A TRIZ-Based Method for New Service Design Kah-Hin Chai Jun Zhang Kay-Chuan Tan National University of Singapore This article demonstrates the viability of applying the theory of inventive problem solving (TRIZ) to services by proposing a new approach to new service design. Tradi- tionally, the effectiveness of new service design is unpre- dictable as service design relies largely on inspiration and the past experiences of service designers. By integrating TRIZ problem-solving tools and its knowledge base, the authors propose a new TRIZ-based approach to address this weakness in service design. Through two case studies, the proposed model is verified. This demonstrates the rele- vance of TRIZ to service design. It is hoped that this article will raise awareness among service researchers so that more studies in this direction are conducted. Keywords: service design TRIZ new service develop- ment systematic innovation In the past two decades, the trend of formalizing the process of service development has led to a series of new service development (NSD) models. Although many NSD models comprise a set of development steps starting from product initialization to final commercialization, they are usually oversimplified and lack detailed guidelines for operational use. In particular, recent studies have found that the idea generation stage, despite its importance to the overall service development process, has not been ad- dressed adequately by researchers (Bowers 1989 Edgett 1996 Kelly and Storey 2000). Hence, service developers often have had to use ad hoc processes (Metters, King- Metters, and Pullman 2003) and tended to not engage in formal ways of idea generation (Easingwood 1986). One potential pitfall of such practices is that the novelty and the quality of new service ideas are severely affected. This is due partly to psychological inertia in human thinking, which focuses the mind on only what is known, and avoid- ing unknown paths (Savransky 2000). Indeed, service design has been identified as an important research area where more studies should be conducted on the opera- tional tools employed for successful NSD (Menor, Tatikonda, and Sampson 2002). Similarly, Johnston (1999) suggests that good design tools and techniques in service development need to be further explored. Building on Zhang, Chai, and Tan (2005), this article demonstrates the viability of applying the theory of inven- tive problem solving (TRIZ) to services by proposing a new method to identify, generate, and evaluate possible solutions to service problems. The outcome of the model is a series of possible solutions that can be further devel- oped into service concepts (Clark, Johnston, and Shulver 2000 Edvardsson and Olsson 1996 Johnson et al. 2001). The article is arranged in the following manner. We first review previous research in service design and the relevant tools. This is followed by a description of the theoretical framework of TRIZ methods. Two empirical case studies are used to illustrate the application of TRIZ in service design. Finally, we discuss the contributions and manage- rial implications of the study. Based on the limitations identified, several directions for future research are suggested. Journal of Service Research, Volume 8, No. 1, August 2005 48-66 DOI: 10.1177/1094670505276683 �� 2005 Sage Publications

LITERATURE REVIEW According to Johnson et al. (2000), NSD comprises four major phases: design, analysis, development, and full launch. Although the NSD process cycle might represent a progression of planning, analysis, and activity execution, the process is iterative and nonlinear in nature (Menor, Tatikonda, and Sampson 2002). Each phase revolves around the design and configuration of service concept elements where resources such as development teams and tools play a key function. For example, the design stage involves formulation of new services objectives, strate- gies, idea generation, screening, and concept development and testing. As one of the essentialcomponents in the NSD process, service design focuses on the operational basics of the development work. It involves understanding and planning the interaction of a variety of physical, elec- tronic, and human elements (Edvardsson, Thomasson, and ��vretveit 1994). Other definitions include ���the idea to design high quality into the service system from the outset, and to consider and respond to customer expectations in designing each element of the service��� (Zeithaml, Berry, and Parasuraman 1990, p. 157), the concretization of the service concept in drawing flowcharts (Gummesson 1991), and ���the work of specifying an idea about a new service in drawings and specifications��� (Norling, Edvardsson, and Gummesson 1992, p. 57). Previous work in service design has identified the need to incorporate more customer elements throughout the process of NSD. One good example of this is the development of an inven- tory of activities that needs to be carried out in involving users in NSD projects (Alam 2002). Another example is the use of consumer choice modeling in e-services, which provides insights in understanding consumer choices and preferences for design and operation strategy formulation in NSD (Iqbal, Verma, and Baran 2003). Among the many activities in service design, idea gen- eration is one of the most critical activities. Ideas for new product can arise almost anywhere (Robinson and Moak 1997). They can arise inside the company by taking employees as an important source of new product ideas (McGuire 1973), or they can come from outside the com- pany by listening to customers���feedback or copying from competitors (Easingwood 1986 Sheuing and Johnson 1989). Conventional idea search methods include brain- storming, solicitation of ideas from employees and cus- tomers, lead-user research, and learning about competitors (Zeithaml and Bitner 2000). However, these methods rely very much on service designers���or customers���past experi- ences, which contain unavoidable psychological inertia. Thus, the adoption of these methods might probably restrictthegeneration of creativeand breakthrough ideas. Although idea generation is widely agreed to be impor- tant, it is one of the weakest links in NSD (Bowers 1989). Many service companies tend not to engage in formal idea generation in that they find it relatively easy to generate new service ideas (Easingwood 1986). However, recent studies report that successful service firms establish sys- tems and procedures for stimulating idea generation on a long-term basis (Edgett 1996 Felberg and DeMarco 1992 Robinson and Moak 1997). Techniques for stimu- lating ideas must be created and maintained on an ongoing basis to keep generating innovations (Crawford 1994). Another focus of research in service design isthe devel- opment of tools and techniques. Service design tools such as service blueprinting (Baum 1989 George and Gibson 1991 Kingman-Brundage 1988 Shostack 1984), func- tional analysis (Berkley 1996), structured analysis and design technique (Congram and Epelman 1995), quality function deployment (QFD) (Ermer and Kniper 1998 Franceschini and Terzago 1998 Selen and Schepers 2001 Stamm 1992), and root cause analysis (RCA Duffy 2000) have been developed to assist service designers in devel- oping design concepts for services. Although these ser- vice design tools may be effective in describing and ana- lyzing service problems, they have limited capacity in generating ideas and overcoming pitfalls identified in the design process. This limitation in overcoming psychologi- cal inertia affects both the quantity and the quality of sub- sequent design solutions. The challenge is compounded by the ill-structured nature of design problems, in which often one or more steps (or states) are either unknown or incoherent, there is insufficient information in the initial state, and the properties of the goal state are never fully specifiable in advance (Goldschmidt 1997). For example, as part of QFD analysis, the house of quality (HoQ) is fre- quently used by service researchers to identify problems, such as unveiling contradictory relationships among dif- ferent service product attributes (Rovira and Aguayo 1998). However, apart from problem identification, it is noted that HoQ often fails to provide effective solutions to eliminatethe contradictions without making compromises between conflicting requirements. Another example is RCA, which is often used in identifying potential service failure points. The preventive solutions generated are based mainly on past experiences of designers rather than as the result of RCA. Moreover, because RCA is closely allied to the optimization of existing processes (Mann 2002), if service designers rely wholly on the outcome of its analysis, it would be very difficult to find novel breakthrough solutions. Chai et al. / A TRIZ-BASED METHOD FOR NEW SERVICE DESIGN 49

SERVICE DESIGN BASED ON TRIZ To overcome the limitations in service design tools and idea generation practices, TRIZ, a systematic problem- solving methodology, is proposed in this study to close the gaps. TRIZ was first developed by Genrich Altshuller and his colleagues in the former USSR, starting in 1946. The hypothesis of TRIZ research is that there are universal principles of invention that are the basis for creative inno- vation. If these principles could be identified and codified, they could be taught to people to make the process of invention more predictable. Through the analysis of more than 2 millionpatents, a number of innovation patterns and laws of ideality were identified. In addition, the work also reveals the following: ��� Problems and solutions were repeated across indus- tries and sciences. ��� Patterns of technical evolution were repeated across industries and sciences. ��� Innovations used scientific effects outside the field where they were developed. Therefore, TRIZ can be described as a structured problem-solving process with the integration of a set of problem definition and resolution tools that was created on the basis of patent analysis. A distinct characteristic of TRIZ problem solving is systematic problem resolution without compromise. The entire problem-solving process is guided by TRIZ tools, which direct the problem solver to explore solutions in directions that have previously been proven successful. A typical TRIZ problem-solving pro- cess comprises three stages: problem definition, problem resolution, and solution evaluation. At each stage, a vari- ety of TRIZ tools can be used. For instance, the Innovative Situation Questionnaire, the idealfinal result, and function analysis are frequently used in problem definition. The 40 inventive principles, 4 separation principles, the algorithm of inventive problem solving (ARIZ), and 76 standard solutions are often used in problem resolution. The Relevance of TRIZ to Services According to TRIZ methodology, most problems have so-called inherent contradictions. A typical contradiction is a conflict due to two conflicting requirements to the same element in a system. It could result also from two conflicting elements in the same system. The identifica- tion of an inherent contradiction is a critical step in TRIZ problem analysis. Once a contradiction is formulated, selected TRIZ knowledge-based tools can be used to elim- inate the contradiction. The resolution of the contradiction usually leads to the resolution of the main problem as well as other minor problems. Not surprisingly, contradictions in technical areas are more tangible and easier to formulate. These contradic- tions are often clear-cut conflicts between two design parameters. These have been summarized in the original TRIZ contradiction matrix (Altshuller 1997). However, it is possible to formulate problems in services in the form of contradiction despite its intangible nature. For exam- ple, ���mass customization��� is a typical service contradic- tion with implications in many service sectors. How can a service firm deliver personalized service offerings, based on individual customer requirements, to a large group of customers without incurring additional costs brought about by the personalization? Is there any breakthrough solution that could lead to a win-win situation without compromising either ���standardization��� (to achieve econ- omy of scale) or ���customization���? Because TRIZ has the ability to eliminate contradictions and generate break- through solutions, it is possible to use TRIZ to find cre- ative win-win solutions. To explore this opportunity, a revised version of the 40 inventive principles (see Appen- dix A, adapted from Zhang, Chai, and Tan 2003) was developed with the aim of applying them in resolving ser- vice contradictions. The other positive overlap between TRIZ and services is the research on the patterns of innovations. Through the analysis of many examples, Berry and Lampo (2000) cate- gorized five typical ways of redesigning service offerings (i.e., self-service, direct service, preservice, bundled ser- vice, and physical service). This suggests that the patterns of service innovations can be predicted. This assumption is similar to the underlying philosophy of the TRIZ 40 inventive principles. The comparison shown in Table 1 indicates many similarities between these five service re- design patterns and the 40 inventive principles. One impli- cation of this comparison is that if service designers under- stand the patterns of innovations that TRIZ has successfully demonstrated for physical products, the process of service design and innovations can be made more predictable. Service Design Based on TRIZ TRIZ can help to eliminate the psychological inertia in the minds of service designers, thus enhancing the capac- ity of service idea generation in service design. The knowledge-based toolkit provided by TRIZ caters exactly to the needs of helping problem solvers to overcome their own psychological inertia, which is often considered the hardest part in solving difficult problems (Altshuller 1984). Mann and Dewulf (2002) also argued that in terms of its toolkit and method, TRIZ is the most comprehensive of all available models. Compared to traditional idea gen- eration methods such as brainstorming, synectics, lateral 50 JOURNAL OF SERVICE RESEARCH / August 2005