, but this is not enough anymore. A more detailed discussion about the role of top management would help to better understand what it takes to achieve strategic collaboration. This could be one of the most important pieces in the big SCM puzzle that is not yet in place.
Notes 1. The dened population in this study is manufacturing companies (companies with SNI codes starting with the letter D). 2. One-sample t-test shows that the values for total logistics costs and service between the companies are signicant separated from 3. 3. Measured on a ve-point Likert scale ranging from not at all to to a high degree. 4. The questions were put as a statement as to what extent the collaboration had changed the planning of supply chain activities. The answers were given on a ve-point Likert scale ranging from do not agree to completely agree. 5. Investigated with ANOVA analysis. 6. Note that the mean values in days are only based on those respondents that share the certain type of information at least once a month (e.g. 111 respondents, or 71 percent, share information about production planning at least once a month and the average number of days between sharing for these respondents is 9.09 days). 7. The answers were given on a ve-point Likert scale ranging from do not agree to completely agree which means a high involvement for high values. 8. Ranging from The collaboration is not performed in this area. to The collaboration is to a great extent performed in this area. 9. The questions were put as a statement and the answers were given on a ve-point Likert scale ranging from do not agree to completely agree. 10. Respondents with customer and triadic collaborations share signicantly more frequently information with their partner/partners. 11. Investigated with ANOVA analysis. References Andraski, J.C. (1998), Leadership and the realization of supply chain collaboration, Journal of Business Logistics, Vol. 19 No. 2, pp. 9-11. Aronsson, H. (2000), Three perspectives on supply chain design, Doctoral dissertation, No. 44, Department of Management and Economics, Linkoping University, Linkoping. Aronsson, H. (2003), Logistikindikatorn Del 2 En uppfoljning av tillstandet och forandringarna av svenska foretags logistic, IMIE working paper series, Division of Management and Economics, Linkoping.

Logistics collaboration in supply chains 291

IJLM 18,2

292

Barratt, M. (2004), Understanding the meaning of collaboration in the supply chain, Supply Chain Management: An International Journal, Vol. 91, pp. 30-42. Barratt, M. and Oliveira, A. (2001), Exploring the experiences of collaborative planning initiatives, International Journal of Physical Distribution & Logistics Management, Vol. 31 No. 4. Bechtel, C. and Jayaram, J. (1997), Supply chain management: a strategic perspective, The International Journal of Logistics Management, Vol. 81. Christopher, M. (1998), Logistics and Supply Chain Management, Pitman Publishing, London. Christopher, M. and Juttner, U. (2000), Developing strategic partnerships in the supply chain: a practitioner perspective, European Journal of Purchasing & Supply Management, Vol. 6, pp. 117-27. Cooper, M. and Ellram, L. (1993), Characteristics of supply chain management and the implications for purchasing and logistics strategy, The International Journal of Logistics Management, Vol. 42. Forslund, H. (2004), The existence of logistics quality deciences and the impact of information quality in the dyadic order fullment process, Doctoral dissertation, No. 62, Department of Management and Economics, Linkoping University, Linkoping. Forza, C. (2002), Survey research in operations management: a process based perspective, International Journal of Operations & Production Management, Vol. 22 No. 2, pp. 152-94. Hoffman, J.M. and Mehra, S. (2000), Efcient consumer response as a supply chain strategy for grocery businesses, International Journal of Service Industry Management, Vol. 114. Ireland, R. and Bruce, R. (2000), CPFR only the beginning of collaboration, Supply Chain Management Review, September/October. Lambert, D.M. and Cooper, M.C. (2000), Issues in supply chain management, Industrial Marketing Management, Vol. 29. Lee, H.L. and Whang, S. (2000), Information sharing in a supply chain, International Journal of Technology Management, Vol. 20 Nos 3/4, pp. 373-87. Malhotra, M.K. and Grover, V. (1998), An assessment of survey research in POM: from constructs to theory, Journal of Operations Management, Vol. 16. Mattsson, S-A. (2002), Logistik i forsorjningskedjor, Studentlitteratur, Lund. Melan, E.H. (1993), Process Management Methods for Improving Products and Service, McGraw Hill Inc, New York, NY. Mentzer, J.T., DeWitt, W., Keebler, J.S., Min, S., Nix, N.W., Smith, C.D. and Zacharia, Z.G. (2001), Dening supply chain management, Journal of Business Logistics, Vol. 22 No. 2. Moss Kanter, R. (1994), Collaborative advantage: the art of alliances, Harvard Business Review, Vol. 72 No. 4. Pinsonneault, A. and Kraemer, K. (1993), Survey research methodology in management information systems: an assessment, Journal of Management Information Systems, Vol. 10 No. 2, pp. 75-105. Sandberg, E. (2005), Logistics collaboration in supply chains a survey of Swedish manufacturing companies, Licentiate thesis, No. 1180, Department of Management and Economics, Linkoping University, Linkoping. Selldin, E. (2002), Enterprise and supply chain planning: a survey of Swedish manufacturing rms, Licentiate thesis, No. 966, Department of Production Economics, Linkoping University, Linkoping.

 Smaros, J. (2003), Collaborative forecasting: a selection of practical approaches, International Journal of Logistics: Research and Applications, Vol. 6 No. 4. Spekman, R.E., Kamauff, J.W. and Myhr, N. (1998), An empirical investigation into supply chain management: a perspective on partnerships, International Journal of Physical Distribution & Logistics Management, Vol. 28 No. 8. Stank, T.P., Keller, S.B. and Daugherty, P.J. (2001), Supply chain collaboration and logistical service performance, Journal of Business Logistics, Vol. 22 No. 1. Willoch, B.E. (1994), Business Process Reengineering, Fagbokforlaget, Bergen. Xu, K. and Dong, Y. (2004), Information gaming in demand collaboration and supply performance, Journal of Business Logistics, Vol. 25 No. 1. Yu, Z., Yan, H. and Cheng, T.C. (2001), Benets of information sharing with supply chain partnerships, Industrial Management & Data Systems, Vol. 101 No. 3, pp. 114-9. About the author Erik Sandberg is a PhD candidate in Logistics Management at Linkoping University. His research includes logistics collaboration and the role of top management in a companys supply chain management practices. Erik Sandberg can be contacted at: erik.sandberg@liu.se

Logistics collaboration in supply chains 293

To purchase reprints of this article please e-mail: reprints@emeraldinsight.com Or visit our web site for further details: www.emeraldinsight.com/reprints