Research on a Service Touchpoint Design Model Driven by Smart Technology Based on Kano–Failure Modes and Effects Analysis Authors: Jian Chen, Zhihan Li, Weiwei Wang, Yi Wang, Zhaoxuan He Published: 2024 ABSTRACT With the development of smart technology and the increasing variety of everyday products, factors influencing product service touchpoint design have become more diverse and complex. Existing service touchpoint design methods and models often focus narrowly on user research, co-design, and risk analyses, lacking a systematic approach. Consequently, they struggle to deliver solutions that align with user needs. This misalignment may result in issues such as increased cognitive load during product use, a diminished user experience, and lower evaluations of the product. In response, this paper proposes a service touchpoint design model, the “BEDFITA” model. It starts with user behavior and follows a structured, systematic process that includes understanding user behavior, recording user emotions, matching user needs, designing product functions, planning interaction experiences, designing service touchpoints, and analyzing failure risks. The Kano model is employed in the user requirement identification phase to provide more precise user requirement parameters, while FMEA is employed in the failure risk analysis phase to generate more accurate failure risk assessments. This ensures that the final service touchpoint design meets user needs and offers reliability and robustness. Finally, the feasibility and effectiveness of the proposed model are validated through a case study on the service touchpoint design of a smart desk. 1. Introduction The essence of service touchpoint design is to shape the complete user behavior during activities such as daily life, work, entertainment, leisure, healthcare, or learning. This behavior encompasses several stages: the intention phase before the activity, the planning phase, the action phase, the completion phase, and finally, the evaluation phase. Service touchpoints are the tangible carriers through which products provide users with visual, tactile, auditory, olfactory, gustatory, emotional, and imaginative experiences at each stage [1]. The form, shape, color, material, and other characteristics of these tangible carriers are significantly influenced by user preferences and needs within their complete behavior pattern. As a result, due to the diverse preferences and behaviors of different users, the design of service touchpoints is inherently varied and complex [2]. With the development of smart technology, several service touchpoints have acquired intelligent attributes. Compared to traditional touchpoints, those with smart features offer users more varied interaction behaviors and modes, resulting in a more convenient, comfortable, and seamless user experience [3]. While smart technology has the potential to enhance user experience, it also increases the complexity of service touchpoint design. This growing complexity necessitates a systematic design model when designing smart touchpoints [4]. Such a model must take user behavior as the starting point, extracting user needs based on the emotions experienced at each behavioral stage. The product’s functions and tangible carriers should be designed according to these needs. Finally, this process ensures a service touchpoint design that meets user requirements, resulting in a more convenient, intuitive, efficient, and comfortable user behavior, thereby improving the overall user experience. The current design of service touchpoints largely relies on the experience of designers and traditional product design methods. The design process typically focuses on aspects such as the shape, size adjustment, and material selection of service touchpoints, but it lacks a foundation in research on user behavior and user needs to inform the design of touchpoint interactions. As a result, many existing products fail to meet user needs or are not aligned with user behavior habits, leading to confusion and dissatisfaction during use. In some cases, designs that do not align with user behavior habits may even mislead users, causing operational errors and unnecessary negative impacts. Therefore, research on service touchpoint design methods and models driven by smart technology is both necessary and significant. 2. Research Status Current research on user needs and user behavior primarily revolves around the Kano model and the FMEA (Failure Modes and Effects Analysis) method. Design researchers typically use the Kano model to capture user requirements, while the FMEA method is employed to analyze potential failure modes and risks in both the design and manufacturing processes, as well as during product use. The Kano model is an efficient tool for classifying and quantifying user expectations, offering a more precise identification of user needs. It has been widely used in product design for user needs assessments, emotional analyses, and design critiques, as well as in industries such as healthcare, management, and services for analyzing personnel needs. For instance, Yao et al. [5] used the Kano model to design a questionnaire for hospital emergency departments, which successfully captures patient needs and improves the patient experience. To verify the feasibility of this method, a sample of 100 people was employed at Guizhou Provincial People’s Hospital. Using the Kano model, 19 patient needs were identified and categorized, effectively improving the patient experience. Similarly, Zhang et al. [6] applied the enhanced Kano model to address customer needs in the e-commerce field. By extracting customers’ expectations from product reviews, they further identified user needs, aiding subsequent product updates and iterations. Moreover, Xiao et al. [7] designed a survey questionnaire using the Kano model to address social demands for aging health management amidst accelerating global aging. They collected data from 143 users and analyzed it across four dimensions, ultimately determining the design direction for aging health management platforms and offering development and optimization strategies for such platforms. Adding to that, Ce et al. [8] conducted a survey of local supermarket consumers using the Kano–QFD model to increase competition in the retail market. This approach aligned supermarket service offerings with consumer needs and optimized the retail service experience to meet the rising demands for service content and quality. In addition, Wen et al. [9] employed the Kano model as a framework to study tourists’ preferences and emotional needs for creative products at a cultural memorial hall. They designed a series of cultural creative products to enhance the hall’s cultural influence and publicity. Furthermore, Qi et al. [10] discussed users’ emotional needs for health water bottles, emphasizing the strong correlation between user senses and product design. By combining the Kano model with the AHP model, they piloted a quantitative analysis of user needs, extracting design elements to serve as a basis for proposals and providing designers with an effective method for innovation. Moreover, Hariri et al. [11] improved the Kano model and the SWARA method to analyze 112 Customer Requirements (CRs), classifying them into different attributes. This improved production systems in the automotive and manufacturing industries, enhancing product quality and market competitiveness. Finally, Dai et al. [12] designed a Kano–AHP model to build an industrial product design and service platform. By analyzing and computing user needs and performance, they verified the platform’s effectiveness in meeting customer service demands. The Failure Mode and Effects Analysis (FMEA) method, along with the Process Potential Failure Mode and Effects Analysis (PPFMEA) method, is a systematic, proactive tool used for risk management to prevent potential issues. In recent years, local and international scholars have applied FMEA to service design, offering guidance on improving service effectiveness. For instance, Haizhe et al. [13] addressed the limitations of FMEA in medical device design by proposing a structured human reliability assessment model, which optimizes a failure risk analysis by introducing scenario parameters. This improvement resolved issues of numerical repetition and weight allocation encountered in traditional FMEA, resulting in more accurate and reliable medical device design. Moreover, I. D. P et al. [14] enhanced the failure risk analysis in FMEA by aligning different detection methods with various types of risks and incorporating a comprehensive impact of different factors to determine failure probabilities. Adding to that, Padash et al. [15] introduced a hybrid fuzzy risk assessment method for the high-risk steel industry, which more accurately quantifies expert judgments in uncertain environments, refining FMEA’s risk assessment outcomes and effectively lowering risk incidents. Furthermore, Kornek et al. [16] highlighted the shortcomings of FMEA in radiotherapy risk assessment, and by testing and evaluating a new event reporting system (IRS), demonstrated its ability to validate FMEA results. In addition, Barsalou et al. [17] improved FMEA’s analytical structure by dividing the process into system analysis, failure analysis, risk mitigation, and risk communication stages. They refined each stage through a seven-step method, enhancing the systematic nature of FMEA. Moreover, Wencai et al. [18] discussed improved vehicle fault mode evaluation from the perspective of user driving experience. They noted that traditional FMEA methods struggle to achieve this task. Therefore, they combined cost FMEA, FAHP, and EFMULTIMOORA. Cost FMEA was used to identify vehicle fault modes, FAHP calculated the weights of various criteria, and EFMULTIMOORA classified these fault modes. Finally, Heiko et al. [19] discussed the importance of fault detection and risk ranking in Industry 4.0 and introduced several methods and models for fault detection and risk ranking, highlighting FMEA’s effectiveness in analyzing failure risks in product design and manufacturing. The research outcomes mentioned above provide valuable insights into product service touchpoint design, smart-driven product design, and user experience optimization methods. Additionally, they offer several decision models for design research. However, in the context of smart technology-driven product design, there is a lack of systematic design models and processes. Existing methods are often too narrow in scope, making it difficult to address the evolving design challenges of intelligent products. Consequently, many market products fail to meet user expectations. To address this gap, the BEDFITA model is proposed for designing product service touchpoints in smart technology applications. This model organizes the design process into phases, systematically outlining research objectives, methods, and outcomes for each phase. This approach enhances the overall integrity, reliability, and feasibility of product design. Ultimately, solutions generated through the BEDFITA model better align with user needs, leading to an enhanced user experience. 3. Service Touchpoint Design Driven by Smart Technology Smart technology, rooted in statistical methods, integrates disciplines such as big data, the internet, the Internet of Things (IoT), and neural networks to emulate human intelligence. This allows smart technology to endow artificial objects, such as products, tools, and software, with intelligent, human-like characteristics [20]. In the realm of product service touchpoint design, the advancement of smart technology has transformed the way humans interact with products. The relationship between users and products has evolved from a complex “control and be controlled” dynamic to one of “inquiry and learning”. Consequently, smart technology-driven product service touchpoint design provides a wider range of interaction options, while increasing the diversity and complexity of these touchpoints. Under these conditions, traditional design approaches, which rely primarily on designers’ experience, often struggle to ensure the reliability of design solutions. Single-stage design models and methods are insufficient to guarantee that the solutions meet user needs [21]. Users interact with product functions to solve problems in social and survival activities, which is the fundamental purpose of using a product. Similarly, these interactions are at the core of the user experience, with service touchpoints being the focal points of these interactions, heavily influenced by user behavior preferences. Therefore, product service touchpoint design must be user-centered, beginning with an analysis of user behavior and the emotions they experience while addressing issues in their daily lives. This approach uncovers user needs across different attributes, enabling the creation of design solutions that meet those needs. To this end, this paper proposes the BEDFITA model, a comprehensive framework for designing product service touchpoints. This model includes a user behavior analysis, emotion recording, needs matching, product function design, interaction experience planning, service touchpoint design, and a failure risk analysis. By adopting this structured approach, the design process becomes more holistic and logical, ensuring that design solutions better align with user needs to optimize the user experience. Moreover, the BEDFITA model is divided into two phases:(1)Function Design Phase Based on User Behavior: Start by clearly identifying the specific behaviors users exhibit when interacting with a product or engaging in an activity. This involves systematically mapping the complete behavioral process, from the starting point to the endpoint of the interaction. For each behavioral stage, employ the Kano model to ask both positive and negative questions, capturing users’ emotions and reactions at each stage. Based on users’ varying emotions at different behavioral points, identify and match their specific needs. This information is used to design the product’s functions.(2)Service Touchpoint Design and Failure Analysis Phase Based on FMEA: Using the functional design results from the Function Design Phase, a comprehensive interaction experience map is created for users. Next, an appropriate service touchpoint is selected for each interaction behavior point in the interaction experience map, resulting in a complete mapping of the product’s service content and touchpoints. To ensure reliability, a failure risk analysis is conducted on each service touchpoint using FMEA. If any service touchpoint presents a high failure risk, re-evaluate and select alternative touchpoints until all pass the failure risk analysis stage. This process leads to a robust, failure-resistant service touchpoint design solution. (1)Function Design Phase Based on User Behavior: Start by clearly identifying the specific behaviors users exhibit when interacting with a product or engaging in an activity. This involves systematically mapping the complete behavioral process, from the starting point to the endpoint of the interaction. For each behavioral stage, employ the Kano model to ask both positive and negative questions, capturing users’ emotions and reactions at each stage. Based on users’ varying emotions at different behavioral points, identify and match their specific needs. This information is used to design the product’s functions. Function Design Phase Based on User Behavior: Start by clearly identifying the specific behaviors users exhibit when interacting with a product or engaging in an activity. This involves systematically mapping the complete behavioral process, from the starting point to the endpoint of the interaction. For each behavioral stage, employ the Kano model to ask both positive and negative questions, capturing users’ emotions and reactions at each stage. Based on users’ varying emotions at different behavioral points, identify and match their specific needs. This information is used to design the product’s functions. (2)Service Touchpoint Design and Failure Analysis Phase Based on FMEA: Using the functional design results from the Function Design Phase, a comprehensive interaction experience map is created for users. Next, an appropriate service touchpoint is selected for each interaction behavior point in the interaction experience map, resulting in a complete mapping of the product’s service content and touchpoints. To ensure reliability, a failure risk analysis is conducted on each service touchpoint using FMEA. If any service touchpoint presents a high failure risk, re-evaluate and select alternative touchpoints until all pass the failure risk analysis stage. This process leads to a robust, failure-resistant service touchpoint design solution. Service Touchpoint Design and Failure Analysis Phase Based on FMEA: Using the functional design results from the Function Design Phase, a comprehensive interaction experience map is created for users. Next, an appropriate service touchpoint is selected for each interaction behavior point in the interaction experience map, resulting in a complete mapping of the product’s service content and touchpoints. To ensure reliability, a failure risk analysis is conducted on each service touchpoint using FMEA. If any service touchpoint presents a high failure risk, re-evaluate and select alternative touchpoints until all pass the failure risk analysis stage. This process leads to a robust, failure-resistant service touchpoint design solution. The model is named the BEDFITA model, as it follows a sequential logic of key terms: Behaviors → Emotions → Demands → Functions → Interaction → Touchpoints → Analysis. The specific technical framework of this model is illustrated in Figure 1. 4. Process of Constructing Functional Solutions Based on User Behavior User behavior is the starting point and core focus for constructing product function solutions. In this paper, user behavior mainly refers to two aspects: the interaction behavior when using a product and specific behaviors during survival and production activities [22]. When users interact with a product, due to factors such as design flaws in the product, new demands driven by market competition, and users’ evolving needs, the original product may no longer provide optimal service. This reduces the efficiency and experience of user interaction. For example, compared to feature phones with physical buttons, smartphones introduced a new interaction behavior for users, offering more efficient and improved user experiences than button-based phones [23]. In survival and production activities, the lack of assistance can result in lower behavioral efficiency and poor user experiences. For example, when children are learning to walk upright, they require support and protection, leading to the design of products like baby walkers [24]. User behavior generates emotions, and these emotions reflect user needs. Naturally comfortable user behavior generates relaxed and pleasant emotions. Under such emotional states, the user’s need may be to prolong the activity. Conversely, prolonged and complex user behavior results in fatigue, and under such emotional states, the user may require a reduction in the complexity of the behavior or seek to avoid the behavior altogether. At this point, the various functions of a product are needed to meet the user’s needs under different emotional states [25]. As described above, the process of constructing a product’s function solution must begin by organizing and analyzing user behavior, recording the emotions generated by these behaviors. The Kano model is then used to describe and quantify these emotions, revealing the various user needs reflected in different emotional states. Finally, product function solutions are developed based on these user needs. The process of constructing function solutions based on user behavior is illustrated in Figure 2. The specific steps and content of constructing functional solutions are as follows: 4.1. User Behavior Analysis To design product service touchpoints driven by smart technology, it is essential to thoroughly understand user behavior when using a specific product or engaging in a specific activity. This behavior comprises multiple behavioral nodes. Users typically begin at a fixed behavioral node, may traverse or skip through intermediate nodes, and ultimately conclude at another fixed behavioral node. Therefore, the goal of analyzing this complete behavior is to ensure the sequence of each behavioral node. Let user behavior be denoted as B, and the m-th (where m = 1, 2, 3, …, n) behavioral node is denoted as Bm. 4.2. User Emotion Recording Throughout the process of engaging in the aforementioned behavior, users will experience different emotions at each behavioral node. Each node has a corresponding relationship with the user’s emotions. Therefore, user emotions are denoted as E, and the emotion experienced by the user at the m-th behavioral node is denoted as Em. 4.3. User Demand Matching Based on the Kano model, user emotions at each behavioral node are categorized into the following: Like (Lk), Neutral (Nr), Expectation (Hp), Indifference (Id), Confusion (Ss), and Dislike (Ng). Each behavioral node has two dimensions: “With this behavior” and “Without this behavior”. User evaluations for each dimension can be classified as “Satisfied”, “Normal”, “Indifferent”, “Tolerable”, and “Dissatisfied”. A user demand matching table was created to record the emotions associated with each behavioral node, serving as the basis for aligning user demands. This table is shown in Table 1. Let the user group participating in this user demand matching record be denoted as H, with the number of users denoted as h, where i = 1, 2, 3, …, h. Let Eim represent the evaluation of the i-th user for the m-th behavioral node. Users’ evaluations for each behavioral node are recorded sequentially in the user demand matching table. After completing this table, the ratio of each type of user emotion to the number of users for each behavioral node is computed and expressed as a percentile, as shown in Equation (1). (1)Lkm%+Hpm%+Nrm%+Idm%+Ngm%+Ssm%=100% where represents the percentage of the user group who expressed “Like” for this behavioral node, denotes the percentage of users who expressed “Expectation”, indicates the percentage of users who expressed “Neutral”, represents the percentage of users who expressed “Indifference”, is the percentage of users who expressed “Dislike”, and highlights the percentage of users who expressed “Confusion” for this behavioral node. The satisfaction coefficient Sam is introduced to represent users’ satisfaction when the m-th behavior is present, and the dissatisfaction coefficient Dsm is introduced to represent user dissatisfaction when the m-th behavior is absent. The user demand coefficient at the m-th behavioral node is denoted as Udm, and the specific calculations are shown in Equations (2)–(4) [26]. (2)Sam=Hpm%+Lkm%Hpm%+Lkm%+Idm%+Nrm% (3)Dsm=Nrm%+Lkm%Hpm%+Lkm%+Idm%+Nrm% (4)Udm=Sam+Dsm The design of product functions must closely align with user needs. Functions based solely on designers’ experience may lead to wasted functionality and increased costs. From a user experience perspective, features that do not address user needs can adversely affect users’ judgments and behaviors. Therefore, the demand coefficients obtained above should be ranked, and product functions should be developed according to this ranking to create an effective product function plan. Product functions are denoted as F, with the j-th product function represented as Fj (where j = 1, 2, 3, …, k). The product function solution constructed at this stage will serve as the foundation for the subsequent service touchpoint design. 5. Service Touchpoint Design and Failure Risk Analysis Process In the previous discussion, we completed the construction of the product function solution based on user behavior, resulting in a solution that contains k different functions. However, when users interact with a product’s functions, they can choose from various interaction methods. For example, controlling water flow is a function of a faucet, and within this function, there are multiple interaction options such as turning a knob, toggling a valve, or using smart sensors [27]. Due to the variation in interaction methods, the specific form of the product’s service touchpoints will also differ. For instance, the interaction method of turning a knob requires the service touchpoint to be in the form of a knob, whereas the smart sensor interaction method necessitates a motion-capture camera as the service touchpoint form [28]. Therefore, the design of service touchpoints must be based on the specific product function solution. First, based on the function solution, we sequentially plan the different interaction methods users will employ when using all functions, resulting in a complete interaction process for the user. Then, for each interaction method corresponding to the product functions, we sequentially design the form of the service touchpoints, leading to the creation of a service touchpoint design plan [29]. During the interaction between the user and the product’s service touchpoints, the touchpoints remain subject to failure risks. A failure in the service touchpoint could impair certain product functions and disrupt the interaction between the user and the product, reducing the overall user experience. Therefore, once the service touchpoint design plan is completed, it cannot proceed directly to production and manufacturing without conducting a failure risk analysis. Due to factors such as the user’s knowledge, the form of the service touchpoint, and other variables, the failure risks of service touchpoints are uncertain. To address this, the FMEA (Failure Modes and Effects Analysis) method is applied to detect the failure causes of each touchpoint in the design plan, analyze its failure modes, and predict the potential consequences of its failure. The failure risks of service touchpoints are quantified by evaluating the severity (S) of failure risks, the detectability (D) of failure causes, and the occurrence (O) frequency of failure modes [30]. This process yields a failure risk value for each touchpoint. Touchpoints with high failure risk values have lower reliability. For such touchpoints, they are returned to the corresponding design phase for redesign. The new design plan undergoes a failure risk analysis, and this iterative process is repeated until the failure risk value decreases. Alternatively, reliability can be ensured through a design that improves robustness, allowing normal interaction between the user and the touchpoint. Touchpoints with low failure risk values exhibit high reliability, so their design plans are retained. Following this process, the final output is the product’s service touchpoint design result [31]. The flow of the service touchpoint design and failure analysis is illustrated in Figure 3. 5.1. Interaction Behavior Planning The design of product service touchpoints should first plan user interaction behaviors for each product function according to the functional solution constructed earlier. The relationship between user interactions and product functions is not always one to one. Multiple interactions may correspond to a single product function. For example, when using a smartphone, various interactions such as video calls, watching movies, and browsing information all correspond to the phone’s image display function. At the same time, a single interaction can correspond to multiple product functions. For instance, when using a smart desk lamp, the single action of repeatedly touching the light adjustment button can correspond to several functions, including adjusting the color temperature, brightness, lighting effects, and setting a timer [32]. Therefore, based on the k product functions included in the functional solution, user interaction behaviors should be planned to ensure that each interaction for using the product functions is as easy to understand, easy to adapt to, and as healthy and comfortable as possible. The relationship between product functions and interaction behaviors is illustrated in Figure 3. The total interaction behavior for using all product functions is denoted as Ib, where Ibz represents the z-th user interaction behavior (z = 1, 2, 3, …, y). 5.2. Service Touchpoint Design The user experience journey map outlines the complete flow of user interaction with product functions to achieve their goals. This interaction flow comprises multiple interaction behavior nodes, each requiring material mediums based on visual, tactile, auditory, olfactory, and gustatory inputs. These material mediums are known as service touchpoints. Based on the interaction experience flow developed earlier, service touchpoints should be selected or designed for each behavior node to create a detailed user interaction behavior journey map. This journey map illustrates the service touchpoints associated with each interaction behavior node, collectively forming a provisional service touchpoint design plan. Service touchpoints are denoted as P, with Pw representing the w-th service touchpoint (where w = 1, 2, 3, …, v). This provisional design plan will undergo a failure risk analysis in the subsequent steps. 5.3. Service Touchpoint Failure Risk Analysis To assess the reliability and stability of the service touchpoint design plan, a failure analysis is required. The FMEA method is commonly employed to identify risks in service systems, product design, and user experience processes [33]. Therefore, this paper utilizes the FMEA method to analyze potential failures for each service touchpoint in the design plan. Since user evaluations often contain ambiguous terms such as “should”, “seems”, “particularly”, “very”, and “slight”, this paper employs triangular fuzzy numbers for the calculation and analysis. A unified evaluation standard or fuzzy term set is constructed for the severity (S), occurrence (O), and detection (D) of failures in the FMEA method. These three indicators are treated as fuzzy linguistic variables. A fuzzy linguistic term set with five evaluation terms is used for each variable, with scores ranging from [0, 10]. The corresponding fuzzy evaluation terms are displayed in Table 2, where the minimum fuzzy number is L, the middle fuzzy number is N, and the maximum fuzzy number is U. Triangular fuzzy numbers, introduced by Zadeh in 1965 as part of the fuzzy sets theory, address uncertainties in various environments. They are extensively used in quality and risk management due to their ability to accurately represent relationships between elements and sets within fuzzy sets [34]. Therefore, triangular fuzzy numbers are employed to quantify user evaluations in the failure analysis of service touchpoints. The membership function for the fuzzy terms listed in Table 2 is used to quantify and describe user evaluations. The membership function for a user evaluation X is given by Equation (5) as follows:(5)μÃ(x)x−LN−L L≤x≤NU−cU−N N