**1. Introduction**

In the current scenario of the house building industry, there is a fierce market competition in different countries, primarily concerned with costs, demanding strategies to increase productivity [1,2] and, at the same time, to consider customers heterogeneous demands [3]. Understanding customers' needs and preferences is a challenge due to their changing lifestyles and different family structures [4–6]. Therefore, customer requirements must be appropriately understood and communicated to decisionmakers, such as investors, developers and designers; otherwise, value generation may be compromised [4]. The progressively increasing diversity of customer requirements has created business opportunities related to product customisation in several different sectors [7,8], including house building [9]. According to Wang et al. [10] this shifting focus from company to customer demand is a driving force in industrial innovation.

Mass customisation (MC) is a strategy that aims to fulfil customer requirements [11–13], and, at the same time, achieve high efficiency and competitive advantage [2,11], through flexible processes and supply chain integration [1,14]. Therefore, companies combine elements of mass and craft production to improve value generation for specific market segments [15–17]. In the house building industry, besides contributing to competitive advantage, the adoption of MC can provide benefits related to environmental and social sustainability, by avoiding waste caused by product changes made after occupancy by users, as well as by increasing their perceived value and sense of ownership [5,18].

Several successful applications of MC in the manufacturing industry have been reported in the literature [7,19,20]. However, its body of knowledge is dispersed and is still growing [7]. According to Piller [21] and Suzic et al. [20], there is a lack of in-depth understanding of the strategies for implementation. Other authors [22,23] argue that the further expansion of the field depends on the development of models and tools to support companies in new product development (NPD). A major challenge in MC is customer integration, i.e., how to improve value generation by understanding and considering requirements from different customers, as well as defining their degree of involvement in NPD [22,24]. Most studies on this topic tend to be technology-focused [19], being often limited to methods and digital tools to generate and display product alternatives, such as configurators and choice menus [22].

In the house building industry, the implementation of MC is still latent [1,25], sparse and more focused on operations [1]. A critical challenge for the adoption of MC in housing is capturing customers' requirements [3,14,25–27], and establishing a balance between offering variety and achieving efficiency and, consequently, housing affordability [1,9,25,27]. Several research opportunities on this topic have been pointed out in the literature, such as the definition of solution spaces, and the support to customers' decision-making during the configuration process [1–3,22,25]. However, Khalili-Araghi and Kolarevic [3] sugges<sup>t</sup> that new methods for customer integration are needed to reduce the trade-offs between customers perceived value and the complexity that results from customisation. Kotha [17] argues that technologies and tools alone are insufficient to achieve MC goals, as the adoption of this strategy requires an organisational context that fosters continuous improvement, learning and knowledge creation.

Some studies have associated the use of MC strategies with prefabricated or industrialised construction methods (e.g., [1,28,29]). However, this strategy has also been explored by companies that adopt traditional construction methods (e.g., [6,9,25,26]). In fact, some of the potential improvements related to MC are not directly related to the type of technology used, such as understanding customer requirements, customer interaction, and visualisation approaches [2,6,25,30]. Rocha [30] suggests that the definition of an MC strategy can be divided into decision categories, and should start by making some core decisions related to the scope of MC, and then move to other areas, including customer integration. Wikner [31] defines decision categories as ways to classify decisions and support the segmentation of complex decision problems into a structured and relatively independent way to facilitate decision-making.

A possible starting point to understand key decision categories is to analyse practices implemented in the industry [20,32,33]. Those practices can be regarded as methods, tools or techniques that have been successfully used in real-life situations for improving performance or solving problems [32]. By understanding the underlying ideas of those practices, they can be adapted to other companies facing similar challenges [33]. This research seeks to further understand practices as an expression of tacit knowledge that can be applied for learning, working, innovating and organising [34].

Therefore, this research study aims to answer the question: How can customer integration in the NPD of mass-customised house building projects be managed? The main outcome of this investigation is a framework of decision categories for customer integration and for devising the scope of customisation to support the definition of MC strategies. It is based on practices identified in the literature and also on an empirical study carried out in a house building company. The framework is meant to be used by companies to support the definition of MC strategies. A secondary contribution of this investigation is a set of constructs that have been used to describe the decision categories and their relationships.

This paper is structured into six sections, including the introduction. In the theoretical background section, MC is discussed, emphasising its core concepts, especially the ones related to customer integration. In the third section, the research method is presented, including the methodological approach and research design. Then, the results of the empirical study are presented in the fourth section. In section five, the framework for customer integration is presented and evaluated. Finally, in section six, the main conclusions and opportunities for future research are presented.

#### **2. Theoretical Background**

#### *2.1. Mass Customisation and Related Concepts*

According to Silveira et al. [35], the success of MC strategies relies on several internal and external factors, such as customers demand for customisation, market and value chain readiness, technology availability, and knowledge sharing. Other studies [22,28,30] point out that the implementation of MC depends on the coordinated e fforts from three di fferent areas of the company: customer integration, product design and operations management. After requirements are captured, the design area must focus on developing product alternatives by translating those requirements into specifications. Finally, operations managemen<sup>t</sup> is concerned with producing and delivering customised goods, by managing resources and the supply chain to achieve time and cost-e ffectiveness [22,30].

MC depends strongly on the company's ability to translate customers' demands into new products and services, in which knowledge creation and information sharing play a key role [22,35]. According to Kotha [17,36], knowledge creation in the MC strategy has two primary sources of information: (i) external, from customers, and (ii) internal, related to internal processes and workers' experiences.

Customers inputs into NPD can be communicated in di fferent ways, such as desires and needs, suggestions towards product solutions, and even insights that may lead to radical innovations [37]. According to Piller et al. [24], by translating customer preferences and needs into product requirements, companies are able to transform subjective information into explicit knowledge. This knowledge can be used to understand customer demands and inspire new developments [17,24,36]. Besides, feedback from customers and previous choices can be used by companies to introduce innovations and also provide guidance on whether to limit or expand product variety [17,36]. Furthermore, Wang et al. [10] discuss emerging methods for collection and storage of customers inputs based on "Big Data" and other IT tools to support decision-making. Therefore, different practices can be used to capture such knowledge [37].

The level of customisation is concerned with the range of customisation options to be offered in order to satisfy different customers [13]. However, this decision needs to be based on the analysis of trade-offs between the company's capabilities and customers' demands [7,35,38]. Moreover, customisation can occur at various points in the value chain, from a minor product adaptation to full customisation defined at the design stage [35,39]. Each one of these points may be related to a specific level of customisation, and requires the definitions of how and when customers' needs are translated into product specifications. A number of taxonomies of customisation types have been proposed in the literature based on the level of customisation, such as the MC generic levels proposed by Silveira et al. [35]: design, fabrication, assembly, additional custom work and services, package and distribution, usage and standardisation. Another example is Barlow et al.'s [40] set of strategies for the house building industry (Table 1), which is based on Lampel and Mintzberg [15].


**Table 1.** House building strategies.

Source: adapted from Barlow et al. [40].

The location of the customer order decoupling point (CODP) is essential to define the customisation level [41,42]. It divides the value chain in processes based on forecasts (mostly standardised) and on customer demands (customised according to orders) [24,38,39] (see Figure 1). The CODP also defines which activities are postponed until the customer's specific requirements are captured, and an order is placed [24].

**Figure 1.** Customer order decoupling point (CODP) in housebuilding. Source: adapted from Barlow et al. [40] and Silveira et al. [35].

Therefore, the extent of customer integration is closely related to the level of customisation [24] and the CODP definition [39]. In fact, the level of customisation usually defines the intensity of customer–company interaction during NPD [24,38]. Moreover, a high customisation level should rely on collaboration with customers from early design stages, while a low one requires less intense participation of customers [38].

When defining the level of customisation, companies should bear in mind that offering too many options not only can make operations inefficient, but also cause customers frustration and confusion, the so-called burden of choice [43,44]. Thus, the definition of a limited solution space plays a key role in MC. The solution space consists of a combination of different customisation units (i.e., customisable attributes and their available options) and rules to combine them, limiting the set of possible product alternatives [30,44]. However, even if there is a limited number of flexible processes, a large number of features and product alternatives may be generated [7,19,21].

Previous studies [14,23,28,30] have pointed out that devising a solution space must be based on the identification of customers' needs and preferences for product customisation, and decide whether and how those will be meet [2,28,44]. It must also be highlighted the importance of post-occupancy evaluations (POE) to capture requirements and provide feedback for the NPD of future house building projects [14,26].

Rocha [30] proposed three core decision categories to define the scope of an MC strategy in house building: (i) the solution space; (ii) customisation units; and (iii) classes of items, which are specific properties of options offered in the customisation units [30]. Additionally, Amorim [45] proposed a decision category named communication of customisation information that defines how the information is made available, when and for whom. This is strongly supported by previous studies [9,25–27,45–47], that highlight the need to improve the effectiveness of information flows between different sectors of the company, in order to facilitate collaboration and improve value generation.

Rocha [30] suggests that the level of customisation should be considered as an operations managemen<sup>t</sup> related decision category, as it is related to the definition of when and how customisation units are defined. However, Schoenwitz et al. [28] sugges<sup>t</sup> that customers' preferences play a key role in the definition of the customisation level, indicating that there is an interaction between customer integration and operation managemen<sup>t</sup> decisions. The same authors also pointed out that the definition of a single CODP neglects the possibility of choices to be made separately for different components and attributes, which are made feasible by prescribing multiple decoupling points.

## *2.2. Customer Integration*

According to Franke, Keinz and Schreier [48], the value delivered by mass customised products is driven by the fit, style and functionality, or utility perceived by customers, and the uniqueness of a product. Customers are often willing to pay extra to obtain customised goods [1,21,38]. Furthermore, Piller [21] argues that the willingness-to-pay (WTP) reflects the value perceived in the increment of utility that they gain from a product that better fits their needs rather than the best standard product available. Therefore, customer integration should start from capturing needs and preferences, and estimating the WTP for a customised good [22].

Kumar et al. [7] argue that customer integration embraces not only co-design but also other types of interactions between companies and customers, which can be enabled by modular design, configurators, and elicitation of needs. It means that customers can have an active role in product definition, configuration or modification within a given solution space [19,21]. Thus, premium prices are charged to cover additional costs resulting from customisation, such as higher costs of sales [17] and operations [24]. Moreover, customer integration can also bring some cost-saving results from collecting consistent market information and establishing a close customer–company relationship [24].

In this context, new relationships must be established between customers and companies [3]. Thus, companies can benefit by expanding the use of traditional customer relationship managemen<sup>t</sup> (CRM) tools [49] to relational marketing ones [50]. These are means to build long-lasting relationships with customers, by improving value generation through interactions, creating trust and increasing loyalty [49–51]. According to Tommaso [50], relational marketing is based on a logic of exchange and learning. It can potentially improve customer experience, which refers to the combination of a number of personal impressions (considering cognitive, affective, behavioural, physical and social aspects of the response), resulting from interactions between a customer and a product or service [50].

According to Silveira et al. [35], the customer–company interface must be tailored to each unique context. Fetterman et al. [25] proposed a set of steps to outline a customer–company interface for the house building industry, which is built on a proposition by Silveira et al. [35]: (i) defining a solution space to be offered to customers; (ii) collecting and storing information on customers choices; (iii) transferring data from retail to production; (iv) translating customers choice into product design features and manufacturing instructions; and (v) delivering customised products and offering post-occupation customisation. In step two, effective ways to present the solution space for customers are needed [30,35], enabling them to deal with the variety of alternatives, avoiding the burden of choice [43].

Rocha [30] suggested two decision categories for customer integration, namely, configuration sequence and visualisation approaches. These are concerned with how the customisation units are presented to customers and how they engage in creating the product. The first one involves defining a sequence of decisions to be made by customers when configuring their product alternatives [30]. The visualisation approaches decision category defines how the customisation units will be displayed and to whom (i.e., customer, company or both), being divided into three types: collaborative, transparent and do-it-yourself [30], similar to the approaches proposed by Gilmore and Pine [16]. For example, in the collaborative approach, both customers and companies are aware of the customisation process and can be applied through choice menus and or a dialogue between the company and customers [30]. However, Rocha [30] only proposed a broad definition of those three approaches, without discussing how to implement or combine them for effectively presenting the solution space to customers.

#### **3. Research Method**

Design science research (DSR) was the methodological approach adopted in this investigation. This approach typically involves the development of innovative solution concepts, named artefacts, to solve classes of practical problems, and at the same time contribute to the development of mid-range theories, i.e., theoretical models that apply to a limited range of situations [52,53]. The main reason for choosing DSR is the prescriptive, rather than descriptive character of this investigation. The practical problem addressed by this research work is how house building companies can use customer integration concepts to support the definition of MC strategies and improve value generation for customers.

There are different types of outcomes in DSR, such as models, methods, constructs, instantiations [54] and technological rules [55]. The artefact proposed in this research is a conceptual framework which prescribes a set of core and customer integration decision categories that can be used to support the definition of MC strategies in house building companies. This research work also proposes new constructs and adapts existing ones, which are useful for describing those decisions categories.

Figure 2 provides an overview of the research design, in which the activities are organised similarly to the DSR steps proposed by Lukka [53]: (i) identify a practical problem and understand it from a theoretical perspective; (ii) devise the solution; (iii) test and refine the solution in an empirical study; (iv) analyse the utility of the solution and discuss the theoretical contributions of the investigation.

**Figure 2.** Research design.

A literature review on customer integration and MC practices was carried out in order to obtain a deep understanding of the topic, in the first step of the research (Figure 2). The aim was also to find descriptions of practices that were successfully used for customer integration, by using the snowballing technique, complemented by an advanced search in the Scopus repository. The search was undertaken in journal and conference papers, from 1998 to 2018 and its results were limited to areas relevant for house building such as engineering, management, and environmental science, from which 24 papers were selected. As a result, two sets of practices were identified, one related to the MC core decision categories and the other to customer integration. Information about those practices was stored and further categorised in a database, according to authors, and country of adoption.

In the second step of the research, the selected practices were associated with decision categories (Figure 2). Some of the decision categories considered were identified in the literature review (see Section 2), such as solution space, visualisation approaches, and configuration sequence. Furthermore, the processes of classifying practices into decision categories available in the literature brought to light some gaps, which resulted in the proposition of some additional decision categories.

The third step of the research consisted of the development of an empirical study in a house building company, named Company P, in which the implementation of MC practices and decision categories was assessed (Figure 2). The aim of this study was to understand further the underlying ideas of practices and to test the utility of the proposed decision categories. It was part of a broader research project, in which the MC strategy of the company was assessed, and some improvements were implemented by the company, which took approximately two years.

Company P was founded in the 1970s as a family company, being currently one of the largest construction companies of the South Brasil, with 252.312 m<sup>2</sup> built so far. They have over 20 years of experience in delivering customised residential building projects for upper-middle and middle-class customers. Their products are made from a combination of traditional methods of construction with industrialised components, such as internal drywall partitions and precast façades. This company was chosen because its business strategy was strongly based on the customisation of products to obtain market differentiation. Moreover, the company was willing to take part in this project and had a department entirely dedicated to customising residential projects. The customisation team (CT) had six architects, including a coordinator.

The focus of the empirical study was on a relatively new market segmen<sup>t</sup> explored by the company in which a limited solution space was offered to customers. Within this context, the productivity– flexibility trade-off had to be managed carefully in order to increase the perceived value for customers without substantially increasing costs and lead time.

The empirical study started by assessing and analysing the customisation process adopted by Company P, based on multiple sources of evidence (see Table 2). Several semi-structured interviews were carried out with representatives of different departments of the company. These interviews were divided into three sections: (i) company's general information (e.g., business model, customers, competitors, history); (ii) description of NPD and customisation practices; (iii) description of products and customisation options. Additionally, one open-ended interview was carried out with the customers and customisation manager about the role of the customisation department and the MC strategy. Based on the interviews and documents analysis, a customisation process map was devised by researchers and discussed with the CT. Simultaneously, the existing customer integration practices were compared to a preliminary list of practices extracted from the literature, and a gap analysis was then carried out, resulting in the identification of some improvement opportunities. Those improvements were discussed with Company P's representatives in two meetings. Then, the company decided to implement some of the suggested improvements.


**Table 2.** Sources of evidence used to understand the customisation process and identify improvement opportunities.


**Table 2.** *Cont.*

Approximately one year later, after the implementation of some improvements by the company, a data collection protocol was used to assess Company P's MC strategy regarding core and customer integration categories. This data collection protocol was based on the final set of decision categories and on the full list of practices, being used as a reference to discuss the adoption of practices with the CT (Table 3). This assessment was based on a 5 point scale. Besides, data about the perspective of customers were captured qualitatively during three open days in construction sites, bringing another perspective to the discussions.


**Table 3.** Sources of evidence used on the assessment of the level of implementation of practices.

Analysis and reflection of the research findings were carried out in the fourth step of the research study. The utility of the research outcomes, i.e., decision categories and MC practices, was assessed based on the following criteria: (i) provide underpinnings to the assessment and monitoring of core and customer integration decision categories; (ii) provide support to understand MC related concepts and its underlying ideas; (iii) support decision-making for defining the MC strategy, particularly in terms of integrating customers in customisation processes. The assessment of utility was carried out in six meetings with representatives of the customisation department, as shown in Tables 2 and 3. Finally, the conceptual framework of decision categories for customer integration was devised.
