Assignment- Doctorate Level- Please READ Entire Post First!!!

R e s e aRc h a Rt i c l e Abstract:  0 this study provides insights into how manufacturers adopt their Global account Management (GAM) activities in response to the increasing expansion of retailers. Specifically, we focus on the manufacturers’ central coordination of two types of GaM activities: strategic and tactical activities. 0 We analyse the manufacturers’ associations with international retailers and with GaM ef- fectiveness and efficiency by using data from 172 manufacturers. Moreover, we consider the suppliers’ dependence on their key retail accounts to be an important moderator within the consumer goods sector.

0 In particular, manufacturers respond to the centralisation of the retailers’ purchasing activities by centralising strategic GAM activities, such as customer strategy, information processing, or price systems. Additionally, manufacturers respond even stronger by centralising tactical activities, such as category management, marketing, or logistics. Although the centralisation of strategic activities drives GAM effectiveness and efficiency, the centralisation of tactical activities does not. This finding might be explained by the specific context of manufacturer- retailer relations. Finally, we find that, although the decision to centralise GAM activities pays off, the benefits are contingent on the particular type of GAM activity and the level of customer dependency.

Keywords:  Retail internationalisation · GaM activities · centralisation · effectiveness · Efficiency Manag Int Rev (2012) 52:727–756 DOI 10.1007/s11575-011-0118-8 Does Centralising Global  Account Management  Activities in Response to International Retailers  Pay Off?

Bernhard Swoboda ·  Andrea Schlüter · Edith Olejnik · Dirk Morschett Received: 15.04.2010 / Revised:  18.10.2010 / Accepted:  02.11.2011 / Published  online: 04.01.2012 © Gabler-Verlag 2011 Prof. B. swoboda ( ) University of Trier, Trier, Germany e-mail: [email protected] Dr . a. schlüter · PhD. cand. e. Olejnik University of Trier, Trier, Germany Prof.

D. Morschett University of Fribourg, Fribourg, Switzerland 728B. swoboda et al.

Introduction Today, leading retail firms conduct up to 70% of their sales abroad (see Swoboda et al.

2009). t he foreign expansion of retailers affects the relationship between retailers and suppliers, particularly with regard to how suppliers coordinate their key account manage - ment (KAM). Scholars define KAM as a strategic approach focused on customer prioriti - sation to obtain higher customer profitability and higher returns on sales (Homburg et al.

2008). In this study, we focus on global account management (GAM). GAM differs from KAM in that GAM considers specific challenges to coordinating customer relations across countries (Gosselin and Bauwen 2006). Prior scholars define GAM as “an organisational form and process in multinational companies by which the worldwide activities serving a given multinational customer are coordinated centrally by one person or team within the supplying company” (Montgomery and Yip 1999, p. 10). The organisational form of the concept appears to be straightforward. it provides customers with one point of contact.

However, for suppliers, the process of implementing centrally coordinated activities is frequently problematic (Birkinshaw et al. 2001, p. 232). Why is central coordination a critical issue? On the one hand, centralising customer-related activities (i.e., coordina - tion at the headquarters level and not at the country level; see hennart 1991) might be a superior way of dealing with globalising customers. On the other hand, GAM programs are often established alongside existing local sales organisations, and, more importantly, the retailers who manage complex store networks demand support at different levels (e.g., at the headquarters level and at the county/store level) (Verbeke et al. 2006). Hence, to analyse how consumer goods manufacturers organise their GaM activities in response to retailer expansion, we focus on the central coordination of two types of GAM activities:

strategic and tactical activities. We further propose that strategic and tactical activities have different impacts on GAM effectiveness and efficiency. Although scholars often study KAM (e.g., a recent review is provided by Richards and Jones 2009), the literature rarely focuses on GAM. The relevance of GAM is undis- puted within the literature, but only six studies empirically investigate GAM coordination issues. t hree of these studies address the issue of centralisation (Birkinshaw et al. 2001; t oulan et al. 2007; s hi et al. 2010) but do not analyse it in further detail. For example, Shi et al. ( 2010) find that the extent of centrally coordinated marketing activities drives GAM performance, and Birkinshaw et al. ( 2001) show that centralisation is related to GaM performance. Thus, two research gaps exist in the literature. There is a lack of empirical findings on the centralisation of different GAM activities and its association with GAM effectiveness and efficiency. Recent studies usually analyse performance in terms of effectiveness or efficiency. On the one hand, GAM may be more effective if manufacturers strengthen their market positions by centralising their strategic GAM activities. On the other hand, GAM may be more efficient if the manufacturers decentralise some of their GAM activities to enhance the local flexibility towards the retailers’ store networks. Both scenarios are relevant to studies on GaM and should be analysed in more detail with regard to their effects on the effectiveness and efficiency of G aM.Prior studies predominantly survey firms without differentiating the firms according to sector. In contrast, we argue that the consumer goods sector is distinct and that its 729 Does centralising Global account Management activities … specific characteristics have not been properly addressed to date. Specifically, this sector is distinctive because of its two-tier distribution structure (Geylani et al. 2007), where retailers play an important filtering role by deciding which products from which suppliers will be offered to the consumers. More importantly, each supplier depends on retailers to a different extent and may respond internally by strategically coordinating its cus- tomer -related activities. although retailers and suppliers frequently collaborate in cat- egory management and logistics within the domestic context (corsten and Kumar 2005), they often experience conflict while discussing prices. For these reasons, we analyse the decision to centralise strategic vs. tactical GaM activities and the suppliers’ dependence on retailers. In sum, this study aims to analyse the following research questions. To what extent do manufacturers centralise their GaM activities in response to the international retailers’ activities? How does the centralisation of strategic and tactical GAM activities influence GAM performance in terms of effectiveness and efficiency? Finally, how does the suppli - ers’ dependence on their key accounts affect these relationships? While investigating these questions, this study contributes to the International Business literature, especially with respect to international retailer-supplier relationships. From a theoretical perspective, we respond to Birkinshaw et al.’s ( 2001) call to investigate cen- tralisation as an important aspect of GaM organisation. We also follow shi et al. (2010) and s hi et al. ( 2004) by analysing the effects of facilitating conditions and the antecedents of GAM. Finally, we answer Shi et al.’s ( 2010) call for sector-specific studies on GAM.

t his study provides a detailed perspective on how manufacturers coordinate two types of GAM activities. The findings of this study might help reduce the concerns of consumer goods managers. as many managers believe that GaM induces global retailers to exert increased pressure for greater transparency on pricing, these consumer goods managers believe that there are several obstacles to the centralisation of customer-related activities ( harvey et al. 2003).

Literatur e on GAM Coordination Prior research on GaM coordination is sparse (see table 1) and distinct from the research on domestic KAM (e.g., Shi et al. 2010). t he existing studies on domestic KaM view structural coordination mechanisms differently from the studies on GAM and define, for example, centralisation as the integration of certain decision makers or specialists into the decision-making process (e.g., Homburg et al. 2002). a lthough we control for structural coordination mechanisms (e.g., formalisation and specialisation) in this study, we review the literature on GaM coordination by focusing on three aspects: the centralisation of GAM activities, the efficiency/effectiveness of GAM, and the customer-related determin - ants of G aM.

Montgomery and Yip ( 2000) conceptually discuss and Birkinshaw et al. ( 2001) empir- ically demonstrate that when the supplier centralises stronger than the customer it has positive effects on the efficiency of the GAM relationship. Toulan et al. (2007) show that the central coordination of supplier activities has a significant impact on the perform - ance of GAM activities. While analysing GAM strategies, Shi et al. ( 2010) show that 730B. swoboda et al.

author(s) Research question t heoryempirical basis Main findings s hi et al.

(2010) Drivers and outcomes of G aM strategies Global marketing strategy203 question- naires (US, Europe), cross-industry • Global strategic priority and glo- balisation drive G aM strategy • GAM strategy, which includes (1) inter-country coordination (i.e., centralisation of marketing activities) , (2) inter-organisational coordination (i.e., involvement of decision makers), (3) marketing standardisation, and (4) global in - tegration, significantly determines performance (achieving objectives) • Global consumer demand moder- ates the effects of (3) and (4) Krentzel (2009) Organisational challenges of G aM None c onceptual • Discussion of organisational chal- lenges in terms of the coordination requirements of domestic K aM vs.

GaM • appropriate central implementa - tion of strategies/activities is the key to success Yip and Bink (2007) a nalysis of GaM Nonec onceptual • Dimensions: structure is based on the degree of international coordi- nation, and activities are based on the importance of customisation • Determinants: products, customer intent and importance, advantages • Customers selection: size, impor - tance, integration potential, fit t oulan et al. (2007) i mportance of organisational fit to G aM success contingency theory, resource dependence theory 106 question- naires in 16 MNE (UK, s We) across industries • Fit is important from structural (e.g., centralisation, manage- ment support) and strategic (e.g., relevance of the relationship, standardisation) perspectives • If fit cannot be realised, then the manufacturer benefits if the power differential (e.g., stronger centrali- sation, local flexibility) is exploited s hi et al.

(2005) Identification of requirements to or ganising GaM Dynamic capability theory c oncep - tual, but also includes 20 interviews with six firms • Important capabilities (i.e., customer/market knowledge, inter-organisational coordination, and adaptations to environmental changes) are identified • Important outcomes (i.e., turnover and customer profitability) are identified Table 1: Review of the titerature on GaM coordination 731 Does centralising Global account Management activities … author(s) Research question t heoryempirical basis Main findings Wilson and Weil- baker (2004) Development of a comprehensive model analysing the decision to incorporate G aMNone c onceptual • Model is developed based on sales research, studies, and practice • Dimensions: Process drivers, such as definition/selection of key accounts, strategy development, strategic and tactical activities, systems • Determinants: suppliers, buyers, GAM roles, organisational issues • Performance: knowledge, values, profits, customer satisfaction h arvey et al.

(2003) a nalysis of the de - cision to introduce G aM Relational contracting theory c onceptual • challenges to introducing GaM activities include motivations, structures, persons, teams, inter-or - ganisational issues, increased costs and reduced profits • a plan for introducing GaM is proposed a rnold et al. (2001) a nalysis of suc - cess factors of G aM None 35 interviews with 10 compa- nies (UK, s We) across industries • three criteria for successful GaM:

• strategically important to custom - ers and suppliers • internationalised marketing and sales/purchasing • support for management from customers and suppliers Birkin- shaw et al. (2001) a nalysis of suc - cess factors of G aM Resource depend- ence theory, information processing theory106 question- naires in 16 MNE (UK, s We) across industries • GAM efficiency is determined by the centralisation of activities and the dependencies of the customers and support/information systems but not by the scope of the ac- counts and communication ( < 0.10) • Good collaborations with cus - tomers are determined only by customer dependency (communication < 0.10) Mont- gomery and Yip (2000) a nalysis of the de- cision to introduce G aM None 191 mixed data and companies across in- dustries and countries Descriptive data show the following:

• customer globalisation drives not only GaM demand and certain activities (e.g., uniform prices and single point of contact) but also services in host markets • Demand drives the use of GaM activities, such as centralising customer information, controlling strategy, and marketing • GaM performance depends on the coordination of GaM Table 1: (continued) 732B. swoboda et al.

the central coordination of marketing activities, which the researchers refer to as ‘inter- country coordination’, significantly impacts performance, as does the standardisation of marketing activities. However, only Wilson and Weilbaker ( 2004; within a domestic con- text, Zupancic 2008) differentiate the scope of KaM activities into two levels: strategic activities represent the general, fundamental and long-term approach to the collaboration with the customer (e.g., customer-related strategies and information processing), whereas tactical activities are operational arrangements that are conducted on a day-to-day basis (e.g., category management, marketing, or logistics). s cholars conceptualise GaM performance in different ways. Birkinshaw et al. (2001) focus on efficiency and sales growth as well as partnership with the customer (e.g. realis - ing joint projects). toulan et al. ( 2007) consider efficiency and the lessons learned from managing global accounts. shi et al. (2010) examine the achievement of GaM objectives.

We find that global supplier-customer relationships might be linked to different objectives (further goals are shown by Wilson and Weilbaker 2004). However, as Homburg et al.

(2002) indicate, GAM should be effective because suppliers may want to grow with the key account in sales and market share. Additionally, GAM should be efficient because suppliers aim to minimise the costs and efforts needed to handle their relationships with their key accounts. Many studies also extensively discuss the forces driving the adoption of GaM pro- grams by addressing the decision to incorporate a GaM program from a conceptual per- spective. For example, Yip and Bink ( 2007) focus on variables such as customer intent and the importance of suppliers, whereas Yip and Madsen ( 1996) argue that the globali- sation of customers and their central purchasing primarily drive the firms’ decisions to adopt a GaM program. shi et al. (2010) empirically show that the globalisation of cus- tomers and global strategic priority influence firms’ decisions to incorporate GAM strate- gies. We observe that especially international customers’ purchasing and standardised international strategies primarily drive the adoption of GaM. some scholars believe that firms are driven to adopt account programs within a domestic context by their dependence on their key accounts; other scholars believe this adoption to be driven by firms’ suppliers ( ivens and Pardo 2008). Birkinshaw et al. ( 2001) view dependence as an antecedent of G aM performance. author(s) Research question t heoryempirical basis Main findings Yip and Madsen (1996) Drivers of the de- cision to introduce G aM Relationship marketing, contingency theoryFew cases across industries • Drivers: global customers, chan - nels, purchasing • Dimensions of coordination: strat - egy, structure and activities (e.g., one point of contact), consistent services, information Table 1: (continued) 733 Does centralising Global account Management activities … Theoretical Foundation Past scholars have examined GAM from different angles, such as the supplier-customer relationship, resource dependence theory, and, more recently, global marketing strategy.

To address our research questions, we analyse two research streams: 1) studies that con- sider how the corporation adapts to inter-organisational changes in its environment (i.e., the globalising customer) and 2) studies that analyse the effectiveness and efficiency of this process. Thus, we focus on resource dependence theory (RDT) and transaction cost theory (tct). Using both theories broadens the existing arguments on customer relations and relational exchange (Fink et al. 2006), which becomes more uncertain for suppliers because customers will seek mutually beneficial relationships or become more sensitive to prices (Birkinshaw et al. 2001). Accordingly, we employ RDT to determine the effects of retailers’ activities on the centralisation of GaM activities and also to determine their associations with GAM effectiveness. Additionally, we employ TCT to determine the relationship between retailer activities and GAM efficiency. RD t addresses the role of inter-organisational cooperation as a strategic response to changing levels of dependence and uncertainty. Pfeffer and salancik (1978) find that three factors influence the extent to which an organisation depends on another organisa- tion: (1) the latter controls a resource that is important to the survival of the former; (2) the latter has power over the use of that resource; and (3) there are no other sources for that resource. Birkinshaw et al. ( 2001) point out that within GAM, successful firms strive to reduce their reliance on or increase their influence over other firms within the task environments. t hese authors view RDt as a way to systematically examine the dynamics of the bargaining relationship between two organisations. this relationship is especially relevant for studies on GaM because RDt provides insights into how one can balance the power between suppliers and customers. Accordingly, we argue that if a manufacturer moves from a country-by-country relationship with a retailer to an international relation- ship, then the manufacturer will become more dependent on the retailer, who may even control the majority of the manufacturer’s sales (Draganska and Klapper 2007). Thus, RDT states that firms have to respond to and manage these dependencies in the face of uncertainty (heide 1994). One possible solution is to seek closer relationships (Fink et al. 2006). As GAM responds to both dependency and uncertainty, centralising customer- related activities should strengthen the relationship and increase the negotiating power of the supplier with respect to the retailer. Thus, centralising customer-related activities drives GAM effectiveness. However, this argument may depend on the task environ - ment (i.e., the behaviours of the key accounts and the suppliers’ dependence on these accounts). TCT allows one to focus on the relationship between a firm and other organisations.

Although RDT focuses on effectiveness, TCT assumes that all exchanges entail transac - tion costs and focuses on the efficient coordination of inter-organisational cooperation.

From a TCT perspective, actors are rationally bounded and opportunistic, and organi - sations are designed to minimise the cost of organising exchanges and cooperation (Williamson 1991; h ennart 1991). One can utilise tct to systematically examine how organisations reduce their transaction costs. this issue is important to studies on GaM because it provides insights into which coordination mechanisms are the most efficient. 734B. swoboda et al.

Thus, GAM represents a crucial control mechanism in two respects. First, GAM can be viewed as a transaction-specific investment employed to improve the coordination between suppliers and customers (ivens and Pardo 2008). Furthermore, the centralisation of customer activities can reduce the probability of moral hazard because retailers will be less likely to play the manufacturers’ foreign units against each other during the pric- ing negotiations (Verbeke et al. 2006). Reduced moral hazard may also improve GAM efficiency (Jap 2001).

Conceptual  Framework and Hypothesis Development In this section, we examine the hypotheses proposed by this study. Our conceptual model summarises the set of relationships examined in this paper (see Fig. 1). We propose that manufacturers respond to the internationalisation of key retail accounts by centralising the strategic and tactical GAM activities, which, in turn, influences GAM effectiveness and efficiency. Moreover, we argue that the relationships are moderated by the manufac - turers’ dependence on their key accounts because the research on retailer-manufacturer relations shows that this dependence changes the scope of the manufacturers’ activities (Geylani et al. 2007).

In the following, we will discuss the relationships from a theoretical perspective and provide empirical evidence for our assumptions wherever possible.

i nternational Retailer activities and the centralisation of GaM activities Following Montgomery and Yip’s (2000) arguments on global distribution channels and s woboda et al.’s ( 2008) arguments on the dimensions of retail internationalisation, we consider the international retailers’ purchasing, market offers, and process integration as important activities that may drive the decision to centralise GaM activities. in par- Fig. 1: conceptual model ‡ 6 W U D W H J L F * $ 0 D F W L Y L W L H V 7 D F W L F D O * $ 0 D F W L Y L W L H V ‡ & H Q W U D O L V D W L R Q R I L Q W H U Q D W L R Q D O S X U F K D V L Q J D F W L Y L W L H V ‡ 6 W D Q G D U G L V D W L R Q R I L Q W H U Q D W L R Q D O P D U N H W R I I H U V ‡ , Q W H J U D W L R Q R I S U R F H V V H V * $ 0 H I I H F W L Y H Q H V V * $ 0 H I I L F L H Q F \ * $ 0 S H U I R U P D Q F H * $ 0 F H Q W U D O L V D W L R Q , Q W H U Q D W L R Q D O U H W D L O H U V ' H S H Q G H Q F \ R Q N H \ D F F R X Q W ‡ * $ 0 V S H F L D O L V D W L R Q ‡ * $ 0 I R U P D O L V D W L R Q ‡ & R P S D Q \ V L ] H ‡ , Q W H U Q D W L R Q D O H [ S H U L H Q F H 5 H J L R Q D O K H D G T X D U W H U V ‡ ) L U P G X P P \ & R Q W U R O V 735 Does centralising Global account Management activities … ticular, we view the retailers’ centralisation of international purchasing activities to be equivalent to their coordination of agreements on products or orders at the headquarters level. t hese agreements are important for retailers because purchasing costs amount to the lion’s share of total costs, and a high degree of centralisation leads to cost savings ( liu and McGoldrick 1996). Standardisation of international market offers represents the degree of global format transfer (Goldman 2001) and drives the retailers’ expansion costs.

Finally, the retailers’ integration of processes includes the implementation of standards and improvements in information processing, logistics, or category management. The integration of their processes is important in supplier-retailer collaborations (corsten and Kumar 2005). According to RDT, GAM is an organisational response designed to cope with the retailers’ centralisation of international purchasing activities. highly centralised purchas- ing strengthens the international retailers’ negotiating power with suppliers. Thus, suppli - ers that centralise their GaM activities may counteract the retailers’ increased negotiating power and obtain more control over the relationship. Prior studies argued that changes to the customers’ purchasing organisation not only drove the suppliers’ GaM activities but may have also driven suppliers to centralise the coordination of the firms’ customer- related activities (t oulan et al. 2007). Furthermore, Workman et al. ( 2003) argue that many buying firms have centralised their procurement process and expect a similarly coordinated domestic KaM approach from their suppliers. We conclude that these effects are similarly relevant for both strategic and tactical GaM activities:

Hypothesis 1:

the retailers’ centralisation of their international purchasing activities has a positive impact on the manufacturers’ centralisation of (1) strategic and (2) tactical GAM activities.

RD t suggests that suppliers may also respond to the retailers’ standardisation of market offers by centralising GaM activities. Global strategies enable retailers to consolidate their purchasing needs and enhance their negotiating power by accumulating orders. in response, suppliers may strengthen their own negotiating power by centralising their cus - tomer-related activities. Moreover, retailers tend to reduce the number of suppliers if the retailers standardise their offers and, thus, force suppliers to employ a centralised approach. Doing so is reasonable for both strategic and tactical GaM activities because if the retailers standardise their international market offers then there is less need for local marketing or logistics. In our literature review, the customers’ global strategies were viewed as a core driver of the suppliers’ centrally coordinated customer-related activities.

Accordingly, we hypothesise the following:

Hypothesis 2:

the retailers’ standardisation of their international market offers has a positive impact on the manufacturers’ centralisation of (1) strategic and (2) tactical GAM activities.

From the perspective of RDT, the retailers’ integration of their processes may also lead suppliers to centralise strategic and tactical GaM activities. as retailers increasingly integrate their processes, manufacturers may better coordinate the manufacturers’ acti - vities. t his improvement may be driven by reduced losses to friction within the interna- tional retailers’ organisations (e.g., different process configurations) (Corsten and Kumar 736B. swoboda et al.

2005). Although few empirical studies analyse this relationship, the general literature supports the theoretical assumption. For instance, Johansson ( 2002) implies that the retai- lers’ process integration leads to less complexity in terms of the functions and the actors involved in the process. Similarly, Coe and Hess (2005) address how retail organisati- ons upgrade the firms’ procurement and information processes regarding suppliers. Thus, suppliers may need to react by centralising their GAM activities, as suggested by the following hypothesis:

Hypothesis 3:

the retailers’ integration of their processes has a positive impact on the manufacturers’ centralisation of (1) strategic and (2) tactical GAM activi - ties of manufacturers.

c entralisation of GaM activities and Performance RDt suggests that the centralisation of strategic GAM activities drives the effective- ness of GAM. An internationally centralised customer-specific strategy or information processing system (i.e., strategic GAM activities) might strengthen the manufacturers’ coordination options with regards to their relationship with their key accounts. By cen- trally developing and controlling key-account-specific strategies, manufacturers can set their priorities within their relationships. Consequently, centralisation reduces the level of dependency of the manufacturer on the key retailer and may drive GaM effectiveness.

Birkinshaw et al. ( 2001) provide empirical support showing that the centralisation of G aM activities in general and of information processing in particular is relevant to the effectiveness of the GaM program. Based on TCT, one can view GAM as a transaction-specific investment designed to improve the coordination of supplier-retailer relationships (ivens and Pardo 2008). c om- panies may introduce GAM specifically to better serve their strategically important cus - tomers; hence, it represents a sunk cost if the relationship is terminated (Jap and Ganesan 2000). The specific nature of transaction-specific investments promotes the implementa - tion of mechanisms to reduce the probability of opportunistic exploitations (heide 1994).

Centralising strategic activities, such as pricing and conditions, is an essential issue because pricing represents a core field of conflict within the supplier-retailer relationship.

a company that centrally coordinates its international pricing strategy will reduce the opportunities for retailers exhibiting opportunistic behaviours (e.g., retailers cannot eas- ily arbitrage price differentials). Accordingly, a manufacturer that centralises its strategic G aM activities helps economise its relationship with the retailer and thus increases GaM efficiency (Jap 2001). Therefore, we conclude the following:

Hypothesis 4:

The centralisation of strategic GAM activities has (1) a positive impact on GAM effectiveness and (2) a positive impact on GAM efficiency.

RD t leads to different conclusions concerning the centralisation of tactical GAM activi- ties. A supplier that centralises its customer-specific category management, marketing, or logistics (i.e., tactical GAM activities) might not meet the retailer’s local objectives, as retailers must address local consumer needs. In contrast, a supplier that decentralises its tactical GAM activities might support the international retailers’ country-specific needs. 737 Does centralising Global account Management activities … Thus, the manufacturer’s position is strengthened; this may lead to superior relationship outcomes. Similarly, Harvey et al. (2003) indicate that centralisation includes the risk of overly general logistics or category management with regard to a key account’s needs in different countries. In a domestic context, the close contacts at the key account’s stores are beneficial for the distribution and promotion of the suppliers’ products (Verbeke et al.

2006).

Based on TCT, we can argue that a firm that centralises its tactical GAM activities will increase GAM efficiency, because it will provide thorough international coordination ( ivens and Pardo 2008). Furthermore, centralisation functions as a control mechanism or safeguard that reduces the retailers’ opportunism (on safeguarding in retailer-manufac- turer relationships, see, e.g., Jap and Ganesan 2000). Conversely, if a supplier implements a decentralised approach, then the firm must establish its local resources, and the organi - sation’s fragmentation leaves room for retailers to play regional subsidiaries against each other . shi et al. ( 2010) provide empirical evidence showing that marketing standardisation drives GaM performance. By considering Jap’s ( 2001) study of customer relationships, we conclude that firms need to centralise their tactical activities to prevent opportunistic behaviours as well. In sum, we hypothesise the following:

Hypothesis 5: The centralisation of tactical GAM activities has (1) a negative impact on GAM effectiveness, and (2) a positive impact on GAM efficiency.

t he Moderating effect of Dependency on the Key account Based on RDT, we argue that the manufacturers’ dependency on the key accounts has a moderating effect on the impact of the international retailers’ activities on the cen- tralisation of GAM activities . Centralisation, standardisation and integration help inter - national retailers pool their bargaining power and enhance their market positions (crook and combs 2007). The higher a manufacturer’s dependency on the retailer, the more the manufacturer is forced to react and strengthen its own negotiating position by develo- ping an appropriate customer relationship. the impact of the retailers’ activities on the manufacturer’s strategic and tactical GAM activities is equally relevant, as manufacturers benefit from adapting to the specific requirements of the retailer in both cases. Conse - quently, we hypothesise the following:

Hypothesis 6: When a supplier is highly dependent on its key account the impact of (1) the retailer’s centralisation of its international purchasing activities, (2) the retailer’s standardisation of its international market offers, and (3) the retailer ’s integration of processes on the centralisation of (a) strategic and (b) tactical GaM activities is enhanced.

We further suggest that the supplier’s dependency on its key account influences the impact of centralisation of strategic GAM activities on GAM performance. Because the retailer controls the distribution options that the supplier desires, the supplier’s dependency enlar - ges the retailer’s power (Workman et al. 2003). Accordingly, the retailer can substantially af fect the supplier’s decision making (ivens and Pardo 2008). Moreover, suppliers have fewer options to control and strategically develop the relationship to achieve superior 738B. swoboda et al.

relationship outcomes and thereby improve the effectiveness of their GaM relationship (Fink et al. 2006). In extreme situations, the retailer may even dictate the terms and con - ditions of the relationship to the supplier (Geylani et al. 2007). Therefore, we hypothesise the following:

Hypothesis 7:

When a supplier is highly dependent on its key account the impact of cen- tralisation of strategic GAM activities on (1) GAM effectiveness and on (2) GAM efficiency is reduced.

Finally, we argue that the supplier’s dependency on its key account moderates the effect of centralisation of tactical GAM activities on GAM performance . For manufacturers that are highly dependent on the key retailer, we assume that the impact of centralising tactical GAM activities will be stronger. Because firms usually try to build closer relationships if the level of dependency is high (Fink et al. 2006), the retail customer will likely realise its advantage and want to exploit the supplier’s dependency. if the supplier’s tactical activities are centrally coordinated, then there is less room for this type of opportunistic behaviour. However, the retailer may also acknowledge the supplier’s coordinated efforts and show greater commitment to the relationship. In this case, we may find superior rela - tionship outcomes. Hence, we hypothesise the following:

Hypothesis 8: When a supplier is highly dependent on its key account the impact of cen- tralisation of tactical GAM activities on (1) GAM effectiveness and (2) GAM efficiency is enhanced.

Empirical  Study sample Design We conducted a survey among consumer goods manufacturers in the year 2009 by fol - lowing h omburg et al.’s ( 2002) procedure. To develop the sample, we selected every consumer goods manufacturer from the food and near-food sectors in Germany by using the “Who-to-Whom” commercial database, which includes approximately 10,000 profi - les of firms from the retail, service and manufacturing industries. This procedure resulted in a sample composed of approximately 1,250 companies. We excluded companies wit - hout international sales and manufacturers that do not deliver to international retailers.

Next, we contacted the association of brand manufacturers and inquired as to whether the manufacturers supply foreign retailers and whether the manufacturers are known to deli- ver to foreign grocery retailers. This procedure narrowed the sample to 670 firms. After conducting a pre-test of the measurements, we sent out 670 questionnaires to the ce Os of consumer goods manufacturers by mail. We asked the ceOs to either complete the questionnaire or to pass it on to the respective global account manager or head of sales responsible for the most important (i.e., in terms of sales volumes) international retail customers (for a similar approach, see Ivens and Pardo 2008; for limitations, see Duffy and Fearne 2004). t he questionnaire could be completed in writing or electronically. two weeks later, we sent follow-up postcards, and another ten days later, we sent e-mails. 739 Does centralising Global account Management activities … Finally, we made telephone calls. Approximately 250 CEOs replied that they were unable to answer the questionnaire. if the ceOs had not answered the questionnaire within the next ten days, then we identified the global account managers and heads of sales by using the aforementioned data sources as well as social networks and directly contacted these people. Although 14 questionnaires had to be excluded because of a considerable number of missing values, we obtained 172 usable questionnaires in total. 61% of respondents were CEOs, head of GAM or head of sales and 39% were account manager. The effective response rate is 25%, which is comparable to those of other studies (Harzing 1997). t he interviewed manufacturers mostly operate in the food and near-food industries (t able 2).

The majority (80%) are national brand producers, whereas 11% produce private label brands.

Measurements With regards to the survey design, we first considered the general aspects of the ques - tionnaire by, for example, using five-point Likert-type scales ranging from 1 (strongly disagree) to 5 (strongly agree) considering the hierarchy of effects and the visual design.

W e adapted all of the questions used in the survey from empirical studies on KaM and retail internationalisation (see table 6). However, we also had to adapt and partly deve - lop appropriate scales for our specific context, as will be described below. We conducted pre-tests with six global account managers and two retail ceOs as part of the scale deve- lopment process and to check for face validity. By discussing the relevance of the items in each factor, we made additional modifications, especially in the measures of GAM effectiveness. We measured the centralisation of GAM activities in accordance with toulan et al.

(2007) by differentiating among the following strategic activities: customer strategy, spe - cific customer information, controlling, and pricing/condition systems. With regard to Table 2: sample characteristics Questioned manufacturer (N = 172) corresponding retail customer indus - try % total sales in Mill.

e UR % Foreign sales share % i ndus- try % total sales in Mill.

e UR % For- eign sales share % Food 41.9 > 1,000 35.5 > 50% 47.7 Food 61.1 > 10,000 50.0 > 50% 20.4 Near- Food 39.6 100– 1,000 41.3 25–50% 20.4 Near- Food12.2 1,000– 10,000 22.7 25– 50% 30.2 Non- Food 18.5 < 100 22.0 < 25% 31.3 Non- Food 25.0 < 1,000 19.1 < 25% 42.4 Miss- ing 0.0 Missing 1.2Missing 0.6Missing 1.7Missing 8.2Miss- ing 7.0 t otal 100.0 total 100.0total 100.0total 100.0total 100.0total 100.0 740B. swoboda et al.

tactical activities , we referred to the literature on retailer-manufacturer collaborations ( aastrup et al. 2008) and investigated the following activities: supply contracts, category management, marketing, and distribution logistics. We measured performance at the GaM level (homburg et al. 2002). Effectiveness addresses superior growth in sales or market share with regards to the key account. Based on our pre-tests, we also included the category profitability at customer level. Further - more, we addressed superior relationship outcomes, such as control of the relationship, to ensure that agreements to maximise sales are executed (t oulan et al. 2007). We measured efficiency in accordance with toulan et al. (2007) by addressing the cost reductions in customer sales, the efforts needed to coordinate the relationship and the usage of cross- selling as an opportunity to minimise the resources used to manage the r\ elationship. We measured the retailers’ centralisation of their international purchasing activities by adopting toulan et al.’s ( 2007) scale, which was adapted according to the retailers’ specific characteristics, such as procurement, contracts on orders/conditions, logistics, merchandising and category management. We measured the retailers’ standardisation of their international market offers (Katsikeas et al. 2006) by using the retailers’ format transfer elements, such as store layout, assortment, and communication (Goldman 2001).

We measured the retailers’ integration of their processes by addressing their implementa- tion of standards, upgrades of logistics, and degree of category management. We adopted the measurement of the supplier’s dependency on the key account from Duffy and Fearne ( 2004). a similar metric was also employed by toulan et al. ( 2007), who inquired about the general importance of the key account. This variable is signifi- cantly correlated with an additional question addressing the loss in sales that would occur if the relationship were to end (0.68). We controlled for additional variables that might influence GAM performance. Aside from centralisation, specialisation and formalisation are important structural coordination elements of GaM. GAM specialisation is defined as the degree to which tasks and activi - ties are divided by the organisation (Olson et al. 2005). Therefore, this measure covers the extent to which specialists from different units and hierarchies, such as logistics and controlling units, are involved in GAM. We adapted Miller and Dröge’s ( 1986) scale to differentiate among the respective units. GAM formalisation is defined as the proc- ess of implementing standardised rules and procedures to manage core customers. We measured this value by asking three questions concerning the companies’ formal com- munication channels, operation procedures and guidelines (Homburg et al. 2002). We controlled for the manufacturer’s company size, which was measured by sales volume, because the related market power might impact the companies’ GaM performances. We also considered the manufacturers’ international experience (i.e., the number of years of foreign operations) because a large amount of international experience might induce more successful implementations of strategies and, thus, influence GAM performance.

Because we interviewed manufacturers in Germany, we employed a dummy variable to control for whether the manufacturer was an international or a regional headquarters of an MNC. Finally, we created dummy variables for the five retail key accounts, which were evaluated several times. However, because the accounts did not exhibit any effects, we excluded them from further analyses. 741 Does centralising Global account Management activities … Method Our methodical approach consisted of three steps. First, we tested the measurements for reliability and validity. Second, we checked for any possible biases, and third, we tested the hypotheses.

With regards to reliability, we initially ensured that any corrected item-to-total-corre- lation would fall below 0.5. To assess the constructs’ reliability, we computed the coeffi - cient alpha and composite reliability. In every case, both values exceed the recommended thresholds of 0.7 and 0.6, respectively (see Table 6). Additionally, the average variance extracted (AVE) is above 0.5 for each of the latent variables. With regards to validity, we assessed the face validity in the pre-tests. Regarding construct validity, we found that the a V e values provide support for convergent validity and that all of the latent variables satisfy the requirement for discriminant validity (see table 3).

We assumed that the probability of non-response bias is limited, as indicated by the comparison of the indexes between early vs. late respondents. We gathered secondary data on the demographics of the randomly selected non-respondents to conduct additional tests comparing the respondent and non-respondent firms on size, age, and sales, wher - ever available. Again, we found insignificant differences. We handled common method bias a priori by employing an appropriate questionnaire design and by using harman’s single-factor test ex-post (Podsakoff et al. 2003). As the first factor accounts for only 16.5% of the total variance explained in exploratory factor analysis, we can assume that common method bias is limited within our study, as we had to rely on the self-reports of executives (spector 2006). Our efforts to control for single response bias with a second respondent resulted in only 12 respondents. However, the comparison resulted in insig- nificant mean differences. To test the hypotheses, we applied a partial least squares (PLS) approach and acknowl- edged the shortcomings of this method, which include a lack of a global optimisation function and, consequently, a lack of measures that can determine the global goodness of model fit (Hulland 1999).

Results t able 6 presents the descriptive statistics of the data. The hypothesis tests show that H1, H2, and H3 are supported by the data (see Table 4). t he centralisation of strategic and tactical GaM activities is determined by the retailers’ centralisation of their international purchasing activities, standardisation of their international market offers, and integration of their processes (R² 0.429 and 0.555). The retailers’ centralisation of their purchasing activities primarily influences the manufacturers’ decisions to centralise their strategic and tactical GAM activities (H1.1 and H1.2). The path coefficients (0.387 and 0.510) show medium to high impacts. In addition, the retailers’ standardisation of their interna- tional market offers impacts the centralisation of strategic and tactical GaM activities but with lower path coefficients (H2.1 and H2.2). A similar effect is observed for the impact of the retailers’ integration of their processes (H3.1 and H3.2). GAM centralisation has a significant positive impact on GAM effectiveness and effi - ciency (R² 0.294 and 0.338). Centralisation of strategic GAM activities has a significant 742B. swoboda et al. Table 3: Discriminant validity squared latent variable correlation aVe 1 2 3 4 5 6 7 8 910 11 12 13 1. Centralisation of strategic GAM activities 0.76 1.0   2. Centralisation of tactical GAM activities 0.73 0.67 1.0   3. GAM effectiveness 0.59 0.18 0.13 1.0 4.

GAM efficiency 0.58 0.27 0.26 0.36 1.0 5. Retailers’ centralisation of int.

purchasing 0.78 0.35 0.49 0.06 0.18 1.0 6. Retailers’ standardisation of int.

market offers 0.81 0.25 0.25 0.10 0.13 0.25 1.0 7. Retailers’ integration of processes 0.72 0.25 0.33 0.12 0.17 0.29 0.26 1.0 8. GAM specialisation 0.74 0.11 0.14 0.12 0.07 0.07 0.11 0.19 1.0 9. GaM formalisation 0.63 0.07 0.08 0.14 0.08 0.07 0.11 0.17 0.13 1.0 10. Manufacturers’ company size 1.0 0.03 0.06 0.00 0.01 0.04 0.02 0.04 0.00 0.00 1.0 1 1. Manufacturers’ international experience 1.0 0.02 0.07 0.00 0.04 0.04 0.01 0.02 0.00 0.00 0.18 1.0 12. Regional headquarters 1.0 0.05 0.04 0.04 0.04 0.03 0.03 0.04 0.02 0.03 0.07 0.04 1.0 13. Dependency on key account 1.0 0.01 0.02 0.02 0.02 0.01 0.01 0.00 0.04 0.05 0.02 0.00 0.01 1.0 D iscriminant validity: s quared correlation < aVe 743 Does centralising Global account Management activities … hypothesis Path s tandardised coefficient Unstandardised coefficient t-value H 1.1 Retailers’ centralisation of international purchasing activities → Centralisation of strategic GAM activities 0.387 0.402 (4.850)*** H 1.2 Retailers’ centralisation of interna- tional purchasing activities → Centralisation of tactical GAM activities 0.510 0.512 (6.948)*** H 2.1 Retailers’ standardisation of inter- national market offers → Centralisation of strategic GAM activities 0.208 0.202 (2.964)** H 2.2 Retailers’ standardisation of inter- national market offers → Centralisation of tactical GAM activities 0.118 0.107 (1.802)* H 3.1 Retailers’ integration of processes → Centralisation of strategic GAM activities 0.189 0.189 (2.303)* H 3.2 Retailers’ integration of processes → Centralisation of tactical GAM activities 0.242 0.253 (3.173)*** H 4.1 c entralisation of strategic GaM activities → GAM effectiveness 0.305 0.312 (2.399)** H 4.2 c entralisation of strategic GaM activities → GAM efficiency 0.293 0.310 (2.483)** H 5.1 c entralisation of tactical GaM activities → GAM effectiveness − 0.023 0.015 (0.182) H 5.2 c entralisation of tactical GaM activities → GAM efficiency 0.195 0.172 (1.668)* c ontrols GAM specialisation → GAM effectiveness 0.146 0.136 (1.800)* GAM specialisation → GAM efficiency 0.040 0.039 (0.551) GAM formalisation → GAM effectiveness 0.214 0.219 (3.225)*** GAM formalisation → GAM efficiency 0.121 0.115 (1.820)* Table 4:

Results of Pls analysis 744B. swoboda et al.

positive influence on GAM effectiveness and efficiency (0.305 and 0.293, respectively) (H4.1 and H4.2). Centralisation of tactical GAM activities has a negative but non-sig- nificant impact on GAM effectiveness (H5.1) but a significant positive impact (0.195) on GAM efficiency (H5.2). Hence, H5.1 is not supported, whereas H5.2 is supported.

With regards to the control variables, we found that the manufacturer’s size impacts GAM effectiveness and efficiency. The manufacturer’s international experience nega - tively impacts GAM efficiency. The dummy variable ‘regional headquarters’ has no impact on the dependent variables. the formalisation of the manufacturer’s GaM activi- ties has positive and significant impacts on GAM effectiveness, and GAM efficiency.

GAM specialisation also has a significant positive impact on GAM ef fectiveness.

H6 to H8 were tested by comparing the manufacturers that are highly dependent on their key account with the manufacturers that exhibit medium to low levels of depend- ency on their key account (see table 5). Neither the impact of the retailers’ centralisation of international purchasing activities on the manufacturer’s centralisation of GaM activi- ties nor the retailers’ integration of processes is moderated by the manufacturers’ depend- ency on the key account. Therefore, H6.1 and H6.3 are not supported. The impact of the retailers’ standardisation of international market offers on GAM centralisation is signifi - cantly different among the groups. Because the effect is stronger for the less dependent manufacturers, we reject H6.2. h ypothesis Path s tandardised coefficient Unstandardised coefficient t-value Manufacturers’ company size → G aM effectiveness 0.116 0.122 (1.735)* Manufacturers’ company size → GAM efficiency 0.110 0.111 (1.748)* Manufacturers’ int. experience → G aM effectiveness 0.060 0.056 (0.738) Manufacturers’ int. experience → GAM efficiency − 0.148 − 0.148 (1.917)* Regional headquarters → GAM effectiveness − 0.097 − 0.096 (1.341) Regional headquarters → GAM efficiency − 0.081 − 0.082 (1.143) R 2 GaM effectiveness 0.294 GAM efficiency 0.338 c entralisation of strategic GaM activities 0.429 c entralisation of tactical GaM activities 0.555 * p ≤ 0 .05; **p ≤ 0 .01; ***p ≤ 0 .001 Table 4: (continued) 745 Does centralising Global account Management activities … hypo- thesisPath h igh dependency on key account (≥ 4; n = 108) l ow dependency on key account (< 4; n = 64) Group comparison coeff. t-value coeff. t-value diff. t-value H6.1.a Retailers’ centralisation of international purchas- ing → Centralisation of strategic G aM activities 0.435 (5.534)*** 0.255 (2.958)*** 0.180 (1.562) H6.1.b Retailers’ centralisation of international purchas- ing → Centralisation of tactical G aM activities 0.509 (6.775)*** 0.417 (5.115)**** 0.092 (0.840) H6.2.a Retailers’ standardisation of international market offers → Centralisation of strategic G aM activities 0.138 (2.202)* 0.370 (4.369)**** 0.232 (2.243)* H6.2.b Retailers’ standardisation of international market offers → Centralisation of tactical G aM activities 0.061 (0.971) 0.233 (3.142)**** 0.172 (1.713)* H6.3.a Retailers’ integration of processes → Centralisation of strategic G aM activities 0.250 (3.303)*** 0.100 (1.125) 0.150 (1.263) H6.3.b Retailers’ integration of processes → Centralisation of tacti - cal G aM activities 0.325 (4.396)*** 0.174 (2.223)* 0.152 (1.318) H7.1 c entralisation of strategic GaM activities → GAM effectiveness 0.190 (1.584) 0.542 (3.919)**** 0.352 (1.847)* H7.2 c entralisation of strategic G aM activities → GAM efficiency 0.145 (1.222) 0.477 (4.115)**** 0.332 (1.840)* H8.1 c entralisation of tactical GaM activities → GAM effectiveness 0.219 (1.910)* − 0.325 (2.665)*** 0.544 (3.109)*** H8.2 c entralisation of tactical GaM activities → GAM efficiency 0.334 (2.657)** − 0.029 (0.287) 0.362 (2.071)* c on - trols GAM specialisation → G aM effectiveness 0.151 (1.989)* 0.183 (2.177)* 0.032 (0.268) GAM specialisation → GAM efficiency 0.064 (.844) 0.004 (0.060) 0.059 (0.527) Table 5: Results of the multigroup Pls analysis 746B. swoboda et al.

Regarding the moderating effect of the impact of GaM centralisation on GaM perfor- mance, we found that both H7 and H8 are supported. For less dependent manufacturers, the impact of centralising strategic GaM activities has a stronger positive impact on their GAM effectiveness (difference in path coefficients of 0.352) and GAM efficiency (0.332). With regard to the centralisation of tactical GAM activities, the path coefficients also significantly differ. The centralising tactical GAM activities has a stronger impact on the performance of manufacturers that are highly dependent on retailers than it does on manufacturers that are less dependent, as the difference in path coefficients between the two groups is 0.544 for the effect on GAM effectiveness and 0.362 for the effect on GAM efficiency. Remarkably, centralising tactical GAM activities is negatively related to GAM effectiveness for less dependent manufacturers.

hypo- thesis Path h igh dependency on key account (≥ 4; n = 108) l ow dependency on key account (< 4; n = 64) Group comparison coeff. t-value coeff. t-value diff. t-value GAM formalisation → G aM effectiveness 0.152 (2.170)* 0.353 (5.921)**** 0.201 (2.063)* GAM formalisation → GAM efficiency − 0.007 (0.104) 0.288 (5.281)**** 0.295 (3.187)*** Manufacturers’ com- pany size → GAM effectiveness 0.185 (3.137)*** 0.065 (0.904) 0.120 (1.267) Manufacturers’ company size → GAM efficiency 0.112 (1.646)* 0.123 (1.908)** 0.011 (0.114) Manufacturers’ int.

experience → GAM effectiveness 0.122 (1.479) − 0.060 (0.751) 0.182 (1.426) Manufacturers’ int. expe- rience → GAM efficiency − 0.155 (1.923)* − 0.171 (2.145)** 0.016 (0.135) Regional headquarters → G aM effectiveness − 0.147 (2.248)* 0.016 (0.192) 0.163 (1.527) Regional headquarters → GAM efficiency − 0.078 (1.096) − 0.112 (1.342) 0.034 (0.300) R 2 GaM effectiveness 0.3180.387 GAM efficiency 0.2980.483 c entralisation of strategic GaM activities 0.479 0.379 c entralisation of tactical GaM activities 0.609 0.490 * p ≤ 0 .05; **p ≤ 0 .01; ***p ≤ 0 .001 Table 5: (continued) 747 Does centralising Global account Management activities … With regard to the controls, the impacts of company size, international experience, regional headquarters, and GAM specialisation do not exhibit significant differences.

However, GAM formalisation significantly and strongly impacts both GAM effectiveness and efficiency for the manufacturers that are less dependent on the key account.

Discussion  and Conclusions summary this study examines the impact of centralising certain GaM activities on GaM perfor- mance. Additionally, we analyse the impact of the international retailers’ activities and the- refore, the specific context of the consumer goods sector on the manufacturers’ decisions to centralise GaM activities. this under-researched area is relevant because retailers are dynamically expanding into foreign markets and because manufacturers are increasingly centralising their GaM activities to change local KaM approaches towards international retailers. With respect to RDT, our results strongly support the argument that international retailers’ activities influence the manufacturers’ decisions to centralise GAM activities.

Moreover, GAM centralisation explains a notable amount of the variance in the firms’ GAM performances. The centralisation of strategic GAM activities (i.e., customer stra- tegy, information, control, and price/condition systems) has a stronger impact on GAM performance than the centralisation of tactical activities (i.e., supply contracts, category management, marketing, and logistics). The results substantially differ if we consider the manufacturers’ dependence on their key accounts. Moreover, the observations are stable with or without considering the control variables formalisation and specialisation, which represent other important structural design elements of GaM. these observations allow for three major theoretical implications and conclusions for managers.

t heoretical implications With respect to our first research question on the extent to which manufacturers centra - lise their GAM activities in response to the international retailers’ activities, the results show that the retailer’s activities strongly influence the manufacturers’ response, as these activities explain half of the variance in GaM centralisation. the retailers’ centralisation of their international purchasing activities drives the manufacturers to centralise strategic and tactical GaM activities much more than the retailers’ international market offers and integration of processes. Intuitively, retailers’ purchasing activities, which serve as the direct counterpart to GAM, most influence suppliers’ centralisation of GAM activities.

Given the strategic importance of pricing, retailers are tempted to utilise the centralisation of their purchasing activities to maximise their negotiating power and generate econo- mies of scale. This tendency is also reflected by the growing importance of international purchasing co-operations. Our result answers shi et al.’s (2004) call to analyse the effects of different facilitating conditions. Moreover, we do not find strong correlations among the international retailers’ purchasing activities, market offers and processes. In domestic markets, retailers attempt to integrate their purchasing, sales and supply processes. The 748B. swoboda et al.

international dynamics may enhance these aims. Retailers prefer to standardise their mar- ket offers while entering foreign markets (Pederzoli 2006) but then adapt these offers to the local conditions during this process (swoboda et al. 2009). At the same time, retailers may integrate their purchasing, sales and supply processes. With respect to our second research question on the impact of the centralisation of GAM activities on GAM performance, the results suggest that the manufacturers’ cen - tralisation of their strategic and tactical GaM activities explains a notable amount of the variance in GAM effectiveness and efficiency. Thus, we agree with Shi et al. ( 2010) and Birkinshaw et al. (2001) that the centralisation of GaM activities is a key driver of GaM performance. We will now discuss these two conclusions in greater depth. First, we must acknowledge the differences between strategic and tactical GAM activi - ties in the consumer goods sector. the results suggest that GaM effectiveness is only driven by the centralisation of strategic GAM activities. Accordingly, manufacturers must primarily develop customer-specific strategies (Zupancic 2008). We also found that the centralisation of tactical GAM activities has a negative but insignificant effect on GAM effectiveness. This finding is characteristic of the consumer goods sector, where manu - facturers need to build close relationships not only at their international headquarters but also at the store level to achieve superior performances. In contrast, GAM efficiency is driven by the centralisation of both strategic and tactical GaM activities. Following shi et al.’s (2010) call to conduct sector-specific studies, we highlight the need to differentiate between strategic and tactical GAM activities in the consumer goods sector. Thus, we call upon future studies to further analyse the roles of strategic and tactical GaM activities as the drivers of GaM performance. Second, our results show that the centralisation of GAM activities has different impacts on GAM effectiveness as compared with GAM efficiency. If manufacturers want to achieve greater efficiency, then they have to centralise both their strategic and tacti - cal activities. With regard to GAM effectiveness, we find that only the centralisation of strategic GAM activities drives GAM effectiveness, whereas tactical activities need to be viewed in the specific context of dependency. Thus, it is useful to differentiate between GAM effectiveness and efficiency. Effectiveness and efficiency represent complemen - tary goals for companies. companies need to achieve superior relationship outcomes and minimise their costs and efforts to manage their relationships with their key accounts. By centralising GAM, a manufacturer can pursue both goals at the same time or prioritise one over the other depending on the retailer’s growth rate or the country’s market char- acteristics. Consequently, future scholars need to conduct differentiated investigations of GaM performance. Only Workman et al. ( 2003) indicate how KaM effectiveness and efficiency interact on a domestic level, but they assume that KAM effectiveness is a determinant of efficiency. In response to our third research question, we conclude that the manufacturers’ depend - ence on their key accounts strongly moderates the impact of the decision to centralise G aM activities on GaM performance. We discuss two major conclusions below.First, the impact of the international retailers’ activities on the centralisation of GAM activities is quite stable for manufacturers with different levels of dependency on key accounts. However, for highly dependent manufacturers, the international retailers’ inte- gration of processes also determines the degree to which strategic GaM activities are 749 Does centralising Global account Management activities … centralised. This finding supports those of previous studies, which found that retailer- manufacturer relations are primarily related to supply-chain processes (coe and hess 2005; c orsten and Kumar 2005). However, the same is not true for less dependent manufacturers, who respond more strongly than highly dependent manufacturers to the retailers’ standardisation of offers with a centralised GAM approach. Accordingly, these manufacturers may be able to maximise their synergies based on the retailers’ standardi- sation to set the strategic direction for the manufacturer-retailer relationship. Second, the performance effects of centralising GAM activities differ in accordance with the degree to which the manufacturers depend on their key accounts. a high degree of dependency leaves less room for manufacturers to set strategic directions for their rela- tionships with their key accounts. Hence, manufacturers need to adapt strongly to the spe - cific needs of their key accounts. Coe and Hess (2005) and Geylani et al. (2007) show that manufacturers must address the channel power of dominant retailers by providing addi- tional support and price concessions. Therefore, only the centralisation of tactical GAM activities has a significant impact on GAM effectiveness and efficiency. Strong central control of tactical activities helps manufacturers ensure that agreements are executed at their headquarters. Furthermore, manufacturers can identify local opportunities to gener - ate additional support in the stores for their products. In contrast, less dependent manu - facturers benefit from centralising their strategic GAM activities in terms of both GAM effectiveness and efficiency. According to Geylani et al. (2007), these manufacturers have more opportunities to develop their relationships with their retailers and set the directions of these relationships. In these cases, the manufacturers can actively develop strategies designed to attain high levels of GAM performance. Otherwise, the decentralisation of tactical GAM activities drives GAM effectiveness. This trend is specific to the consumer goods sector because consumer goods manufacturers maximise their local opportunities.

Additionally, decentralisation addresses the current situation in the consumer goods sec - tor where category management, marketing and logistics are adapted nationally (Fernie and s taines 2001). Given its strong moderating role, dependency should be researched by future studies.

Managerial implications this study also provides managerial implications by highlighting the characteristics of the consumer goods sector and, thus, the dynamic retailer-manufacturer relationship.

The decision to centralise GAM activities is strongly influenced by the retailers’ buying organisations. This finding implies that the international retailers’ activities are the main determinant of the decision to centralise GAM. Accordingly, a manufacturer must unders - tand the specific needs of the retailers and the direction of their development in order to determine the degree to which it should centralise its GAM activities. In particular, expe - rienced manufacturers should continually evaluate their GaM activities to ensure that their GAM is efficiently configured. Moreover, this study does not confirm the general assumption that all GaM activities should be centralised. although the centralisation of strategic elements drives GAM performance, the effects of centralising tactical activities on GaM performance are contingent upon the manufacturers’ dependence on the retai- lers. If the manufacturers are less dependent, then they can maximise the success of their 750B. swoboda et al.

GaM activities and their regional competitive advantages by building close local relati- onships. Hence, if a manufacturer is rethinking the configuration of its GAM activities, it should consider its level of dependency on its key accounts and re-organise accordingly.

Limitations  and Further  Research Because the present study has several limitations, scholars must conduct additional rese - arch to better understand international retailer-manufacturer relationships. With respect to data collection, we made every effort to include all consumer goods manufacturers in the sample, but we cannot guarantee the generalisability of the sample.

Because we conducted interviews with only a single person in each company and only one partner in each partnership in only one country, this study is somewhat limited in its scope (Duffy and Fearne 2004; John and Reve 1982). Further, we have implied causality while testing covariation between constructs measured at the same time. Thus, broaden- ing the database will avoid these limitations and allow for additional conclusions. For example, scholars may analyse the retailers’ perspectives or dyadic relations. However, the latter is not easy to examine, especially in the consumer goods sector, because most international retailers are not obligated to publish their financial data and usually refuse to participate in surveys. With respect to the conceptualisation process, we conceptualised and measured the elements of GAM as precisely as possible. However, GAM is complex in nature, and our conceptualisation focuses on the centralisation of GaM activities as a key success factor. Thus, there are other elements in the manufacturer-retailer relationship that might be examined in a future study. some examples include additional GaM organisation ele- ments (homburg et al. 2002), antecedents (e.g., the collaborative orientation) (Shi et al.

2005), the intensity of the retailer-manufacturer partnership (McDonald et al. 1997), or the business configuration of the international retailers beyond their levels of centralisa- tion, standardisation and integration. Furthermore, our attempt to adapt the scales to the consumer goods context is still exploratory in nature and is primarily based on previous studies as well as a pre-test. More fine-grained measures may enhance the precision of the results and enlarge the scope of the implications that can be drawn from an empirical study. 751 Does centralising Global account Management activities … MW/STDi-t-t cor- relation Factor loadings t-test (Boots- trapping) c ronbach’s alpha a Ve composite reliability s ource the following strategic GAM activities for this customer are coordinated only centrally and not on a country-by-country basis: 0.894 0.761 0.927 a d apted from toulan et al.

( 2007 ); Zupancic ( 2008); Wilson and Weilbaker ( 2004) • customer strategy 4.1/1.1 0.81 0.90 45.09 • Prices and condition systems 3.3/1.5 0.62 0.78 21.34 • Customer specific information 3.0/1.3 0.80 0.89 49.04 • customer controlling 3.2/1.1 0.82 0.91 61.08 t he following tactical GAM activities for this customer are coordinated only centrally and not on a country-by-country basis: 0.874 0.726 0.914 • supply contract 2.8/1.4 0.77 0.87 36.69 • category management 2.7/1.3 0.73 0.83 26.02 • Marketing 2.7/1.3 0.70 0.84 30.86 • Distribution logistic 2.8/1.4 0.73 0.86 31.23 a re you more successful (compared to your average accounts) in the rela\ tion with this key account in terms of: GAM effectiveness 2 0.827 0.592 0.878 New scale based on toulan et al.

(2007) • Growth in sales to customer 3.7/1.0 0.67 0.83 24.58 • Growth of market share at customer 3.7/1.1 0.67 0.83 21.61 • Growth of category profitability 3.2/1.1 0.61 0.76 15.44 • Greater control of relationship 3.4/1.1 0.59 0.72 12.41 • Reduction of conflicts 3.1/1.1 0.56 0.70 11.42 Appendix Table�:

able �: Reliability and validity of measurements 752B. swoboda et al. MW/STD i-t-t cor- relation Factor loadings t-test (Boots- trapping) c ronbach’s alpha a Ve composite reliability s ource GAM efficiency 0.764 0.584 0.847a ccording to toulan et al. ( 2007) • coordination of sales to customer operations internationally 3.3/1.2 0.51 0.67 8.75 • Reduced cost of sales to customer 2.8/1.0 0.57 0.84 34.18 • More efficient use of salespeople’s time in serving customers 2.9/1.1 0.67 0.87 41.50 • cross- selling into weaker divi - sions of customers’ operations 2.8/1.2 0.50 0.65 9.82 t he following purchasing activities of the key account are undertaken centrally and not on a country-by-country basis 0.927 0.776 0.945 New scale based on toulan et al.

(2007) • Contracts on orders/conditions 3.4/1.5 0.93 0.83 36.93 • Logistic activities/contracts 2.7/1.5 0.73 0.88 42.89 • category management 2.8/1.4 0.81 0.89 45.79 • Merchandising activities 2.5/1.4 0.82 0.84 30.86 • international purchasing activities 2.9/1.3 0.75 0.96 137.70 t he key account standardises his international activities, e.g., applies everywhere the same retail offers/concepts 0.921 0.807 0.944 a dapted from Katsikeas et al.

(2006); Goldman ( 2001) • Format and store layout 3.7/1.2 0.79 0.86 31.31 • a ssortments and product offerings 3.3/1.1 0.81 0.89 40.05 • c ommunication and merchan- dising 3.3/1.2 0.80 0.90 51.52 • international format transfer strategy 3.3/1.1 0.87 0.94 82.85 T able �:

able �: (continued) 753 Does centralising Global account Management activities … MW/STDi-t-t cor- relation Factor loadings t-test (Boots- trapping) c ronbach’s alpha a Ve composite reliability s ource the key account internationally integrates following processes/activities 0.867 0.716 0.910New scale based on coe and hess (2005); Johansson (2002) • implementation of standards 3.5/1.1 0.66 0.80 19.06 • Upgrading of logistics systems 3.2/1.1 0.72 0.85 33.99 • Greater degree of category mgmt. 3.9/1.1 0.72 0.86 40.44 • integrated supply chain processes 3.3/.8 0.76 0.87 35.34 GAM specialisation (integration in the G aM) 0.825 0.741 0.895New scale based on Miller and Dröge (1986) • Finance and controlling 2.8/1.2 0.71 0.89 16.79 • logistic and supply chain 3.1/1.2 0.70 0.86 17.57 • Marketing, market research, R&D 3.2/1.2 0.64 0.83 13.04 GAM formalization 0.706 0.629 0.835a d apted from homburg et al.

( 2002 ) • Within our organisation formal communication channels are fol- lowed when working with the key accounts. 3.3/1.1 0.44 0.70 6.81 • t o coordinate the parts of our organisation working with key account standard operating proce- dures have been established. 3.1/1.1 0.63 0.84 12.05 • We have put a lot of thought into developping guidelines for work- ing with this key account. 2.4/1.2 0.64 0.84 13.66 Dependency on key account Duffy and Fearne (2004) • We view the customer as one of our most important retail key accounts 3.7/1.1 1.0 1.0 Q uestions were asked on a scale between 5 = strongly agree to 1 = strongly disagree Table �:

able �: (continued) 754B. swoboda et al.

References Aastrup, J., Kotzab, H., Grant, D. B., Teller, C., & Bjerre, M. (2008). A model of structuring effi- cient consumer response measures. International Journal of Retail & Distribution Manage- ment, 36(8), 590–606.

Arnold, D., Birkinshaw, J., & Toulan, O. (2001). Can Selling be globalized? The Pitfalls of global account management. California Management Review, 44(1), 8–20.

Birkinshaw, J., Toulan, O., & Arnold, D. (2001). Global account management in multinational cor - porations: t heory and evidence. Journal of International Business Studies, 32(2), 231–248.

coe, N. M., & Hess, M. (2005). The internationalization of retailing: Implications for supply net - work restructuring in east asia and eastern europe. Journal of Economic Geography, 5(4), 449–473.

corsten, D., & Kumar, N. (2005). Do suppliers benefit from collaborative relationships with large retailers? An empirical investigation of efficient consumer response adoption. Journal of Mar- keting, 69(3), 80–94.

Crook, T. R., & Combs, J. G. (2007). Sources and consequences of bargaining power in supply chains. Journal of Operations Management, 25(2), 546–555.

Draganska, M., & Klapper, D. (2007). Retail environment and manufacturer competitive intensity. Journal of Retailing, 83(2), 183–198.

Duffy, R., & Fearne, A. (2004). The impact of supply chain partnerships on supplier performance. International Journal of Logistic Management, 15(1), 57–72.

Fernie, J., & Staines, H. (2001). Towards an understanding of European grocery supply chains. Journal of Retailing & Consumer Services, 8 (1), 29–36.

Fink, R. C., Edelman, L. F., Hatten, K. J., & James, W. L. (2006). Transaction cost economics, resource dependence theory, and customer-supplier relationships. Industrial & Corporate Change, 15(3), 497–529.

Geylani, T., Dukes, A. J., & Srinivasan, K. (2007). Strategic response to a dominant retailer. Mar- keting Science, 26(2), 164–178.

Goldman, A. (2001). The transfer of retail formats into developing economies: The example of c hina. Journal of Retailing, 77(2), 221–242.

Gosselin, D. P., & Bauwen, G. A. (2006). Strategic account management: Customer value creation through customer alignment. Journal of Business & Industrial Marketing, 21(6), 376–385.

Harvey, M., Myers, M. B., & Novicevic, M. M. (2003). The managerial issues associated with global account management: a relational contract perspective. Journal of Management Devel- opment, 22(2), 103–129.

Harzing, A.-W. (1997). Response Rates in Mail Surveys. International Business Review, 6(6), 641–665.

Heide, J. B. (1994). Interorganizational Governance in Marketing Channels. Journal of Marketing, 58(1), 71–85.

Hennart, J. F. (1991). Control in multinational firms: The role of price and hierarchy. Management International Review, 31(1), 71–96.

homburg, C., Workman, J. P., & Jensen, O. (2002). A configurational perspective on key account management. Journal of Marketing, 66(2), 38–60.

Homburg, C., Droll, M., & Totzek, D. (2008). Customer prioritization: Does it pay off, and how should it be implemented? Journal of Marketing, 72(3), 110–130.

Hulland, J. (1999). Use of Partial Least Squares (PLS) in strategic management research: A review of four recent studies. Strategic Management Journal, 20(2), 195–204.

ivens, B. S., & Pardo, C. (2008). Key-account-management in business markets: An empirical test of common assumptions. Journal of Business & Industrial Marketing, 23 (5), 301–310.

Jap, S. D. (2001). Perspectives on joint competitive advantages in buyer-supplier relationships. International Journal of Research in Marketing, 18(1/2), 19–35. 755 Does centralising Global account Management activities … Jap, S. D. & Ganesan, S. (2000). Control mechanisms and the relationship life cycle: Implica - tions for safeguarding specific investments and developing commitment. Journal of Marketing Research, 37(2), 227–245.

Johansson, U. (2002). Food retail buying processes: A study of the UK, Italy and Sweden. Interna- tional Journal of Retail & Distribution Management, 30(12), 575–585.

John, G., & Reve, T. (1982). The reliability and validity of key informant data from dyadic relation - ships in marketing channels. Journal of Marketing Research, 19(4), 517–524.

Katsikeas, C. S., Samiee, S., & Theodosiou, M. (2006). Strategic fit and performance consequences of international marketing standardization. Strategic Management Journal, 27(9), 867–890.

Krentzel, G. (2009). Going global. Marketing Management, 18(5), 20–25.

Liu, H., & McGoldrick, P. (1996). International retail sourcing: Trend, nature and process. Journal of International Marketing, 4(4), 9–33.

McDonald, M., Millman, T., & Rogers, B. (1997). Key account management: Theory, practice and challenges. Journal of Marketing Management, 13(8), 737–757.

Miller, D., & Dröge, C. (1986). Psychological and traditional Determinants of Structure. Adminis- trative Science Quarterly, 31(4), 539–560.

Montgomery, D. B., & Yip, G. S. (1999). Statistical Evidence on Global Account Management Programs. Thexis, 4, 10–13.

Montgomery, D. B., & Yip, G. S. (2000). The Challenge of global Customer Management. Market- ing Management, 9(4), 22–29.

Olson, E. M., Slater, S. F., & Hult, G. T. M. (2005). The performance implications of fit among business strategy, marketing organization structure, and strategic behavior. Journal of Market- ing, 69(3), 49–65.

Pederzoli, D. (2006). Conception and test of a comprehensive model of international strategy for retail companies. International Review of Retail, Distribution & Consumer Research, 16(4), 415–431.

Pfeffer, J., & Salancik, G. R. (1978). The external control of organizations: A resource dependence perspective. New York: Harper & Row.

Podsakoff, P. M., MacKenzie, S. B., Lee, J.-Y., & Podsakoff, N. P. (2003). Common method biases in behavioural research: a critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5), 879–903.

Richards, K. A., & Jones, E. (2009). Key Account Management: adding elements of account fit to an integrated theoretical framework. Journal of Personal Selling and Sales Management, 29(4), 305–320.

Shi, L. H., Zou, S., White, J. C., & Cavusgil, S. T. (2004). A conceptual framework of global account management capabilities and firm performance. International Business Review, 13(5), 539–553.

Shi, L. H., Zou, S., White, J. C., McNally, R. C., & Cavusgil, S. T. (2005). Executive insights, global account management capability: insights from leading manufacturers. Journal of Inter- national Marketing, 13(2), 93–113.

shi, l. h., White, J. C., Zou, S., & Cavusgil, S. T. (2010). Global account management strategies:

drivers and outcomes. Journal of International Business Studies, 41(4), 620–638.

Spector, P. E. (2006). Method variance in organizational research. Organizational Research Meth- ods, 9(2), 221–232.

Swoboda, B., Foscht, T., & Cliquet, G. (2008). International value chain activities of retailers and wholesalers. Journal of Retailing and Consumer Services, 15 (2), 63–77.

Swoboda, B., Zentes, J., & Elsner, S. (2009). Internationalisation of Retail Firms: State of the Art after 20 Years of Research. Marketing–Journal of Research and Management, 5 (2), 105–126.

t oulan, O., Birkinshaw, J., & Arnold, D. (2007). The role of interorganizational fit in global account management. International Studies of Management & Organization, 36(4), 61–81. 756B. swoboda et al.

Verbeke, W., Bagozzi, R. P., & Farris, P. (2006). The role of key account programs, trust, and brand strength on resource allocation in the channel of distribution. European Journal of Marketing, 4(5/6), 502–532.

Williamson, O. E. (1991). Comparative economic organization: The analysis of discrete structural alternatives. Administrative Science Quarterly, 36(2), 269–296.

W ilson, K., & Weilbaker, D. (2004). Global account management: A literature based conceptual model. Mid-American Journal of Business, 19(1), 13–21.

Workman, J. P., Homburg, C., & Jensen, O. (2003). Intraorganizational determinants of key account management effectiveness. Journal of the Academy of Marketing Science, 31 (1), 3–21.

Yip, G. S., & Bink, A. J. M. (2007). Managing global accounts. Harvard Business Review, 85(9), 102–111.

Yip, G. S., & Madsen, T. L. (1996). Global account management: The new frontier in relationship marketing. International Marketing Review, 13(3), 24–42.

Zupancic, D. (2008). Towards an integrated framework of key account management. Journal of Business & Industrial Marketing, 23(5), 323–331. Copyright of Management International Review (MIR) is the property of Springer Science & Business Media B.V. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.