Towards a holistic view of customer value creation in Lean: A design science approach

Abstract While “identifying customer value” is the first principle of Lean thinking, the concept of customer value has largely remained unchanged in the Lean discourse - quality, cost and delivery. This research examines the problem of working from such an internal process point of view in today’s highly dynamic world, where customer needs get increasingly more sophisticated. Aiming to contribute to solving this problem, this paper develops a new artifact, the so-called Customer Value Matrix (CVM) Instrument by bridging Operations Management, Strategy and Marketing literatures and following the design science approach. This paper builds the CVM instrument based on five theoretically grounded design principles, which are derived from a systematic review of 49 articles, and proposes three empirically grounded technological rules as a result of multi-stage evaluations. Supporting Service-Dominant Logic and Lean Consumption views, the CVM instrument can be considered as a novel addition to the Lean toolkit and enable managers to view what the customer truly values in a more holistic fashion and, as a result, promote a balanced view of process and customer focus.

Towards a holistic view of customer value creation in Lean: A design science approach Erdoğan Gülyaz 1 *, Jack A.A. van der Veen 1 , Venu Venugopal 1 and Sam Solaimani 1 Abstract: While "identifying customer value" is the first principle of Lean thinking, the concept of customer value has largely remained unchanged in the Lean discourse -quality, cost and delivery. This research examines the problem of working from such an internal process point of view in today's highly dynamic world, where customer needs get increasingly more sophisticated. Aiming to contribute to solving this problem, this paper develops a new artifact, the so-called Customer Value Matrix (CVM) Instrument by bridging Operations Management, Strategy and Marketing literatures and following the design science approach. This paper builds the CVM instrument based on five theoretically grounded design principles, which are derived from a systematic review of 49 articles, and proposes three empirically grounded technological rules as a result of multi-stage evaluations. Supporting Service-Dominant Logic and Lean Consumption views, the CVM instrument can be considered as a novel addition to the Lean toolkit and enable managers to view what the customer truly values in a more holistic fashion and, as a result, promote a balanced view of process and customer focus. ABOUT THE AUTHOR Erdoğan Gülyaz is a Senior Project Manager -New Product Development (NPD), working for a large multinational organization. He pursued a PhD in 'Lean and Competitiveness'. Jack van der Veen is Professor of Supply Chain Management and holder of the evofenedex Chair for SCM at Nyenrode Business University. Venu Venugopal is a Professor of Supply Chain Operations at Nyenrode University. Sam Solaimani is Associate Professor at Nyenrode Business University.
As a research group, Erdoğan, Jack, Venu and Sam focus on areas including Lean management, SME Competitiveness, Value creation and appropriation, Supply Chain Collaboration and Innovation Management.
This research is part of a broader PhD research where Erdoğan investigated how organizations can achieve competitiveness through Lean and proposed new ways of creating and capturing value in a Lean system.

PUBLIC INTEREST STATEMENT
For organizations that aim to achieve operational excellence, Lean has long been a powerful method. Unfortunately, associated efforts have the tendency to become internally focused. However, in today's dynamic world, centering processes around the customer is more important than ever. Although Lean embraces the idea of creating customer value, it can be observed that the available tools in this regard are somewhat limited and do not always adhere to the new requirements.
In this paper, a new tool, coined the Customer Value Matrix (CVM) Instrument, is proposed as an addition to the Lean toolbox with the objective to facilitate organizations to identify and capture customer value. Combining important new insights from Strategy and Marketing with Operations, the CVM can help managers of different functions to get a holistic and structured view on what customer value entails and provides an associated Lean-based method to enhance customer value and become more competitive.

Introduction
In the post-economic recession environment where efficient use of resources is paramount, more and more organizations across a wide range of sectors are embarking on the Lean journey along its mantra "creating customer value by reducing waste". While creating value focuses on retaining and attracting more customers thereby increasing revenue, eliminating waste helps companies improve their efficiency thereby reducing cost. However, according to a large survey, a majority of the firms (an alarming 74%) failed to extract the expected value from Lean and are not making good progress (Pay, 2008). One possible explanation would be that for many companies the major focus of Lean implementation is still at the shop floor level without any "strategic approach", which led to many unsustainable Lean transformation programs (Hines, Holweg, & Rich, 2004). Not only production-oriented organizations, but also many service organizations that adopt Lean, typically tend to work from an internal, cost-focused point of view (Seddon & O'donovan, 2010). As efficiency is becoming a necessary order qualifier (MacBryde et al., 2013), many organizations might be adopting Lean, in fact, to avoid the "competitive disadvantage" coming from inefficiency.
Although continuous customer value creation is the core of Lean philosophy, Lean practice and theory development over the years appears to have gained a somewhat one-sided approach, i.e., focusing on the firm's exploitative capability. March (1991) relates exploitative capability to efficiency, routinizing and performing better in meeting the existing needs of customers under given constraints. Many Lean firms arguably strive to provide the same value for less money or more value for the same money. In other words, in everyday Lean practice, there appears to be a strong focus on efficiency (on "doing the job right") and a weaker concern about effectiveness ("doing the right job"). In today's highly dynamic world, where customer needs change rapidly, focusing too much on efficiency can eventually reduce a firm's offerings to commodity level.
Several scholars argue that firms need to be ambidextrous, i.e., to develop a balanced view of exploitation (e.g., increasing efficiency, improving quality) and exploration (e.g., searching creative ways to satisfy customer needs, explore new markets) if they are to survive and prosper (e.g., Birkinshaw & Gibson, 2004;He & Wong, 2004;Lubatkin, Simsek, Ling, & Veiga, 2006;March, 1991). Similarly, Radnor and Johnston (2013) stress that although process focus is essential to be more efficient, Lean firms need to develop a more balanced process and customer focus in service environments.
Understanding, creating and managing customer value (CV) has been long seen by many marketing scholars as the cornerstone of a firm's business strategy (e.g., Almquist et al., 2016;Porter, 1996;Slater & Narver, 1994;Verhoef & Lemon, 2013;. Over the years, the marketing perspective on CV has evolved. For instance, Service-Dominant logic (S-D logic) stream (e.g., Grönroos & Voima, 2013;Vargo & Lusch, 2004), which emphasizes the importance of intangibility, exchange processes, and relationships in value offerings, and customer experience stream (e.g., Verhoef et al., 2009), which stresses myriad touch points in multiple channels, become the new frontiers in CV research. Despite such advancements in CV research, CV concept for many Lean practitioners and scholars has largely remained unchanged: providing customer right quality with lower cost when needed (Womack & Jones, 1996), also known as Quality Cost Delivery (QCD). Typically in Lean, value is defined by the ultimate customer in terms of a specific product or service, which meets the customer's needs at a specific price at a specific time (Womack & Jones, 1996) and it is created through a series of value-added steps (Liker, 2004). This view has three limitations when compared to contemporary marketing literature: (i) CV can be expressed by the customer; (ii) CV is embedded in a product or service through a series of value-added steps; and (iii) CV meets the practical needs of a customer. The foregoing three limitations can blindside any firm (and particular Lean firms as they have a limited view on CV) to realize full value potential in customer's consumption space. The problem is not the key ideas of Lean management, but the lack or incompleteness of tools and techniques in the Lean toolkit that can enable practitioners to adopt true customer perspective and capture the holistic notion of CV.
This study bridges operations and marketing disciplines to design a novel management instrument to capture CV within a Lean system, so-called Customer Value Matrix (CVM) Instrument. The CVM Instrument aims to provide a holistic view of customer value creation through encouraging explorative thinking, i.e., revealing blind-spots in the CV understanding and looking for creative ways to meet customer needs, which many Lean organizations seem to be missing.
The remainder of this paper is organized as follows. First, the research method, Design Science Research (DSR) is explained. Second, the key CV characteristics in literature are systematically reviewed to develop an initial solution design. Third, based on these CV characteristics, the main design principles are derived and the CVM Instrument is built upon. Fourth, the CVM is evaluated through multiple stages. Next, the discussions about the findings are provided. Finally, the conclusions, limitations and several suggestions for future research are given.

Research approach: design science research (DSR)
To iteratively design and evaluate the CVM Instrument, this research uses the DSR approach (Hevner, March, Park, & Ram, 2004;Holmström, Ketokivi, & Hameri, 2009;Van Aken, 2004). The design science paradigm seeks to extend the boundaries of organizational capabilities by creating new and innovative artifacts, and by focusing on how well the artifacts work rather than why they work (Hevner et al., 2004). Similarly, Van Aken (2004) states the mission of a design science is to develop knowledge for the design and realization of artifacts, i.e., to solve construction or improvement problems.
In DSR, March and Smith (1995) highlight two main activities, building and evaluating. An artifact is built to perform a specific task and evaluated to determine if any progress has been made. Throughout this study, the artifact that is worked on is referred as the CVM Instrument, and in line with March and Smith (1995), the artifact combines a model and a method of how to use the model. Inspired by Vaishnavi and Kuechler (2008) as well as the design procedures provided by Hevner et al. (2004), Holmström et al. (2009) and (Gregor & Hevner, 2013) the development and evaluation of the CVM Instrument followed the following five steps.

Solution design (Section 3)
To build the CVM Instrument on a theoretical foundation, the CV literature is systematically reviewed. Holmström et al. (2009) refer to this phase as "solution incubation" in which the CV knowledge domain is scanned to develop an initial solution design. In this phase, the main CV characteristics are identified.

Building the CVM (Section 4)
Based on the main CV characteristics, this research develops several design principles to build the CVM Instrument. The design principles and the CVM construction have gone through multiple thought processes and trials, during which solution design and building phase are intertwined.

Evaluation of the CVM (Section 5)
Once the instrument has been constructed, it is evaluated based on explicit criteria. Since designing is inherently an iterative and incremental activity, the evaluation phase provides essential feedback to the building phase (Hevner et al., 2004). Typically, during the evaluation phase, the strengths and weaknesses of the instrument are noted for the next iteration.

Discussions results (Sections 6 and 7)
In this step, theoretical generalizability of the findings is sought and the technological rules, linking general knowledge to the desired outcome (Van Aken, 2004), are outlined. The technological rules are derived from the build-evaluate iterations.

Solution design: systematic review of CV literature
In order to identify the key CV characteristics, a systematic literature review was conducted (Tranfield, Denyer, & Smart, 2003). Given the high number of publications in CV, the review focused on high-impact articles. First, we selected three-and four-star journals according to ABS 1 journal rating system. Second, to make sure the influential works are not missed, Google scholar citations are used, leading to more than 100 Google citations. In all, the keyword search, screening, snowballing and filtering yielded 49 publications for the full review. The selected papers were systematically reviewed, based on the following queries: How is CV defined? What are the key characteristics of CV? What are the proposed CV models? A careful exploration of the commonalities and differences observed in the literature resulted in nine key CV characteristics that are explained in the sub-sections below.

The nine key characteristics of CV
To identify the characteristics, the various concepts discussed in the publications were coded. Following Miles and Huberman (1994), the coding process started with manageable broad categories and extended/adapted every time new categories were added. Next, the study continued to explore new categories and identify important arguments/evidence for each category. In all, nine distinctive categories have been identified, which from here on will be referred to as CV characteristics. Those characteristics are briefly discussed below (see Appendix A for the evidences from the selected CV literature for each characteristic).

CV is subjective
Almost all scholars agree on the subjective and individualistic nature of CV; it varies from one individual to another . In other words, a product that is highly valuable to one customer may have little value to another.

CV has multiple attributes
CV has both product related and non-product related attributes . For the sake of product differentiation, many firms provide products with a range of accompanying services. In addition to product-related services such as financing, delivery, customer service and recycling the product after usage, there are other possible value elements, such as providing information about the use and maintenance of the product.

CV has abstraction hierarchy
It is a common understanding in CV literature that customers evaluate a product at different levels of abstraction. For instance, the desired attributes at the lowest level and customer goals at the highest level . Or, tangible needs such as physiological and safety-related needs are at the bottom of Maslow's pyramid and intangible needs such as self-actualization are at the top .

CV is longitudinal
The value offering is not first assessed at the time of purchase, but (long) before the exchange takes place when the customer considers the purchase and after the purchase (e.g., Huber et al., 1997;. The entire process, from start (i.e., realizing a need and beginning the search) to finish (i.e., consumption and disposal of the product) can be viewed in episodes (Ravald & Grönroos, 1996). This is related to the well-known concept of "customer journey" in Marketing discourse.

CV involves trade-offs
CV includes perceived benefits as well as perceived sacrifices. Following the utility theory, many scholars (e.g., Anderson & Narus, 1998;Eggert & Ulaga, 2002; formulate CV as the perceived benefits minus the perceived sacrifice. The perceived sacrifices include monetary as well as non-monetary costs (e.g., ).

CV is aggregated
Although customers experience distinctive value elements at different points of time and circumstances, they assess a product holistically and form an overall judgment (e.g., Lemke, Clark, & Wilson, 2011;, which can lead to positive behavior (e.g., repurchase, passing on positive reviews) or negative behavior (e.g., switching to a competitor, posting negative reviews). In other words, CV accumulates over temporal space, as the customer moves from prepurchase to post-purchase assessment, and over cognitive space from attribute level to higherlevel abstraction.

CV is co-created
Over two decades ago, Normann and Ramirez (1993) noted that the distinction between physical products and intangible services were breaking down and the customer's role was changing from consuming to co-creating value with the producer. According to S-D logic (Vargo & Lusch, 2004), goods are appliances that provide services to customers and, in order for these services to be delivered, customers still need to learn to use, maintain, repair and adapt the appliances to their unique needs. Therefore, by using a product, customers create value (Grönroos & Voima, 2013). As such, suppliers and customers are co-creators of value (Gummesson, 2008b).

CV is relative
As  suggests, CV is relativistic, which suggests it involves the comparison of objects, people and situations. Perceived value can also be influenced by competitors' deliberate actions, such as "negative advertising" (Brandenburger & Stuart, 1996). Customer experiences are more social in nature, and peer customers are influencing experiences as well (Lemon & Verhoef, 2016, p. 69).

CV is dynamic
As technology, market, and customer need change, CV also changes over time. Companies cannot remain competitive if they address only today's customer needs and fail to anticipate their future needs. In fact, customers expect companies to anticipate their needs and desires, even though they themselves cannot (Flint, Blocker, & Boutin, 2011).
The research builds on the idea that capturing all nine key characteristics in a single instrument would provide a more holistic view of CV. The next logical step would be to identify the gap between the ideal model (that incorporates all nine characteristics) and the existing CV models. Such a gap analysis would indicate what additional features are needed in the new design. For that purpose, the next sub-section extracts and compares different CV models from the selected publications.

Review of influential CV models
The selected papers were screened for CV models and taxonomies. Table 1 compares the selected CV models from the perspective of the nine CV characteristics.
Two broadly known CV analysis methods in the Lean discourse, Quality Function Deployment (QFD) (see Marchwinski, Shook, & Schroeder, 2008) and Lean consumption (Womack & Jones, 2005b) are also added in Table 1 for comparison. The last row relates to the question as to whether the models are accompanied by a method of use (which makes it an instrument).
The various authors discuss many of these characteristics either explicitly, incorporating them in their CV model, or implicitly, as part of their CV discussion. If the selected publication conceptualizes a given characteristic by providing taxonomy or sub-constructs, the characteristic in question considered as covered and illustrated by a filled circle. Table 1 reveals some patterns and limitations. First, while CV Char. # 1, #2 and #3 are strongly emphasized in almost all the CV models in the Marketing stream, the remaining CV characteristics are emphasized less. Second, there is a strong emphasis on multiple attributes (CV Char. #2) and trade-off (CV Char. #5) aspects of CV in Lean tools, which can be linked to a more functional and operational view of CV. Finally, all the nine key CV characteristics are not incorporated in a single model that has a managerial use, which is a promising starting point for the next phase, i.e., the development of a comprehensive CV model. The CVM model proposed in this research is inspired by the matrix approach of the buyer utility map proposed by Kim and Mauborgne (2000). As it will be presented in the next section, this study expands the matrix in several dimensions to arrive at an analytical exhaustion.

Development of the CVM instrument
The CVM instrument, consisting of the CVM Model and the method of use, went through three design iterations where all the nine CV characteristics have been incorporated. The first version of the instrument (V1) had only six out of nine characteristics of CV. As V1 of the instrument was evaluated and literature was revisited, it became clear that improvements were needed to make the instrument more complete and at the same time more efficient. Based on the early evaluations, it was understood that experiential value has to be distinguished from the rest of the value categories, which agrees with some of the CV models in the literature (e.g., Smith & Colgate, 2007;Sweeney & Soutar, 2001). As a result of this CVM Model became a (7 x 6)-matrix. As is common under DSR, similar updates of the artifact under development (here the CVM Instrument) occurred as a result of further evaluations (see the next section for more details).  Ravald andGrönroos (1996) Holbrook (1996) Woodruff (1997) Kim and Mauborgne (2000) Woodall (2003) Sweeney and Soutar (2001) Smith and Colgate (2007

Constructing the CVM model
This study first converted generally abstractly defined CV characteristics into the design principles upon which the instrument could be built. The selected literature was revisited to validate the five design principles and revised them when needed where some of the design principles were further broken down to make them more precise. Four of the five design principles and corresponding CV characteristics fall into CVM modeling efforts and are presented below. The fifth design principle and corresponding CV characteristics are related to the method of use and will be discussed in the next sub-section.
Design Principle 1: The CVM Instrument should support the individualistic perspective and involve both tangible and intangible elements (incorporates CV Char. #1, #2, and #3).
CV encompasses a broad spectrum of tangible and intangible value elements. Building on existing literature, five value categories emerged as outlined below. For illustration purpose, some examples relating to a situation where a customer is in the process of purchasing a car are given.
Utility Value (I)-product/service attributes: It is related to the bundle of product/service attributes , which satisfies the practical needs of the customer. For example, the customer would be looking for ABS brake, GPS navigation system, after sales service, financing options, etc.
Utility Value (II)-operational attributes: It is about appropriate performances (Smith & Colgate, 2007) in meeting customer needs such as quality, speed, reliability, availability and flexibility. This category is essentially related to the producer's operational capabilities in delivering the bundle of product/service attributes.
Examples: Customers value on-time delivery, car quality performance and dealership performance.
Experiential (hedonic) value: Experiential value is related to the sensory and cognitive stimulation (Hirschman & Holbrook, 1982) facilitated by the provider, which allow the consumer to experience enjoyment, pleasure and fun during the stages of seeking, purchasing and consuming the product.
Examples: The customer interacting with the automaker's website, dealership atmosphere, and the fun of driving.
Relational value: This has to do with the benefits customers receive as a result of engaging in long-term relational exchanges with the firm. Ravald and Grönroos (1996) suggest that in a close relationship, the customer most likely evaluates the relationship as a whole, not necessarily just the offering. Gwinner, Gremler, and Bitner (1998) proposes three categories of customer perceived relational benefits: confidence benefits, social benefits and special treatment.
Examples: A trusted relationship with dealer, relationship with after sales service.
Symbolic (brand) value: Symbolic value, sometimes referred as brand value, relates to a customer's underlying needs for social approval, personal expression, shared value and outward-oriented self-esteem (Keller, 1993, p. 4). It essentially corresponds to the beliefs about the product/service and the firm. Clearly, the symbolic value satisfies the more psychological needs of customer and is highly intangible.
Examples: Some car brands profile themselves for 'safety' such as Volvo and some others for 'speed' such as Jaguar.
Design Principle 2: The CVM Instrument should focus on the customer's consumption processes, from beginning to end (incorporates CV Char. #4).
Customer process is considered as distinct temporal positions where customers assess the firm's offerings (e.g., Huber et al., 1997;Parasuraman, 1997;. Huber et al. (1997) define two positions-ex-ante and ex-post, whereas  explicate four, which are ex-ante, transaction, ex-post and disposition. Based on the concepts from the selected publications, this study delineates three temporal positions, which are similar to the stages outlined by Lemon and Verhoef (2016).
Pre-purchase: The customer's presumptions about value when considering a purchase, which may be driven by brand, relationship, earlier purchase experience and word of mouth.
Exchange: The customer's experience during transaction, which may be influenced by many factors, such as the convenience and safety of the transaction environment, friendliness of sales personnel, delivery performance (speed and reliability).
Post-purchase: The customer's experience with the product/service during consumption, including ease-of-use, effectiveness of product features, maintenance and ease-of-disposal.
In this study, for designing the CVM Model, each phase was divided into two stages. Pre-purchase was divided into search and select, exchange into transact and receive, and post-purchase into consume and dispose. As such, the six columns in the CVM Model represent the longitudinal perspective. In addition to developing a more fine-grained approach to capture the longitudinal aspect of CV, the main reason for expanding three stages to six stages is that, at each stage, the customer and the firm may meet in distinct forms (e.g., during search, visiting the website of the car manufacturer, and during select, visiting the dealer) and engage in distinct activities (e.g., during consume, carrying out periodic maintenance of the car, and during dispose, selling the used car).
Design Principle 3: The CVM Instrument should capture explicitly both benefits and sacrifices (incorporates CV Char. #5 and #6).
Perceived value is the difference between the perceived benefits (tangible and intangible benefits discussed above) and the perceived sacrifice, monetary and non-monetary (Bowman & Ambrosini, 2000;Eggert & Ulaga, 2002;Ravald & Grönroos, 1996;. Perceived sacrifice has two distinct components: Monetary sacrifice: The financial costs incurred by the customer from the start of the search and selection process up to the disposal of the product/service. Ravald and Grönroos (1996) divide total monetary cost into two broad categories: direct costs, which include purchase price, acquisition costs, order handling, transportation, installation and maintenance, etc., and indirect costs, which include delayed delivery, risk of failure, poor performance, repair, etc. From this perspective, 'value-adding activities' are related to direct costs, while 'waste' is related to indirect costs.
Non-monetary sacrifice: This category includes (undesirable) time and effort spent on acquiring and using the product/service (e.g., searching for the desired product in the shop, time spent in queues and experiencing difficulty using the product). In addition to time and effort, which is also mentioned in Lean consumption (Womack & Jones, 2005a), many scholars also consider psychological costs, which Ravald and Grönroos (1996, p. 26) describe as "worry about whether a supplier will fulfill his promises".
The three design principles provide the foundation of the CVM Model, which is depicted in Figure 1. The sub-categories given for each design principle essentially form the rows and the columns of the matrix. The rows represent five benefit and two sacrifice categories. The columns represent the longitudinal journey from search to disposal. Each field of the matrix refers to an option of creating value by improving customer benefits and/or reducing customer sacrifice.
Consisting of two distinct dimensions (time and perceived value), the CVM Model illustrates that CV indeed stretches along two orthogonal directions: temporal and cognitive space. At the center are the core benefits (+) such as physical product or service and the core sacrifices (-) such as price. Moving away from the center means moving away from the core product, indicated by arrows and expanding ovals, and reaching the surrounding value elements that the customer experiences at different times and at different abstraction levels. It can also be logically inferred that the (7 x 6)matrix provides a holistic view of what the customer values throughout the consumption episode; hence, it incorporates CV Char. #6 (CV is aggregated).
So far, the CVM Model as illustrated in Figure 1 does not distinguish whether the customer is evaluating the purchased (or to be purchased) product/service or the other encounters with the firm (e.g., joint solution development, customer's call for information). Vandenbosch and Dawar (2002) argue that customers value how they interact with their suppliers as much as or more than what they actually purchase. If interactions (also referred to as encounters) were not explicated in the design, then the source of value would remain limited with the product/service, which takes us back to the G-D logic. This observation leads to an additional design principle.
Design Principle 4: The CVM Instrument should explicate all the encounters beyond product usage (incorporates CV Char. #7).
As mentioned, the columns in the CVM Model as depicted in Figure 1 relate to a longitudinal journey where customer and firm interact, the so-called encounters. Building on the co-creation conceptualizations of Payne, Storbacka, and Frow (2008) and Ballantyne and Varey (2006), this study proposes four broad forms of encounters, each of which is treated as an additional "layer" on the CVM Model foundation: Provider-led encounters: This form of encounter is initiated by the provider and related to the provider's proactive orientation. Essentially, the provider pushes information and resources to influence the customer's value creation process (e.g., advertisements, website, user manuals, installation instructions, customer visits).
Customer-led encounters: This form of encounter is initiated by the customer and related to the provider's reactive orientation. The customer pulls information and resources from the provider to manage his value creation process (e.g., service request, order status check). Dialogical encounters: This type of encounter matches the notion of relationship development and knowledge generation, which can provide opportunities for learning together and developing new solutions (Ballantyne & Varey, 2006).
Product/service encounters: This is related to consumption of the product/service. Payne et al. (2008) calls this type of encounter 'usage encounters' and relate it to customer practices in using a product/service and the services supporting such usage.
Combining the (7 x 6)-matrix in Figure 1 and the four forms of encounters, the CVM Model takes its final form. To demonstrate the complete CVM Model, the car repair example discussed in Womack and Jones (2005a) is used and expanded to cover the other areas that this research proposes; the table in Appendix B illustrates the benefits and sacrifices that the customer can experience through four possible encounters.

Developing the method of use of the CVM model
While using the CVM model can help organizations unlock many value potentials, the method of use, which is built on key ideas of Lean thinking, can guide them how to prioritize and act on them effectively and efficiently. CVM Model as discussed in the previous sub-section cover CV Char. #1 through #7, clearly, also the remaining two characteristics (CV Char. #8 and CV Char. #9) are also fundamental to be able to offer a holistic approach to value creation. Therefore, a fifth (and final) design principle is needed.
Design Principle 5: The CVM Instrument should support the relativistic view, i.e., the value is relative to other offerings and time (incorporates CV Char. #8 and #9).
Consisting of seven steps, the method of use (see Table 2) has been developed iteratively by incorporating the feedback after each evaluation stage.
Note that step 5 emphasizes the importance of differentiation, hence incorporates CV Char. #8 (CV is relative). To develop CV learning capability in the organization, this exercise, step 1 through 7, should be set-up and viewed as one of the organization's strategic processes, i.e., following the PDCA principal of Lean, the CVM Instrument incorporates CV Char. #9 (CV is dynamic). To summarize, the CVM Instrument proposed consists of various building blocks, namely, the CVM Model sub-structure as depicted in Figure 1, the four types of encounters (which together build the full CVM Model) and the method of use highlighted in Table 2. It has been demonstrated that indeed the CVM Instrument does adhere to all five design principles derived from Marketing literature (therefore covers all nine CV characteristics) and some key ideas and techniques in Lean.

Evaluating the CVM instrument
This section presents how the CVM Instrument is evaluated. Note that the scope of the evaluation is the CVM Model and the first five steps of the method of use. This is because step 6 (QFD tool) and step 7 (PDCA approach) as explained in Table 2 are well known and proven methodologies in Lean toolkit, hence out of scope of this evaluation.
Following Gregor and Hevner (2013), this study uses validity, utility, quality, and efficacy as a complete set of criteria to assess the CVM. Gregor and Hevner (2013) describe validity as "the artifact works and does what it is meant to do; that it is dependable in operational terms in achieving its goals" and utility as "whether the achievement of goals has value outside the development environment". Although the other two criteria, quality and efficacy, are not explicitly defined by the authors, this study derives the following working definitions based on DSR literature. Quality implies how accurately and completely an artifact achieves its goals, while efficacy refers to how easy and efficient it is to use the instrument.
In line with Gregor and Hevner (2013), who argue that a rigorous design evaluation may draw on multiple techniques, including expert reviews, workshops, and semi-structured interviews, this research follows a "two step analysis" approach of the case-based research (Eisenhardt, 1989;Yin, 2013). The first step is "analyzing evaluation stages individually" (Section 5.1). Here the collected data, including discussions (voice-recorded), survey input and observations during each formal evaluation stage are transcribed. The transcriptions are reviewed by the lead author and one of the co-authors, and finally by the company contacts for in-company workshops. Those transcripts are later coded for data reduction purpose. The second step is "looking across evaluation stages" (Section 5.2) to demonstrate the evolution of the design with corresponding evaluation criteria.

Analyzing evaluation stages individually
In evaluating the various versions of the artifact under construction (i.e., the CVM Instrument), two phases are distinguished, see Figure 2 that illustrates how various versions (indicated as V) have evolved through various evaluation stages (indicated as E). Within the first phase, the evaluation was focused on face validity, which refers to whether or not the model appears to perform what it claims to perform (Gravetter & Forzano, 2015), which is also known as proof-of-concept (Gregor & Table 2. The method of use-seven-step guideline

Steps
Description and incorporating some key ideas from Lean 1. Map the customer's key actions from search to dispose Womack and Jones (2005a) suggest that the best way to identify opportunities for improvement in the customer's processes is to draw a map showing the consumption steps. Also, in using the CVM, the starting point is to capture the current status of the customer's process, from search to dispose.

Identify the customer's tasks in each key action and categorize them in the four forms of encounters
Identifying various sources of value, including information, products and interactions, Smith and Colgate (2007) suggest that firms can develop strategies to enhance each source of value at a functional, hedonic and symbolic level. In other words, each customer action should be projected from the four forms of encounters (sources of value) onto the value categories.

Identify perceived sacrifices and highlight the main issues (pain points)
The activities that do not add value should be identified, as described in the value stream mapping technique in Lean management (Rother & Shook, 1998). In addition to that, both monetary and non-monetary sacrifices should be identified at each stage and entered in the corresponding fields in the matrix.

Assess strategically cost and differentiation advantage in reducing customer's sacrifices
Before allocating limited resources to improve the customer's process, organizations should first check the prospects for realizing cost and/or differentiation advantage relative to the competition if the selected pain point is resolved. This requires close monitoring of market trends and competitors' actions, which is generally known as 'market orientation' (e.g., Jaworski & Kohli, 1993;Slater & Narver, 1994).

Inject value into five value categories (perceived benefits) to reduce perceived sacrifice and differentiate
As discussed earlier, firms need to focus both inward, at processes designed to improve efficiency (exploitation) and outward, towards market (exploration), to identify new opportunities. It is essential to combine the findings from step 3 and new insights from the market to decide in which areas to inject the right amount of value in the perceived benefits area. That way, the firm can both solve the customer's problems and achieve differentiation.

Prioritize value elements that create synergy-reduce sacrifices, increase benefits and increase differentiation
After identifying value elements in the CVM, managers need to strategically refine and prioritize those value elements based on their importance to the customer, competition and the firm's capabilities, and bring them to a more manageable level. QFD tool can be very effective for such purpose and complement CVM.

Experiment-test value elements by following PDCA
Drawing lessons from Toyota's success, Spear (2004) states that proposed changes should always be structured as experiments, during which hypotheses are tested and the results are used to refine or reject the hypotheses. In this respect, the CVM should be seen as an instrument designed to formulate and test the hypothesis. Steps 1-6 can be seen as the plan part of a larger PDCA cycle.
Step 7 has to do with testing the hypothesis (do), controlling results by analyzing customer feedbacks (check) and standardizing and expanding what works and learning from what does not work (act). Hevner, 2013). Following the proof-of-concept, a more formal evaluation is planned by incorporating all four evaluation criteria in the second phase. At the end of each stage, the model has been adjusted to the new insights drawn from the collected data.

Phase 2: formal evaluation
To determine the validity, utility, quality, and efficacy of the instrument more rigorously, a more indepth evaluation phase took place. In this phase, two types of workshops were conducted, the first of which took place within individual companies and brought together managers from various disciplines, including supply chain management, finance, purchasing, new service development and sales (the aim was 6 to 10 participants). The second type of workshop was conducted with business representatives of organizations from various sectors, in two events. Different workshop settings provided us both the breadth (variety of sectors) and the depth (company specific) of feedback to assess the instrument.
In this evaluation round, validity is measured by asking whether the CVM leads to a better understanding of CV. Quality is related to completeness of the model, that is to say, whether or not the model is able to capture all essential aspects of CV. Utility has two aspects: interest in using the tool within the organization and fitness of the tool to the organization. User input about the efficacy of the model-how easy and efficient it is to use the model-was solicited through an open question and through group discussions.
Data are analyzed based on the coding technique proposed by Miles and Huberman (1994). First, a detailed write-up of each evaluation stage, which involves survey analysis, transcription of voice recording and observations, was carried out immediately after each workshop. For in-company workshops ([E6], [E8]), a written draft was written by the lead author, reviewed by one of the co-authors and by the key informant in the company, which increased construct validity (Yin, 2013). Second, the transcript was coded manually for data reduction, first with open coding and then with axial coding (Miles & Huberman, 1994). The codes and their descriptions are provided in Appendix C. Finally, reduced data is displayed in a single overview (see Table 3) to facilitate searching for a pattern (Miles & Huberman, 1994, p. 239).
Further details on the set-up of the workshops at the various evaluation stages [E5]-[E8] and the finding are discussed below.

[E5] event workshop (supply chain management congress)
The workshop took place at a Supply Chain Management conference in the Netherlands. The participants − a total of 19, excluding two facilitators − were a mix of experienced professionals from various organizations and backgrounds, including SC Managers, Project Managers, CEO, CFO, Business Development Managers, Controllers and Product Managers. After introducing the basic CVM Model, the participants were asked to fill out one specific area of the matrix in pairs. They were instructed how to come up with ideas to improve CV in this specific area. After the exercise, each group was asked to give a short presentation of their best ideas. After the wrap-up, postworkshop questionnaires (translated in Dutch) were handed out to collect individual feedback.
The CVM appeared to have strong face validity and utility, which had already been concluded in the proof-of-concept phase, and was now re-confirmed. The model was considered valuable for enabling customer orientation in firms.

[E6] in company workshop (wholesaler_comp)
Wholesaler_Comp is a wholesaler involved in the installation, construction and industrial articles (lighting, plumbing, sanitary, central heating, appliances, roofing, etc.) in the Netherlands, with about 2,000 employees. The seven people who took part in the workshop represented supply chain management, project management, finance, business development, marketing & e-commerce, and purchasing.
The concept of intangibles (brand value and relational value) appeared too abstract for some participants. Although these participants found it easier to think along the customer journey with utility value and experiential value, it was less clear how brand value and relational value related to the timeline. The CVM was found to be a complete and structured approach to determining CV, but some participants indicated that it was time-consuming and ambiguous. Accordingly, efficacy, compared to other criteria, was rated the lowest.
Based on the observations and feedback, the following actions were planned for the next design iteration: providing more clarity for the value categories, making users focus more on the customer's processes and sacrifices, and simplifying the process of filling in the CVM. A second version of the method of use (V2) was developed, which is used in [E7].

[E7] event workshop (lean summit)
The workshop was carried out in a breakout session during a Lean Summit that took place in the Netherlands. The 20 participants (not including two facilitators) were professionals with a varying level of experience with Lean and coming from different sectors such as management consulting, public services and railway services. First 20 minutes, the CVM Instrument was introduced and the remaining 40 minutes spent on group exercise and discussion. For the exercise, four groups were formed. Each group was asked to choose a specific product/service line from one of the companies that the group members represented.
According to the participant's comments and survey feedback, the CVM Instrument appeared to have medium validity and quality. As indicated in the earlier evaluation, many participants found the model too complicated to use. Therefore, the instrument was scored medium-low on utility

Medium-Low
The instrument is easy to use and efficient way to develop a broader understanding of CV and efficacy. Three important learning points emerged from this evaluation. First, demonstrating the model in such a short time (1 hour) frame appears to be challenging. Many users from public services struggled to identify their customers. Scoping seemed to be a crucial element of the exercise, which requires preparation before the workshop. Second, starting from "perceived sacrifice" encouraged users to focus more on the customer's processes and issues, which was the key objective. As a conclusion, further improvement, especially in utility and efficacy, were needed for the next design iteration.

[E8] in company workshop (Logistics_Comp)
Logistics_Comp is a large logistics service provider that provides solutions in freight management and contract logistics worldwide, in over 1,000 locations in 160 countries. Headquartered in the Netherlands, Logistics_Comp employs about 40,000 employees and serves to a wide range of sectors. The company adopted Lean management in 2008 and considers Lean as the key element of their business strategy. For this workshop, Logistics_Comp selected one of the services for a major telecommunication client. This was because at that moment Logistics_Comp was looking for new ways to improve their services for this particular client. Selected service was "delivering replacement hardware equipment (internet + TV) to the existing residential customer when the hardware is defect".
The discussions and survey results indicate that the CVM Instrument has strong validity and utility. In other words, using the CVM Instrument can lead to a better understanding of CV and make organizations more customer-oriented. Most participants agreed that the main strength of the CVM Instrument is its structured and complete approach to capturing CV. Although some steps were considered to be lengthy, the overall logic was relatively easy to understand. Due to multiple steps and many fields to fill in the CVM Model, the efficacy of the CVM instrument was scored medium-low. The main area of future improvement appeared to be "scoping the customers", which will be elaborated in the discussion sections.

Looking across evaluation stages
By reviewing the results as summarized in Table 3, the following observations can be made. Of the formal evaluation stages, [E5] received the highest rating. This may be due to the fact that the CVM Instrument was not intended to be fully completed, but to be tried out for the purpose of demonstration. However, in company workshops, [E6] and [E8], the users were able to examine each value category and experience the tool thoroughly. Because the value categories were not appropriately (user-friendly) defined until [E7], the users in [E6] experienced difficulty completing the exercise. In the formal evaluation phase, [E6] was the first in-depth and the lowest rated evaluation stage. Based on the rich feedback received from [E6], the method of use was improved. In [E7], it was observed that validity, quality and efficacy improved, while the utility rating fell, due to several negative comments from the users coming from public services. In [E8], the model version V3 of the CVM was introduced, which has both (7 x 6)-matrix and four forms of encounters, i.e., was theoretically more complete and more complex than earlier versions. However, when comparing the two in-company workshops, [E6] and [E8], it can be observed that there are noticeable improvements in all four criteria.
Further analysis revealed four notable patterns. First, the VAL-SCOP(-) rating, which is about the process and customer scoping, worsens, because, in [E7] and [E8], determining who the customer was posed a challenge and led to lengthy discussions. Second, the overall quality of the instrument increased noticeably; while QUAL-STRUC (+) and QUAL-VIS (+), which are related to structural and visual aspects of the instrument, increased, QUAL-VAG(-), which is related to the vagueness of the value category definitions, decreased. Third, although the final version of the method of use (V3) improved the ease of use, the overall exercise remained "lengthy" in the eyes of the participants. In other words, EFF-TIME(-) did not improve. Therefore, it can be concluded that from the perspective of efficacy, the instrument has not matured yet. Fourth, out of a total of 50 workshop participants in [E5]-[E8], an overwhelming majority of the participants (about 80%) does not have a well-structured approach to determining CV.
The overall conclusion from all evaluations can be that, although there can be deviations in the performance of the instrument in DSR as Vaishnavi and Kuechler (2008, p. 21) point out, the results at the end of [E8] can be considered as "satisfactory".

Discussion
The first point of discussion is the theoretical generalizability of the findings (Holmström et al., 2009) and grounded technological rules (Van Aken, 2004) to contribute to the CV knowledgebase. Formal evaluation phase revealed four challenges in the use of the CVM some of which are addressed in the next iteration and improved.
Firstly, the common feedback from many users in [E5]- [E8] was that the CVM Instrument is complete and structured but at the same time ambiguous and time-consuming. Clearly, the aim of being complete by incorporating all nine CV characteristics in the CVM Instrument has reduced the efficacy of the model. Although a trade-off between completeness and efficiency was to be expected, this should be managed in order to make the tool useful for organizations. In fact, adding the four layers of encounters during [E8] increased completeness and improved efficiency at the same time as well-defined encounters helped the users to sort their thoughts better. Based on these findings, the following technological rule is derived: Rule 1. Striving to capture the holistic notion of CV with a fine-grained instrument such as CVM Instrument can reduce efficiency of the use of instrument, hence the organizational interest in it.
Secondly, at [E8] it was observed that in fact, Logistics_Comp had to satisfy two types of customers: the residential customers (consumers receiving telecom service and that Logistics_Comp delivered hardware to) and non-residential customer (the Telecom provider as the direct customer). Because the CVM Instrument is designed from a single customer perspective, it was not immediately clear how to capture the value of the end user. Similarly, during [E7], some users from the public sector voiced their concerns about how to scope their customers and deal with multiple stakeholders. One Lean expert in [E7] suggested that using the Lean SIPOC (Supplier, Input, Process, Output, Customer) tool might provide a relational overview of the key stakeholders. Gummesson (2008a) state that value is not created just by the supplier and the customer but in a network of activities involving multiple stakeholders. Therefore, the following rule can be derived: Rule 2. Before determining CV in a multiple stakeholder context, the scope has to be first, broadened from dyad (seller-buyer) to network and then, the network of interactions has to be mapped. Only after this exercise, organizations should decide which customer category to focus.
Thirdly, during [E6] it was observed that without guidance the users tend to focus on the top part of the CVM (adding more benefits) while the bottom part (perceived sacrifices) was largely ignored. At [E6] many of the ideas that were proposed on enhancing CV were about how to add to the current perceived customer benefits (i.e., on how to do the current tasks better). However, as stressed in Ravald and Grönroos (1996), "adding more value" or introducing "extras" may not be driven by genuine customer needs can only offer short-term solutions. In [E8] the users were instructed to focus on first to customers' perceived sacrifices. This simple change made a striking difference in terms of generating more relevant ideas. The users were more engaged in brainstorm sessions about understanding the process of customers (customer journey) and customers' gains and pains. This finding leads to the following rule: Rule 3. In order to generate rich and relevant CV ideas, it is necessary to first understand the customer's perceived sacrifices, which can promote customer focus. Conversely focusing first on perceived customer benefits, organizations might fall into trap of 'adding more of the same', which can reinforce an internal process focus.
Another point of discussion is on why the CVM Instrument is positioned as being a contribution to the Lean toolbox. Or, more broadly: what is Lean about the CVM Instrument? First, at the outset of the paper, it was argued that the current CV tools are limited and are basically focused on the internal processes. Although the field of Operations Management is frequently considered as the "natural habitat" of Lean, arguably the essence of Lean management lies beyond internal processes; after all the objective of Lean is to remove waste and improve customer value in processes that span the entire organization (or rather the entire end-to-end supply chain). From this perspective, taking a multi-functional approach, i.e., integrating ideas from the Strategy and Marketing Management into the CVM Instrument, can be considered as a Lean way of working.
Second, several components of the CVM Instrument fit well within the Lean approach. Some examples are; (a) using a disciplined and standardized way of getting a broader and deeper understanding of customer value by considering all fields in the CVM; (b) incorporating the existing Lean QFD tool to prioritize customer value elements in Step 6 of the method of use; and (c) using PDCA on a continuous basis to develop, test and refine CV hypothesis in Step 7 of the method of use.
Third, as has been explained in Rule 3, using the CVM instrument can benefit largely by first focusing on "perceived customer sacrifices". This can be associated with the typical Lean practice to look for "waste". That is, rather than only looking for internal waste, the CVM Instrument suggests that in order to create customer value, a good starting point would be to remove waste in integrated firm-customer processes, which supports Lean consumption view.
Fourth, the practitioners involved in the various experiments had no difficulty in seeing the use of the CVM Instrument from a Lean perspective. For example, at [E8] a user stated that CVM is an interesting addition to their Lean way of working. Another user stressed that CVM is a hands-on tool and that the tool can be adopted straight away in many Lean organizations.

Conclusions
In today's highly competitive and fast moving business world, organizations increasingly need to look deeper into the needs of their customers. However, it can be observed that many organizations lack a sophisticated technique to capture CV in a holistic way. Therefore, in this paper, a new tool dubbed CVM Instrument is developed and tested.
Theoretically speaking, this study aims to add to the observed limited conceptual understanding of CV in the Lean literature by bridging between Operations, Strategy and Marketing disciplines. Based on design theory approach, this research integrates various CV concepts from Strategy and Marketing management research to develop five design principles needed to build the CVM Instrument. Compared to previous studies, which typically focus on only some characteristics of CV this research incorporates nine CV characteristics in a comprehensive design. The proposed CVM Instrument includes key elements from S-D logic (Vargo & Lusch, 2004) and Lean consumption (Womack & Jones, 2005a), thereby responds to the research calls from Gummesson (2008b) and Lusch (2011) on having a more holistic research agenda in the Lean discourse. More specifically,  proposes that a new research frontier in supply chain management is integrating the customer into the supply chain-moving from viewing customer as the destination of supply to someone to co-create value with. This research does exactly this, hence contributes to the discussions about the boundaries of the Operations Management field, which is frequently considered as the "natural habitat" of Lean. Furthermore, build-evaluate iterations have provided new insights about how a CV capturing exercise can support a more holistic understanding of CV as well as the challenges that need to be overcome. In addition, this study contributes to the limited but growing DSR-based research in the Operations Management community (Van Aken, Chandrasekaran, & Halman, 2016). Finally, the CVM instrument should be seen as a "boundary object" (Spee & Jarzabkowski, 2009), which, as an artifact, enables knowledge sharing across semantic boundaries of organizational departments (e.g., between Production and Marketing departments). As Talmar, Walrave, Podoynitsyna, Holmström, and Romme (2018) argue, boundary objects are needed to carry scientific knowledge to management practice and bring practical wisdom into science thereby expand our knowledge.
More from a practical viewpoint, this study demonstrates that the CVM Instrument enables firms to deepen their understanding of CV. It enables managers to view CV in a more structured, explicit, balanced and complete, i.e., holistic fashion. By thinking through each cell of the (7 x 6)-matrix from multiple encounter perspectives, organizations can discover value possibilities that would otherwise remain unexplored. For instance, car repair service example in Appendix B illustrates many value potentials in a structured and complete way. Such a holistic view of CV can help the service providers tap into different value areas that are typically ignored such as explaining the repair to the customer with a friendly video. In addition, the CVM encourages managers from different areas to increase their customer focus and align on a common end-to-end product/ service vision. In that sense, CVM as a management instrument might be the right recipe to develop a more balanced customer and process focus for organizations adopting Lean, which can create a cycle of improvement as pointed by Radnor and Johnston (2013). Finally, the seven-step approach in the method of use can facilitate a systematic, continuous CV learning process in organizations.
There are various areas in which the CVM Instrument requires further work. First, the "intangible dimensions" of the CVM Model should be made more practical to think through at each stage such as how organizations can increase brand value during the transaction stage. Second, more workshops and various setting with different sectors can be organized to improve the reliability of data and theoretical generalizability. Finally, to be able to thoroughly evaluate the CVM Instrument, a longitudinal study is needed, making it possible to not only fully evaluate the instrument, but also to validate whether indeed the usage of the CVM Instrument leads to a better customer orientation. Each of these limitations points to potential future research through which CV understanding and conceptualization can be further advanced, within Lean discourse as well as in CV literature.