Supply chain strategy in the luxury fashion industry: impacts on performance indicators

Abstract

Purpose

The purpose of this paper was to investigate what types of supply chain strategies (SCS) are implemented within luxury fashion companies, according to the drivers that regulate competitiveness in this sector (brand positioning, distribution channel, type and line of product). Moreover, the objective was to define which key performance indicators (KPIs) should be measured according to the chosen strategy, and finally to evaluate the alignment of luxury fashion companies with the proposed indicators.

Design/methodology/approach

The literature review was the first step performed. Thereafter, a case study was conducted and the sample, composed of six companies, was selected, a questionnaire was then developed to guide the interviews, after which the data were collected. From the data, a primary case analysis was conducted, from which cross-case patterns were also researched.

Findings

From the results obtained, it was possible to state that companies involved in the case study adopted different SCS within the same company according to the drivers that regulate the sector competitiveness. As a result, the product line was shown to be the only driver that affected both the alignment between the expected and implemented SCS, respectively, and the alignment with the selected KPIs.

Originality/value

The paper provides valuable insights to companies that are trying to align SCS and KPIs. The close link between these aspects had not yet been explored previously. In particular, there were no indications about the KPIs that have to be measured for a specific SCS.

1. Introduction

The luxury fashion (apparel and footwear) and leather goods accounted for 50% of the luxury goods revenue in 2021. The worldwide revenue in 2021 are estimated to be US$150 billion (‘Luxury Fashion – Worldwide | Statista Market Forecast’, 2021). In particular, luxury fashion and leather goods industry in Europe is expected to overcome US$55 billion, attained by more than 185,000 companies comprising 2 million employees (“Luxury Fashion – Europe | Statista Market Forecast”, 2021).

The luxury fashion sector represents a distinct example within the manufacturing industry, showing several characteristics rendering it difficult for the management of both the productive and logistic processes. In particular, three principle aspects have been identified (Christopher et al., 2004). The first aspect pertains to the short product life cycle of the fashion product designed to capture the current mood of the moment. The second relates to the unpredictable and volatile demand of customers that are subject to complete change within a relatively short period of time (Wang et al., 2012). Finally, there is the impulsive purchasing behavior (Newman and Foxall, 2003), attributable to the nature of fashion shoppers as well as the impulsiveness surrounding their purchase behavior.

The strong competitiveness within this sector has led many companies to undertake initiatives in rationalizing the operating processes, aimed at improving the ability to respond to the continuous changes in customer demand, while simultaneously, improving both the efficiency and speed of the entire supply chain (SC) (Vona, 2003).

Therefore, in the face of the constant changes in consumer needs, successful companies can be defined as those with the ability to respond to these rapid changes by reducing lead time (Christopher and Peck, 2004). Within this context, there is the need to define an SCS that can maximize the performance of fashion companies. The SCS is defined as a set of management methods and activities by which a SC obtains advantages over competitors (Brun and Castelli, 2008). It is necessary to specify that, when talking about an SCS, there is no “one best way” or best choice in any specific case. Rather it is an approach that depends on both the internal characteristics of the company as well as the environmental characteristics (Lee, 2003). Therefore, in order to define the correct SCS, it is necessary to, firstly, identify the context in which the SCS has to compete and then define the objectives of the single SC.

The definition of the SCS is strongly linked to the issue of performance monitoring, in order to assess whether the strategy is implemented correctly and to direct the SC toward strategic objectives. For the fashion sector, the literature provides information about various SCS and performance indicators (Brun et al.., 2017; Brun and Castelli, 2008). However, the close link between the two concepts has not yet been explored. In particular, there is no evidence regarding the types of performance indicators that have to be measured for a specific SCS.

Based on a literature analysis and case studies, the present article aims to provide necessary information to fashion luxury companies regarding the type SCS that should be implemented, based on the factors that regulate competitiveness in this sector. Moreover, the article also aims to identify the correct KPIs to measure, according to the chosen strategy.

In total, two research questions (RQ) are, therefore, addressed in the present article:

1. What types of SCS are currently adopted by fashion companies and why are these strategies adopted?

2. Which KPIs should be measured by fashion companies according to the SCS identified in RQ1?

The contribution of this article includes:

1. Investigating what types of SCS are implemented within fashion companies and understanding the reasons why companies choose to implement these strategies.

2. Defining which performance indicators should be used to measure luxury fashion companies in relation to the strategy adopted and the evaluation of the alignment of luxury fashion companies with the proposed indicators.

Given this background, the paper has been developed as follows. Section 2 presents a literature review on the SCS implemented within the fashion companies and the factors that influence these choices, as well as the KPIs measured within fashion companies. Section 3 outlines the RQ and methodology, also introducing the principle features of the companies involved in the case studies performed. Section 4 and Section 5 present the findings and analysis of these results, respectively. Finally, section 6 offers some concluding remarks with suggestions for further developments.

2. Literature review

2.1 Supply chain strategy

Regarding the SCS strategies proposed in the literature, particular attention should be paid to the characteristics between lean and agile strategies, mentioned in all papers regarding this topic, as well as the hybrid strategy that results from the combination of the two aforementioned strategies.

The lean SCS (Table 1) is aimed at creating a cost-efficient SC, by effectively managing inventory and focusing on improving the quality in the SC (Cagnazzo et al., 2010), thus eliminating waste (Huang et al., 2002; Christopher and Towill, 2002). Christopher and Peck (2004) have argued that lean SCs perform well under conditions where demand is relatively stable and predictable, and variety is low.

On the other hand, an agile SCS is aimed at being flexible by adapting quickly and effectively to rapidly changing customer needs (Christopher et al., 2004; Huang et al., 2002; Christopher and Towill, 2002; Lin et al., 2006). Effectiveness is the primary goal of this strategy. Finally, the hybrid strategy is derived from the combination of the two aforementioned strategies. In the hybrid SCS, the practice employed to ensure adaptability to the constant changing customer demands is that of postponement (Cagnazzo et al., 2010). The latter strategy is aimed at moving the differentiation of the product downstream to the final assembly, and as a consequence closer to the customer (Mason-Jones et al., 2000). In particular, the decoupling point (DP), is the point at which real demand penetrates upstream in a SC. Christopher and Towill (2002) contend that processes are designed to be lean up until the DP and agile beyond that point, respectively. Therefore, the hybrid strategy pursues the cost effectiveness up until the DP and the high service level, necessary in the volatile marketplace, downstream of that point (Christopher, 2000).

Depending on the positioning of the DP, different order fulfillment strategies can be defined, for instance make to stock (MTS), in which the SC activities are based on forecasts, and make to order (MTO), in which the supply chain activities are triggered by orders (Wortmann, 1992). In line with Christopher and Towill (2002), lean SC strategies are expected to be mostly associated to MTS situations, while agile for MTO. Interestingly, given the SC perspective of the paper, the term lean is used to consider the entire SCS, not the specific tools used in each stage of the SC, differently from operations management literature that is traditionally focusing on the techniques, such as Kanban systems (Slack et al., 2010), used within a factory.

2.2 Factors that influence the choice of a supply chain strategy

From an analysis of the literature, there is the possibility of implementing, within a single company, multiple SCS on the basis of factors that drive the competition, namely brand positioning, distribution channel, type of product (Brun and Castelli, 2008) and line of product (Brun et al., 2008). Based on these four factors, a company is in a position to make various decisions. These decisions may include the application of the same strategy to the entire SC, the application of a different strategy based on one factor (i.e. business segment), selecting an intermediate level, thereby segmenting the SCS on the basis of two of the proposed factors and finally segmenting the SCS according to all four factors proposed (Brun and Castelli, 2008). In addition, from this segmentation, it is possible to understand the hierarchy of the four factors (i.e. which is the main determinant of the strategic segmentation).

Table 2 reports the four factors and highlights how these can influence the choice of the SCS.

2.3 Link between supply chain strategies and key performance indicators

The KPIs can be defined as the physical values which are used to measure, compare and manage the overall organizational performance (Gosselin, 2005). The KPIs may include the quality (De Toni and Tonchia, 2001; Gosselin, 2005; Badri et al., 1995; Neely et al., 2005), cost (De Toni and Tonchia, 2001; Neely et al., 2005; White, 1996), financial (Parmenter, 2015; White, 1996), flexibility (De Toni and Tonchia, 2001; and White, 1996), delivery reliability (White, 1996), employee satisfaction (Leong et al., 1990; Ishaq Bhatti and Awan, 2014; Parmenter, 2015), customer satisfaction (Ittner and Larcker, 1998 and Neely et al., 2005; Parmenter, 2015), safety (Flin et al., 2000; Mearns et al., 2003; Parmenter, 2015), environment/community (Neely et al., 2005; Parmenter, 2015; White, 1996) and learning and growth (Parmenter, 2015; Sadler-Smith et al., 2001; Utterback and Abernathy, 1975).

KPIs are used by the organizations in order to ensure that they are proceeding in correct direction, achieving targets in terms of organizational goals and objectives. The performance measures are used to evaluate and control the overall business operations. These indicators are also used to measure and compare the performance of different organizations within the industry, plants, departments, teams as well as between individuals, respectively (Ghalayini et al., 1997; Ishaq Bhatti and Awan, 2014; Parmenter, 2015).

Therefore, KPIs, beyond the objective of improving SC performance, may also represent the means to align the SC with the business strategy.

In order to identify the KPIs measured by fashion companies, a structured literature analysis was conducted (Figure 1).

From the analysis of the selected papers, an initial set of 33 KPIs was firstly identified starting from the following keywords: (“Key performance indicator” OR “performance measurement”) AND (fashion OR luxury) and (SCOR AND fashion). These were both generic for manufacturing and specific for the fashion sector. The identified KPIs were then integrated and detailed through the metrics of the supply-chain operations reference (SCOR) model and design chain operations reference (DCOR) model. For example, the supplier service level was subdivided into three indicators identified in the SCOR model as follows: % of orders received on time with respect to the requested date, % of orders received with the correct quantity requested and % of compliant orders received. Thus, the number of KPIs identified in the literature as a result of the integration of the metrics present in the DCOR and SCOR models increases from 33 to 53. Finally, the number of identified KPIs increases again from 53 to 71 as a result of the previously detailed breakdown operation (Figure 2).

Subsequently, a further analysis to understand how to classify and categorize the identified KPIs was then conducted. The literature provides several examples of classification, but for the purpose of the present article, KPI classification was performed using the classification framework of the DCOR and SCOR models, respectively (Thilakarathna et al., 2015).

Following this classification, a second classification was performed according to the three types of SCS previously identified.

The classification of KPIs within a single strategy was made according to the order winner-qualifier matrix. Hill (1995) introduced the concept of order winner (OW) and order qualifier (OQ) by contending that manufacturing companies should choose their strategy based on these parameters. The correct definition of the OQ and OW leads to a correct definition of the business strategy which, in turn, results in strategic choices at the SC level.

It' is essential to understand the connection between the concept of OW and OQ, and lean and agile strategies, respectively. For example, a lean strategy is more effective when the OW is the cost, whereas an agile SCS is more effective when the OW is the service level (Mason-Jones et al., 2000).

Starting from the OW and OQ matrix, it was possible to produce a second classification by positioning the KPIs within the identified strategies. The association between KPIs and SCS, identified from the literature analysis, was carried out using the OW concept, since it was chosen to report and classify only those KPIs related to the cost (OW of the Lean strategy) for the lean strategy and those related to the service level (OW of the agile strategy) for the agile strategy.

The results of the KPIs process classification and the association to a strategy are shown in Table 3.

During the case studies, the above classifications were used in order to verify whether the companies measure and monitor at least the set of identified KPIs, even if under a different name, necessary to achieve the objectives and implement a lean or agile SCS.

3. Research methodology

The methodology selected was case-based research since this was considered a more suitable approach in answering the type of RQs formulated. The case-based research was carried out in the first phase and included the analysis of secondary sources and interviews carried out in all the companies. These interviews were conducted using a questionnaire.

The methodology was organized into two steps. The first step was the selection of a sample of luxury fashion companies (Eisenhardt, 1989; Maimbo and Pervan, 2005). Multiple-case sampling was used to increase confidence in the findings (Miles and Huberman, 1984) and support their external validity. The research involved selecting a sample of six companies, belonging to the luxury fashion industry, with different sizes and turnovers that were representative of most of the companies in this sector. Of the six sample companies selected, there were two international brands belonging to leading groups in the luxury fashion sector, one smaller company with a strong vocation for craftsmanship and exclusivity, and three brands with smaller turnover and company size. According to Eisenhardt (1989) selecting a sample size ranging between four and ten cases is effective in this type of research approach. Hence, a sample size of six was considered sufficient in providing an accurate representation in an empirical based research. The sample was composed of companies with heterogeneous characteristics regarding the drivers that potentially influence the choice of SCS identified in the literature. In particular, the selected companies were also represented by some common drivers, thereby permitting us to compare the strategies and KPIs implemented by these companies. Additionally, the companies were also represented by different drivers, thereby highlighting all the differences and permitting us to cover all the distinctive features of the sector.

Then in the data collection phase, all the information in answer to the RQ was obtained. In the first phase, a semi-structured two-hour interview was conducted. Initially, information was collected from indirect sources and from previous knowledge given by projects carried out within the company, and thereafter, included the SC and Information Manager (IM) of the companies identified within the sample. The interviews were carried out using a questionnaire (i.e. consisting of a first part of open-ended questions and a second part of closed-ended questions) divided into three sections. In the first section, general information about the companies was gathered. Then, the second section was dedicated to obtaining information about the drivers (i.e. distribution channel, type of product and line of product) that could be influenced by the choice of SCS. In turn, the third section was aimed at discovering the KPIs monitored, controlled and/or optimized by the luxury fashion companies examined. Finally, both the main characteristics of the individual companies and any points of contrast or similarity between all the companies in the sample were identified in the analysis and synthesis phases, respectively. In the data analysis phase, a transcription and reworking of the information obtained by each individual interviewed within the company was carried out in order to compare all the answers obtained. The realization of these documents was necessary in order to be able to make a representative report of each case study, and then to compare the case studies through a cross-case analysis.

In Table 4 the individual luxury fashion firms involved in the present case study were presented.

A brief summary of the companies is presented below.

Company 1 forms part of the luxury fashion industry, founded in Spain in the 1900s. Last year alone, the company invoiced about 239 M€. Currently, this company is experiencing a period of strong growth and as a result is planning to be autonomous from the owner group, particularly in the footwear sector.

Company 2 is one of the most important luxury fashion companies both in Italy and worldwide. This company was founded in the early 1900s with an increasing turnover (1 B€ in 2019). It manages multiple product lines, including handbags, small leather goods, footwear, ready to wear, bijoux, fashion accessories, furs, belts and fashion jewels.

Company 3 was founded in Milan and is focused on leather goods (bags and accessories). The turnover, according to the latest available data, is approximately 3.2 B€. This company manages many lines, including ready to wear, footwear, leather goods, bijoux, accessories, bags, belts, jewels, perfumes and cosmetics.

Company 4 was founded in Florence in the second half of the 1900s. Originally a small family-run luxury fashion boutique, specialized in the creation of ties, it has currently expanded to become an important international luxury group pillar of made in Italy haute couture with about 600 employees worldwide. The specialized production area is the “total look for men” (suits, shirts, jeans, polo shirts, casual wear and ties), which comprises completely hand-made items.

Company 5 was founded in France about 25 years ago and is positioned within the luxury fashion sector of women's footwear. The core business is represented by women's footwear, which has permitted the brand to attain a turnover of 1B€.

Company 6 originated in 1856 in Hampshire. The iconic items of the English luxury fashion houses are yarns, scarves, shawls and trench coats. The turnover (year 2017) was 2.8 B£ with more than 10 thousand employees (in February 2018) and a capitalization of 7 million pounds (in April 2018).

For every single company, different case studies have been carried out, according to the specific product types realized by the companies. In total, 20 different case studies have been analyzed.

3.1 “Within case” analysis of the SCS – RQ1

In this section the results of the RQ1 were analyzed. In particular, the results of each case study are reported in Table 5.

Company 1 represented the high market segment of the luxury fashion industry. Regarding the second driver, distribution and sales channels, this company used both the wholesale channel (about 40%), as well as the retail and commerce channels (about 60%). The core business of the company is footwear and leather goods that are managed internally. Within this company, four different case studies have been carried out, according to the combination of different product types and product lines realized by the company. According to this, for each type of product there are different lines: carry over and seasonal. The carry over products represent 30% of the total products while the remaining 70% is composed of seasonal products. The raw materials (RM) are managed in stock and are, therefore, purchased in advance (before the arrival of the orders). For the carry over, the management of the RM stock is based on the order history, while for the seasonal products RM stock management is attained from “blind orders” through sales forecasts. At the finished product (FP) level, carry over, seasonal products are managed through an MTO type of production. In particular, the decoupling point is positioned in the warehouse of the RM.

To summarize the results related to the drivers identified in the literature that influence the choice of the SCS, it can be seen that the company 1 belongs to the luxury fashion market segment, that employs the use of the retail channel (60%) as the preferential sales channel. Finally, the core business is made up of two types of product, and each comprises two product lines, respectively. The results of case studies 1 and 2 show a misalignment with the strategies that are expected (based on the literature analysis) at FP level. In particular for these case studies, company 1 implemented a hybrid SCS with an MTO type of production instead of a lean SCS with an MTS type of production. Instead, for the case study 3 and 4 company 1 implements an SCS in harmony with those suggested and expected by the literature.

Company 2 was also shown to manage many product typologies; however, the core business is in leather goods, shoes and ready to wear (RTW). The sales channels used by the company are both retail and wholesale. For leather goods (bags) approximately 70% is sold through the retail channel and the remaining 30% through the wholesale channel, respectively. For the footwear, 60 and 40% are sold through the retail and wholesale channels, respectively. The company 2 deals with both carryover products (about 20% for shoes and 40% for bags) as well as seasonal products. Within the company 2, carry over product planning is accomplished through a sales forecast with a time horizon spanning six months on the basis of data derived from marketing. These data are then cross-referenced with real demand. For case studies 5, 6 and 7 all the types of product and carry over lines are obtained according to an MTS type of production, while for the case studies 8, 9 and 10 all the type of products and in particular for seasonal lines are managed according to an MTO type of production.

Summing up, company 2 uses the retail channel (60%) as a preferential distribution channel. Unlike company 1, for company 2 there was a complete alignment with the strategies expected downstream the literature analysis for all the case studies conducted.

The results reported for Company 3 and 4 are similar to those of Company 2. Both companies produce the same type of products and implement for all the case studies 11, 12 and 13 an SCS in harmony with those suggested and expected by the literature.

Company 5 represents an icon for luxury fashion women's footwear. As far as sales channels are concerned, the company owns 126 mono brand shops located all over the world. The ecommerce channel still has rather limited sales volume and is not exploited by the company compared to the more classic retail channel, which is comprises 60% of total sales. For the case studies 14 and 15 and in particular for footwear, both the carry over and seasonal lines are managed by a hybrid SCS and an MTO type of production. Overall, 80% of the collection comprises seasonal products, with the remaining 20%, composed of carry over products. From the data, it is evident that this company is focused on the development of new products to meet the needs of the customers for each season.

To sum up, company 5 belongs to the luxury fashion market segment that uses the retail channel as the preferred SCS channel (60%). It is possible to highlight a misalignment for the case study 14 since for footwear type of product and carry over product line, company 5 implements a Hybrid strategy through an MTO type of production contrary to what the literature suggests. On contrary, it is possible to find an alignment for the case study 15 with the SCS suggested by the literature in particular for footwear product type and seasonal line, company 5 implemented a Hybrid SCS with an MTO type of production.

Finally, company 6 is a world-renowned brand operating in the luxury fashion sector. The core business of the company is represented by men's and women's clothing, comprising 70% of the turnover, with the remaining 30% relegated to accessories, particularly leather goods, which have been expanding in recent years. The distribution channels used by the company 6 are both retail and wholesale, respectively, comprising 45% of the sales volumes. In company 6, five different case studies have been conducted (from 16 to 20) for the three different types of products managed by the company. These include permanent products, such as the traditional trench coat, waterproof coats and various bags and scarves, which have a long-life cycle. In addition, there are the carry over products with a life-span of least one year and finally, the seasonal products that are designed to respond to the customers' needs. A total of 40 and 60% of the collection are composed of carry over products and seasonal products, respectively. The company produces all products externally, ordering finished items directly from the suppliers. For the permanent products, the supplier's stock both RM and FP as sales forecasts for these products have a time span of two years. Also, seasonal products are made by suppliers who exercise freedom to purchase RM. In case study 17 and 18 for both product types and carry over line are made with an MTO type of production. No stock is produced because in case the products remain unsold, the supplier is directly responsible for the goods left in the warehouse.

As before, the company 6 uses the retail channel as the preferred SCS channel (80%).

For the case study 16, there is total alignment with the strategies expected after the literature analysis. In particular, it is possible to highlight how a lean strategy was implemented for products that are permanent fixtures on the market. In case study 17 and case study 18 considering different type of product (RTW and leather goods) but the same line of product (Carry Over), for both case, there was a misalignment with the literature. For these cases, a Hybrid strategy was implemented instead of a lean strategy suggested by the literature, as for the seasonal products (case studies 19 and 20), because if left unsold, the supplier is responsible for the goods remaining in stock.

3.2 “Cross case” analysis of the SCS – RQ1

In order to carry out this analysis a “clustering” of the sample was made. The clustering was obtained distinguishing between companies that produce leather goods (bags) and footwear with leather as the RM and companies that produce the RTW as the principle product. This type of cluster was created because the SC of leather goods and footwear are very similar, characterized by a network of local suppliers and production planning that are connected to leather as the main RM. Instead, the SC of RTW is very different because is much more extensive and has a production planning that is linked to all components and not only to the availability of the principle RM.

As can be seen in Table 6, for both types of clusters identified, the only product line not aligned with the expected SCS is the carry over (case study 2, 14, 17, 18). Instead, the seasonal line was always shown to be aligned with the expected SCS, regardless of the type of product being produced. It is possible to conclude that the only driver that affects the alignment between the expected and implemented SCS, respectively, is the product line.

It can be noticed that carry over products are managed by most companies using a hybrid strategy and MTO type of production. The reasons why the companies examined opted for such strategy are attributable to the following factors:

1. “The forecast of the demand of carry over products is affected by high error” as mentioned by the SC manager of the company 1 for the case studies #1 and #2. In fact, the accuracy of the predictions is affected by the number of years the product remains on the market. For this reason, it is feasible to make reliable predictions when dealing with iconic and stable products that remain on the market for many years. This scenario is very different for seasonal products that become carry overs from one season to the next. Market predictions for these carry over products have duration equal to one year. The high value of the RM (especially fine leathers or accessories) necessitates that companies store such RM inside warehouses that have high purchase and management costs.

2. Also, the SC manager of the case study #5 and #14 for footwear type of product said that “There are difficulties in managing sizes when it comes to clothing or footwear products”.

3. For the case studies # 1, # 2, # 3, # 4, #14 and #15 there is also increased complexity in the simultaneous management of different types of productions according to the product category (carry over, seasonal and capsules)

4. Moreover, both the SC and IT manager of the case studies # 5, # 6, # 7, # 17 and # 18 said that “To pursue the saturation of the façonists, companies are ordering the production of carry over”.

3.3 “Within case” analysis of the KPIs – RQ2

Among the KPIs selected in the literature, the main performance parameters measured by the companies were identified. In a second step, the alignment between the KPIs actually measured by the six companies comprising the case studies and those identified in from literature classification of KPIs within a single strategy were verified.

Below, in Table 7, the results that emerged from the individual case studies carried out are reported:

From the results of the single case analysis it was found that depending on the product line of the brand, it was possible to identify different approaches to KPIs measurement.

3.3.1 Carry over products

For all companies implementing this product line, regardless of the SCS implemented, there was a partial alignment with the cost KPIs characteristic of a lean strategy, expected after the classification realized and the indicators measured by companies. Within the design phase, the companies measure only one cost KPI among those identified. This aspect was justified by the number of years the product has been on the market. Therefore, for a structured product, subject to only occasional changes, it is less important to monitor KPIs related to the prototyping and engineering phase. These indicators were measured only in the first year, following the introduction of the product, and possibly in the following years, when the product was classified as seasonal. As far as the plan phase was concerned, the companies in the sample only measure the indicator related to the “total supply chain cost”. For this type of product, the demand is more stable and therefore indicators such as “accuracy of forecasts” and “volatility of forecasts” are not subject to continuous monitoring and measurements.

There was a complete alignment with the source, make, delivery and return phases and the cost KPIs characteristic of a lean strategy expected after the classification realized and the indicators measured by companies. An aspect that emerges from the source and production phases is that the attention of some of the companies is addressed, not only to the monitoring of costs, but also to the quality of the RM and the product produced. This aspect is confirmed from the luxury fashion brands examined, indicating that the companies view both the quality of the RM and FP as an essential value for the product produced. In these aforementioned phases of the process, companies monitor lead time indicators. This is because the company implements continuous process improvement techniques in order to streamline the life cycle and reduce the product throughout time. Specifically, the parameters related to quality (% compliant orders received) and lead time (% of orders received on time compared to the requested date, % of orders received with the correct quantity requested, % orders with change in delivery date, # order change requests satisfied, # accessible information/# total information) are monitored and reported by companies, in addition to the cost KPIs for the source and make phases.

3.3.2 Seasonal products

For seasonal and capsule products, the KPIs are fully aligned with the cost indicators characteristic of a lean strategy, as could be expected from the classification of KPIs obtained from the literature those indicators measured by companies. For this type of product, all the KPIs of new product development (NPD) are measured and considered very important by the companies. This is attributable to the fact that the costs related to the development and the industrialization of new prototypes have a significant impact on the total cost of the collection.

Within the plan phase, the companies in the sample study measure all the KPIs very carefully because the demand for these products is both uncertain and difficult to predict. Within this phase, some purchase orders of “critical” RMs or production orders are made in advance, especially in the case of commercialized products. The reason why these orders are made is twofold. This is related to the limited availability of some RMs and the high supply lead time, often not compatible with the market demands.

For the source phase, the sample companies monitored all the KPIs related to the service level present in the set of indicators characteristic of an agile strategy identified from the literature analysis. In particular, among the service level KPIs measured, are those pertaining to the capacity of the supplier to change both the delivery date and the quantity ordered, after the first issue of the order by the brand. For some of the companies in the case study, the supplier's ability to meet the requirements for small batches of materials and to be willing to work with the brand in order to develop new products is critical. It is important to emphasize that KPIs related to the service level are influenced by parameters such as lead time and quality (OQ for an agile SCS) in both the source and make phases. Companies are often compelled to change their production planning either because of delays in delivery times or because of non-conformity relating to the RM. In the case in which the suppliers of RM (like leather or precious materials) deliver a material with high lead supply time and with non-conformities, the combination of these two parameters could impact both on production time and, as a result, on the service level.

In the source phase, some companies also monitor KPIs relating to costs including the total cost of the RM, warehouse and obsolescence. In particular, companies also monitor the cost of the RM that they purchase from additional suppliers. This serves to always guarantee the availability of RM and to obtain better prices, attributable to the competition between the various supply companies.

The KPIs monitored in the production phase by the companies include “production efficiency”, but also the “% of non-conformity on the total product”, the “% of available capacity” (both internally and at the façonists) and the “% of orders with delivery date variation during the production lead time”.

Having to guarantee a high quality of the FP, “production efficiency” is a particularly important KPI, and is monitored by the company. This KPI is translated in specific measures related to the quality of the FP but also on the capacity of the suppliers.

With regard to the service level KPIs, similar to that of the sourcing phase, companies monitor the “ability to adapt to the variation in demand” both in terms of priority on orders and in terms of quantities ordered by suppliers. In the luxury fashion sector, the ability to correctly predict demand is a fundamental requirement for all types of products, especially when the production of seasonal products is fundamental. For seasonal products, it is very important to monitor the capacity of suppliers in order to guarantee the same service level in the event that delivery dates and quantities are subject to variation. These variations are due to the continuous alignment between the sales predictions and the purchased orders.

For the delivery and return phases, managed in an agile mode by all six companies, the main indicators included the “% shipments managed on time compared to the customer's request”, the “% compliant orders processed” and the “average time associated with the receipt of a product to the store”. The principle KPIs measured in the retail phase are closely linked to the preferred sales channel implemented by the company. All six companies in the sample study employed the retail channel as the predominant channel. As a consequence, attention was primarily focused on the performance of the stores, and in particular on the KPIs such as the success rate of new products, coined by the phrases “Sell In”, “Sell Out” and “Sell Through”.

3.4 “Cross case” analysis of KPIs – RQ2

Following the single case analysis, a cross case analysis of the companies was then carried out.

To perform this analysis, we used the same clusters proposed in the cross-case analysis for DR1. In order to verify the alignment of the companies with the KPIs expected by the literature analysis, the percentage of indicators measured by each company with respect to the total KPIs proposed was calculated. The analysis was carried out for each company and for each type and product line produced by the respective company. In particular, the first step was to verify the alignment between the KPIs measured by the companies, for each product line, and the expected KPIs identified in the classifications obtained from the literature analysis. There was a total alignment (100%) of the KPIs for the seasonal product line and capsule products for both Cluster 1(composed by the companies producing bags and footwear) and Cluster 2 (composed by all the companies that producing RTW) For the carry over product line, a partial alignment of 76% was found compared to the KPIs presented in the literature analysis (see Figure 3).

Subsequently, a second analysis was carried out, in each cluster, for the carry over product line (for which a partial alignment was found). The percentage of indicators monitored for each phase of the process was then calculated. The first phase of the process examined was the new product development. It was shown that all companies producing the carry over product line, for both clusters, measured only one indicator compared to the five indicators identified for this phase in the literature analysis. This was suggested to have contributed to 8% of the misalignment. The second phase of the process examined was the plan phase. The analysis showed that even within this phase the companies of both clusters measured a lower number of KPIs than those presented within literature analysis. Both clusters for the carry over product line were shown to have measured only one KPI of the three proposed indicators. In turn, this would have contributed to 16% of the misalignment. Hence, the results showed the misalignments reported were attributed only to these two phases within the process, given that the remaining phases were completely aligned with all the KPIs proposed by the literature analysis.

4. Discussion

The results allow us to provide an answer to the two research questions posed and to discuss the results against the literature. We focus our analysis on the differences between carry over, seasonal and capsule collection because this product line classification is widely used also in practice, thus creating a bridge among academic research and the industry.

As far as RQ1 is concerned, results suggest that there is a misalignment between what literature suggests and what companies implement for the carry over products, while literature is confirmed for seasonal product. In particular, although the literature suggests a lean strategy for carry over products (Brun et al., 2008), this strategy is not always followed by companies. Carry over products are managed by most companies using a hybrid strategy, differently from what is suggested by the literature, i.e. lean (Bruce et al., 2004).

The analysis of the reasons why the companies examined opted for such strategies allow us to provide some contributions to the literature in the field.

First of all, the demand of carry over products is not always easy to forecast, differently from what the literature (Caniato et al., 2008) suggests. Indeed, the accuracy of the predictions is affected by the number of years the product remains on the market. For this reason, it is feasible to make reliable predictions when dealing with products that remain on the market for many years, i.e. those called permanent (Rigaud-Lacresse and Pini, 2017). This is not the case of those seasonal products that become carry over, for which therefore demand forecast is based on limited sets of data. Moreover, forecast accuracy is affected by the number of different sizes to forecast, and they can be an issue when it comes to clothing or footwear products.

Secondly, literature is suggesting a different SCS for each product family (Brun and Castelli, 2008). However, companies are managing portfolio of product families. So, to reduce the complexity in the simultaneous management of different types of SCS according to the line of product, they tend to not apply the SCS in its pure form. This might result in a reduction in the efficiency of the management of each product line, but in an increase in the global efficiency. This is the case, for instance, of the use of carry over as products to fill the production capacity of the façonists to compensate the low volumes of production of new products, which is strongly seasonal.

As far as RQ2 is concerned, the KPIs monitored confirm the suggestion by literature for any product line, but for carry over for the KPIs for NPD and plan process. This misalignment with the literature can be due to the nature of the product produced. The carry over line is by definition a product line that is sold season after season until market demand runs out (Caniato et al., 2008). For this reason, NPD indicators are occasionally measured on sample products. In addition, the planning phase of this product line is carried out on the basis of the average sales trend of previous seasons given that the product has been on the market for several years. Indicators such as uncertainty and volatility of demand do not need to be constantly monitored as unsold products can be left in stock and disposed of in the sales of the following season.

These results can be used to build a decision-making tool that build upon literature (e.g. Brun and Castelli, 2008) to propose companies a tool for selecting the SCS and the KPIs to monitor its deployment. First, the company defines the specific theoretical SCS for each combination of brand positioning, line of product, type of product and distribution channel, then, considering portfolio effects, the applied SCS is defined. Finally, based on the SCS chosen, the set of KPIs to be used in each process are selected, as suggested in Table 8Table 9.

5. Conclusions and future developments

This article stems from the desire of companies in the luxury fashion industry, which operates within a highly dynamic context, to implement SCS strongly linked to action. Quite often, it is the inability to connect SCS to the actions, and the consequent lack of attention in the implementation phase, that result in failure pertaining to the concretization of initiatives and ideas, even if supported by apparently coherent plans.

Within this context, the request to monitor KPIs arises in order to understand whether the business strategy has been translated into operational levers. The difficulty encountered by companies resides in identifying which KPIs need to be measured, according to the SCS implemented by the company.

Therefore, the objective of this article was to investigate the alignment between the SCS implemented within fashion luxury companies and the KPIs monitored within them.

The requisite of the present study was to provide answers to the two aforementioned research questions.

1. Regarding RQ1, although the literature suggests adopting a lean strategy for carry over products, the case studies have shown that this type of strategy is not always adopted. As far as seasonal products are concerned, a hybrid SCS, consisting of a lean phase, followed by a subsequent agile phase, appears to be the approach adopted. This result was confirmed from both the single case analysis and by the cross-case analysis. Finally, the capsule product was shown to follow the same strategy used for seasonal products, despite the fact that the literature recommends an agile SCS for these types of products that focuses on reactivity and flexibility in order to meet customer demand.

2. From the RQ2, through the literature analysis, it was possible to identify the KPIs measured by companies and to subsequently integrate them with those present in the SCOR and DCOR model. These indicators were associated with the strategies identified in DR1 and compared with those monitored by the companies involved in the case study through the use of questionnaires and direct interviews.

The scope and results of the overall research are broader than those presented in the previous sections. In terms of managerial implications, one of the most interesting contributions is the identification of the product line as a driver influencing the SCS choice and the different approaches to measuring KPIs. Managers could adopt the listed elements in order to define the most suitable SCS for the specific targets of the brand. Since this decision is strictly related to the company positioning on the market, it is important to fully understand and identify the set of drivers that the brands pursue on the market. As a consequence, it is possible to evaluate the actual SCS, in terms of structure and practices. This will permit the company to understand in which aspects it is aligned toward the KPIs identified in literature and where not, allowing those aspects to be revised.

A limitation of the present study is related to the sample dimension and composition. While case studies permitted us to understand and compare different SCS in the fashion luxury industry, a wider sample should be applied in order to evaluate different sectors, as the fast fashion or the mass market. On the other hand, sample could be extended including other luxury segments, as jewels, boat or cars. Despite this fact, the present study is a starting point for further research in these directions. Last, the study and the sample could be extended to include additional types of fashion luxury products in order to understand different approaches to SCS and KPIs derived from different type of distribution channel and product typology.

Figure 1

Literature review strategy

[Figure omitted. See PDF]

Figure 2

KPI selection

[Figure omitted. See PDF]

Figure 3

Results of RQ2 cross case analysis

[Figure omitted. See PDF]

Table 1

Supply chain strategies

Strategy	Characteristics	References
Lean	Cost of purchased materials Material transport cost Customs material cost, duties, taxes and tariffs	SCOR (CO.2.003)	L
S	Defective inventory percentage	SCOR (AM.3.28)	A
S	Delivery item accuracy	SCOR (RL.3.33)	A
S	Delivery quantity accuracy	SCOR (RL 3.35)	A
S	% orders received on time to demand requirement	SCOR (RL 3.20)	A
M	Total production costs Production labor costs (direct) Cost of automation Cost of productive assets Production management costs	Beamon (1999), Cai et al. (2009), Shepherd and Günter (2011), Thilakarathna et al. (2015), SCOR (CO.2.004)	L
M	Production lead time	Beamon (1999), Chan (2003), Shepherd and Günter (2011), Thilakarathna et al. (2015)	N/A
M	# Stock keeping unit (SKU)	Shepherd and Günter (2011), Gunasekaran et al. (2004), Bhagwat and Sharma (2007)	N/A
M	Productivity	Gunasekaran et al. (2004), Kumar Hurreeram et al. (2014)	L
M	Production capacity	Bhagwat and Sharma (2007), Gunasekaran et al. (2004), Otto and Kotzab (2003), Kumar (2013), Kumar Hurreeram et al. (2014)	L
M	% of non-compliance on total products	Gunasekaran et al. (2004), Sellitto et al. (2015), Kumar Hurreeram et al. (2014)	A
M	Make cycle time	Yeh et al. (2010), Pinna et al. (2017), SCOR (RS.2.2)	N/A
D	Inventory turnover	Lapide (2000), Shepherd and Günter (2011), Otto and Kotzab (2003), Battista and Schiraldi (2013), Papakiriakopoulos and Pramatari (2010)	N/A
D	Transportation costs	Beamon (1999), Cai et al. (2009), Chan (2003), Gunasekaran et al. (2004), Kumar Hurreeram et al. (2014), Otto and Kotzab (2003), Shepherd and Günter (2011), Thilakarathna et al. (2015), SCOR (CO.3.022)	L
D	Customer response time (service level)	Beamon (1999), Bhagwat and Sharma (2007), Cai et al. (2009), Chan (2003), De Felice and Petrillo (2013), Gunasekaran et al. (2004), Lapide (2000), Otto and Kotzab (2003), Sellitto et al. (2015), Shepherd and Günter (2011)	A
D	# of on-time deliveries	Beamon (1999), Chae (2009), Chan (2003), De Felice and Petrillo (2013), Lapide (2000), Sellitto et al. (2015), Shepherd and Günter (2011), Thilakarathna et al. (2015)	A
D	Distribution Lead time	Shepherd and Günter (2011), Gunasekaran et al. (2004)	N/A
D	% of urgent deliveries	Gunasekaran et al. (2004), Bhagwat and Sharma (2007)	A
D	# stock out	Beamon (1999), Cai et al. (2009), Chan (2003), Otto and Kotzab (2003), Shepherd and Günter (2011), Thilakarathna et al. (2015)	L
D	# back order	Beamon (1999), Shepherd and Günter (2011), Thilakarathna et al. (2015)	L
D	# lost sales	Mattila et al. (2002)	L
D	Sell-thought (%)	Beamon (1999), Cai et al. (2009), Mattila et al. (2002), Moore and Fairhurst (2003), Shepherd and Günter (2011)	L
D	New product sales rate	Cai et al. (2009)	A
D	% of orders delivered in full (quantity)	Beamon (1999), SCOR (RL 2.1)	A
D	% defect free conformance orders delivered	SCOR (RL 3.42)	A
R	% of products replaced (non-compliant)	Mattila et al. (2002), Sellitto et al. (2015), Kumar Hurreeram et al. (2014)	A

Note(s): NPD: new product development; P: plan; S: source; M: make; D: delivery; R: return; L: lean; A: agile; N/A: not applicable

Table 4

Sample

Company	country	No. of lines	Main product	Age of the company (years)	Turnover	No. of SKU	Roles of the people interviewed
1	Spain	4	Footwear	60	239,2 M€	500	SC manager
			Leather goods – bags			<# SKU<
						1,000
2	Italy	10	Footwear	90	1 B€	>1,000	SC manager, IT manager
			Leather goods – bags
3	Italy	7	Leather goods – bags	>100	3.2 B€	>1,000	SC manager
4	Italy	25	RTW –tailored clothing for men	41	91,8 M€	500	SC manager, IT manager
						<# SKU<
						1,000
5	French	1	Footwear	30	1 B€	500	SC manager
						<# SKU<
						1,000
6	England	15	RTW	>100	2.7 B£	>1,000	SC manager, IT manager
			Leather goods – bags

Table 5

RQ1 single case analysis

Company	Case study	Type of product	Line of Product	SCS expected	SCS implemented	Note
1	1	Footwear	Carry over	LeanMTS	HybridMTS(RM)MTO(FP)	Optimization in the purchases of the RM Levelling the production capacity of façonists RM stock management on order history
	2	Leather goods – bags
	3	Footwear	Seasonal	Hybrid/AgileMTO		Management of the stock of RM through forecast orders
	4	Leather goods – bags
2	5	Footwear	Carry over	LeanMTS	LeanMTS	Planning on a 6-month forecast by cross-referencing data from marketing with real demand
	6	Leather goods – bags
	7	RTW
	8	Footwear	Seasonal	Hybrid/AgileMTO	HybridMTO	Production flexibility Low volume of each SKU
	9	Leather goods – bags
	10	RTW
3	11	Leather goods – bags	Carry over	LeanMTS	LeanMTS	RM stock management based on order forecast Complete procurement of RM only after the sales campaign and order collection
	12		Seasonal	Hybrid/AgileMTO	HybridMTO	No stock either at the RM or FP level Sourcing of RMs only after sales campaign and order collection
4	13	RTW	Seasonal	Hybrid/AgileMTO	HybridMTO	No stock either at the RM or FP level Sourcing of RMs only after sales campaign and order collection
5	14	Footwear	Carry over	LeanMTS	HybridMTO	Stock management at the RM and FP level Use of carry over to fill suppliers' production capacity
	15		Seasonal	Hybrid/AgileMTO		Stock management at the RM and FP level Reactivity to customer needs
6	16	RTW	Permanent	LeanMTS	LeanMTS	Stock management of RM and FP by suppliers Management of FP stock at point of reorder based on sales forecast from the previous period
	17	RTW	Carry over	LeanMTS	HybridMTO	Stock management of RM by suppliers Stock management of FP at central warehouse Orders launched based on sales campaign
	18	Leather goods
	19	RTW	Seasonal	Hybrid/AgileMTO
	20	Leather goods

Note(s): RTW: ready to wear; RM: raw material, FP: finish product, MTO: make to order, MTS: make to stock; SKU: Stock Keeping Unit

Table 6

Cluster according to the product lines

Cluster	Carry over		Seasonal
	Aligned	Not aligned	Aligned	Not aligned
RTW	7, 16	17	10, 13, 19
Bag and footwear	1, 2, 5, 6, 11	2, 14, 18	3, 4, 8, 9, 12, 15, 20

Table 7

Results of RQ2 single case analysis

Table 8

KPIs measured according to the SCS

Line of product		Carry over	Carry over	Seasonal/capsule
SCS		Lean	Hybrid	Hybrid
Process	NPD	Total cost of new product development	Total cost of new product development	Total cost of new product development
				Total cost of the realized prototypes
				Total cost of samples produced
				Total cost of prototypes/total production cost (%)
				Total industrialization cost/production cost (%)
	PLAN	Total cost of supply chain management	Total cost of supply chain management	Total cost of supply chain management
				Forecast accuracy
				Forecast volatility
	SOURCE	Total supplier-related costs	% of orders received on time compared to the requested date	% Of orders received on time compared to the requested date
		Total RM cost
		% Of orders received on time compared to the requested date	% of orders received with the correct quantity requested	% Of orders received with the correct quantity requested
		% Orders with delivery date variation (during the lead time associated with the supplier)	% Orders with delivery date variation (during the lead time associated with the supplier)	% Orders with delivery date variation (during the lead time associated with the supplier)
		% of compliant orders received	# Change order requests fulfilled	# Change order requests fulfilled
		Total inventory costs	% of compliant orders received	% of compliant orders received
		Obsolescence costs	# accessible information/# total information	# accessible information/# total information
		Returned transport cost	Total RM cost	Total RM cost
		RM shipping cost
	MAKE	Total production cost	Productive effectiveness	Productive effectiveness
		Productive effectiveness	% non-conformity on total product	% non-conformity on total product
		% non-conformity on total product	% available capacity	% available capacity
		% Orders with delivery date variation during the production lead time	% Orders with delivery date variation during the production lead time	% Orders with delivery date variation during the production lead time
		Capacity saturation	Time to market	Time to market
	DELIVERY	% Of shipments handled on time compared to customer request	% Of shipments handled on time compared to customer request	% Of shipments handled on time compared to customer request
		% Compliant orders filled	% Compliant orders filled	% Compliant orders filled
		Average time associated with receiving a product at the store	Average time associated with receiving a product at the store	Average time associated with receiving a product at the store
		Service level	Service level	Service level
		New product success rate	New product success rate	New product success rate
	RETURN	# Products returned for non-compliance	# Products returned for non-compliance	# Products returned for non-compliance

Note(s): RM: raw material

Process	Performance measures	Utilization		How much important is it in a scale from 1 to 5?
		Yes	No	1 = not important	2 = low importance	3 = moderately important	4 = important	5 = very important
New product development	Total cost of new product development
	Total cost of the realized prototypes
	Total cost of samples produced
	# Of new models planned
	# Of new products introduced
	Average time associated with new product changes
Plan	Cost of supply chain management
	Forecasting technique accuracy
	Volatility of forecasts
	Cost of supply chain management
Source	Supplier service level (product compliance, quantity)
	Supplier lead time
	Select supplier and negotiate cycle time
	Total cost of raw materials/semi-finished products/finished warehouse products
	Warehouse obsolescence costs
	Inventory level of raw materials/semi-finished products/finished products
	% Orders received on time to demand requirement
	% Of wrong deliveries (quantity, non-conforming products)
	Identifying sources of supply cycle time
	Verify product cycle time
	Material landed cost cost of purchased materials, material transport cost, customs material cost, duties, taxes and tariffs
	Return cost
	Defective inventory percentage
	Delivery item accuracy
Make	Total production costs production labor costs (direct), cost of automation, cost of productive assets, production management costs
	Production lead time
	# Stock Keeping Unit (SKU)
	Productivity
	Production capacity
	% of non-compliance on total products
	Make cycle time
Delivery	Inventory turnover
	Transportation costs
	Customer response time (service level)
	% of on-time deliveries
	Distribution lead time
	% of urgent deliveries
	# stock out
	# back order
	# lost sales
	Sell-thought (%)
	New product sales rate
	% of orders delivered in full (quantity)
	% defect free conformance orders delivered
Return	% of products replaced (non-compliant)