گسترش تجزیه و تحلیل مهندسی تجارت کردن با یکپارچه سازی ترجیحات کاربر در تحلیل متقارن

The ongoing technical improvements in architecture design with improved features of mobile or smartphones do not automatically guarantee user acceptance, because technical and commercial aspects primarily drive the development of mobile communication systems and devices. Especially in early stages of technology development, user preferences and values are not adequately considered, which might even have a negative impact on acceptance issues. The aim of this study was the implementation of a quantified understanding of user needs in terms of values into the system design process of cell-phone processors. Moreover, we aimed for an extension of the engineering’s trade-off analysis by using conjoint analysis in order to investigate trade-offs between specific device characteristics. Finally, our aim was the evaluation of empirically based user-oriented research methods. Results of the first study revealed that battery life, speech quality, signal quality and data-transmission rate are the most important device characteristics. Results from conjoint analysis indicated a clear trade-off between battery life and the three other characteristics. Moreover, this research demonstrated that technology acceptance research benefits considerably from an interdisciplinary and multi-method approach. Besides, implementing the users’ preferences into early stages of the product development process offers several advantages concerning effectiveness as well as economic aspects of development.

FACING the continuous improvement and growth of mobile phone networks, the demand for technical developments in the area of mobile devices rises as well. Since mobile internet is accessible via mobile or smartphone, a multitude of services and applications has been developed and is used by a growing number of users (Cisco visual networking index, 2012). Technical improvements and changing user demands require a higher performance of mobile devices, which, in turn, require new and more powerful system architecture designs. Additionally, short technical life cycles and growing market pressure demand fast and customer tailored solutions. There are many options to improve today’s mobile systems. One of the main optimization targets in mobile system design is throughput, e.g. Universal Mobile Telecommunications System (UMTS) towards UMTS LTE (Long Term Evolution). Furthermore, system designers can choose algorithms for implementation that enhance the connection stability in certain environments (e.g. high velocity). If, for example, a mobile phone user is in an area with weak radio signals or he is travelling with high velocity, algorithms could improve the connection stability. On the other hand, such algorithms increase the computational load of the system, which directly leads to higher energy dissipation. This has to be taken into account during the system design process.

The system design process for mobile device innovations typically focuses on an optimization of technical parameters. Due to feasibility reasons and growing financial pressure in R&D, the selection of these optimization targets is primarily based on technical and economical considerations. The inclusion of user preferences usually happens at the end of the design process, when the user is asked to evaluate the product in user-tests or – in the end – on the shop floor, when the user decides to buy a product or not. However, in these late stages of product development, the design process is usually finished or only marginal changes are made to the product. As purchase decisions by the user are strongly determined by specific mobile device characteristics, the inclusion of user-oriented input into early stages of the design process might lead to improved design decisions. Therefore, the aim of this study was the implementation of a quantified understanding of user needs in terms of values into the system design process. Moreover, we aimed for an extension of the engineering’s trade-off analysis by using conjoint analysis in order to investigate trade-offs between specific device characteristics, e.g. “How much energy efficiency in terms of battery life is the user willing to lose for an improvement of device features e.g. signal-quality?” Finally, our aim was the evaluation of empirically based user-oriented research methods. We focused on the question whether conjoint analyses yield additional insights into acceptance issues in early stages of the system design cycle and whether its results provide improved practical guidelines for system developers in comparison to conventional approaches in system engineering. Therefore, in this study two empirically based user-centered methods were applied (user survey and conjoint analysis), which were supposed to provide valuable information in supporting decisions regarding optimization targets in system design. The first study was a user preference survey, which focused on the identification of relevant mobile device characteristics. In the second study, these device characteristics were used for a trade-off analysis by conducting a conjoint analysis. First, the results and their practical implications for mobile device system design are discussed in detail. Secondly, the additional value of conducting conjoint analysis for this purpose in comparison to conventional surveys will be evaluated. Finally, managerial implications of these results are discussed. 5.1. Contributions to system development The findings of the user survey (study I) showed that users have in fact clear preferences in mind regarding mobile device characteristics. The most important characteristics, which affect purchase intentions, were battery life, speech quality, connection stability and connection quality. Taking different user groups into account (mobile phone vs. smartphone users), we found that smartphone users more strongly prefer device features which are related to mobile internet usage such as internet access and data rate. Although current mobile device design already strongly focuses on the development of smartphones for Internet usage, the finding confirms the market potential of devices suitable for the demands of the mobile internet. Based on the findings of the user survey we derived four hardware characteristics to be further evaluated by conjoint analysis: Battery life, speech quality, signal quality and data-transmission rate. Results indicated that speech quality and signal quality are the most important features in the decision for one of the queried ‘service packages’ for both user groups. Speech quality, however, was important insofar as users do not want to pay for an additional service. The more detailed analysis of part-worth utilities gives an overview of which attribute features are only just acceptable for users of smartphones or mobile phones. Results indicated that both user groups would accept a minimum of 9 h of battery life and for signal quality at least a partial good signal in dead spots, whereas a good signal was preferred more strongly. Due to technical feasibility respondents had to choose whether they would opt for 9 h battery life or a good signal in dead spots. This interesting question was answered with a sensitivity analysis that revealed that a device which is constant in all other characteristics (speech quality and data-transmission rate) would be more valuable to smartphone users when it provides good signal in dead spots, even if it has only 6 h of battery life. The opposite was shown for mobile phone users. The same trade-off revealed a clear preference for a longer battery life in this user group. Today’s smartphones provide battery runtime for a talk time between 7 and 12 h, which is less than conventional mobile phones offer. The design of a new phone typically has to target this battery runtime as well, because it is already limited due to a greater variety of different task within a smartphone. The system designer therefore has to trade-off the reception quality versus energy dissipation carefully. In the presented study we brought this trade-off decision to the users. The results show that the smartphone users could live with far less runtime in exchange for a better reception. It has to be assumed that smartphone users have a greater tolerance towards less battery runtime in contrast to mobile phone users because they are used to it and the benefit of the multifunctionality of a device outweighs this aspect. Therefore, smartphone users are the more interesting user group in this study because it is essential to ascertain the minimum acceptable amount of battery runtime in contrast to other optimization targets on the basis of a sample that is used to current technical standards. 5.2. Contributions to technology acceptance research The present research work demonstrated that technology acceptance research benefits considerably from an interdisciplinary and multi-method approach. The combination of research questions from system engineering, i.e. the specification of optimization targets in system design, with technology acceptance research methods provided a novel research framework regarding the investigation of acceptance-relevant trade-offs for users. Moreover, the application of a multi-method-approach, i.e. the combination of two empirical methods, allowed (a) to identify acceptance-relevant hardware features of mobile devices (in the user survey), and (b) to get concrete information about the specification or configuration of these features for the design of a novel product (in the conjoint analysis). In most acceptance and usability studies the potential of a multi-method-approach regarding a holistic and productive investigation of acceptance issues is disregarded as only single methods are applied (e.g. Gefen and Straub, 1997 and Venkatesh and Davis, 2000). As a further contribution to acceptance research, this work was a first step to overcome criticism on established acceptance models such as the TAM (Bagozzi, 2007 and Benbasat and Barki, 2007). These acceptance models neither provide information about the evaluation of single product characteristics nor do they deliver decision criteria or concrete practical guidelines for system designers. In our study, user preferences regarding single device features were uncovered and concrete product configurations were specified which were attractive for potential users. Moreover, technology acceptance models are primarily job-related, which implies a mandatory system interaction by the user. As today’s mobile devices are also widely privately used on a “voluntary” basis, user preferences referring to private usage should be included in acceptance research. This also requires a differentiated view on specific user groups (e.g. mobile phone vs. smartphone users), which significantly differed according to their usage profiles and preferences. The successful integration of acceptance research methods and knowledge into early stages of system design showed that acceptance research should not start at the end of the system development process. As present acceptance approaches mainly have a static view (Benbasat & Barki, 2007), an early and iterative inclusion of acceptance issues throughout the system development process will also contribute to a more valid, dynamic perspective on technology acceptance. 5.3. Managerial Implications Our findings have managerial implications, particularly in terms of R&D management. Implementing the users’ preferences into early stages of the product development process offers several advantages concerning effectiveness as well as economic aspects of development. R&D managers should consider the users’ input not only in late stages of system design lifecycles but also for decision processes in the early design phase. This study demonstrated how the design phase could be enriched: A user survey focused on the question which features are relevant for the user. In a second user study we demonstrated that conjoint analysis is an adequate tool for finding the minimum requirements users have regarding technical specifications as well as answering the question “which features are more important to the user”. Furthermore, applying conjoint analysis demonstrated how the engineers’ trade-off analysis can be facilitated and delimited by taking the users’ perspective into account since the engineer does not have to decide between all technical possibilities but now has a framework that tells him e.g. “how much battery life the user is willing to lose for a better signal quality in dead spots”. By applying user preferences to early development stages the product development process could be more efficient and, as a consequence, more economic due to reduced potential misspecifications that otherwise will be detected in later stages when conducting user tests with prototypes or, the worst case, on the shop floor. Especially with regard to growing market pressure and the increasing demand for fast and customer tailored solutions the integration of user values into the development process could be crucial for a successful market launch of new products. Differences in results between potential (mobile phone user) and current users (smartphone users) suggest that a differentiated view of the individual user groups is indispensible. Although this study demonstrates a successful implementation of user input into the development of mobile device architecture, managers should consider that the procedure depends on several aspects. For example it might be a difference whether it is about the development of a new or an already existing (or established) technology (Callahan & Lasry, 2004). In the case of conjoint analysis it has to be assured that potential users have sufficient knowledge of the attributes in order to make a valid decision. The question of which method should be applied also depends on the kind of question the engineer needs to answer, as we discussed earlier. 5.4. Limitations and future research First of all, a sample size of 100 may be appropriate, but having a larger sample size may provide more fruitful results. Secondly, the focus of this study was to show the added value of integrating the user into the product development process on the basis of an example. Future researches should always consider the specific user group they want to address (Arning, Gaul, & Ziefle, 2010). The low preference for the attribute “speech quality” could also have been evoked by methodology. Due to its connection to price this attribute was considered to be less important as long as it did not cost the user anything. We can conclude that price should always be handled with care in a conjoint analysis. In order to prove and validate the additional value of the extension of the engineering trade-off analysis by integrating user preferences in early stages of system design, further research should investigate the effect on purchase intentions or sales figures in comparison to “conventionally” developed products.