关键词：顾客满意  顾客满意度晴雨表

1、引言

瑞典顾客满意度晴雨表（Fornell,1992）作为全国乃至国际性的顾客满意度指数的基础，已在美国（Fornell,Johnson,Anderson,Cha,&Bryant,1996）、挪威(Andreassen&Lindestad,1998)实施。而且指数也已在新西兰、奥地利、韩国和欧洲联盟试行。这些指数是否能发展成为全球指标，更重要的它会以何种形式发展，这都是有待研究的。这种模型中用以衡量顾客满意度与关系构建中至关重要的指数有效性和可靠性在不断学习，而且更贴近实际，并且被反复改进。

作为对顾客满意度测评的方式，它也符合Poiesz Von Grumbkow (1998)为了解经济“繁荣”而对于公共事业所作的总体框架。这一框架作为一件经济福祉使每个人组成一个整体的生活品质。其他领域包括健康评估、社会文化背景、政治自由和安定。经济福祉本身是有三个方面组成，工作满足感、收入评价、消费者或客户满意。在整体层面上，Poiesz Von Grumbkow等同于顾客福利满意。这种福利型或满意累积正是国家顾客满意度指数建立的基础。

2、满意度指数模型的演变

瑞典顾客满意度晴雨表（SCSB）1989年创建以来，是首次真正的为国内购买、消费产品和服务测评的全国性顾客满意度指数（Fornell,1992）。它历史性的涉及了32个瑞典最大的产业中大约130家公司。美国顾客满意度指数（ACSI）建立于1994年秋季，它的测评结果涉及34个产业中的近200家公司(Fornell,1996)挪威顾客满意度晴雨表(Andreassen&Lervik,1999;Andreassen&Lindestad,1998)在1996年实施，截至1999年，报告结果涉及12个不同行业的42家公司（企业对消费者、企业对企业）。近些年，在开发中试用的欧洲顾客满意度指数（ECSI）横跨四大行业，欧洲联盟的11个国家（Eklof,2000）。

再看全国指数，我们特别注意ACSI模型的规格，这种模型是原瑞典模型的演变，它适用于范围较小的新西兰和台湾（Fornell,1996）、奥地利（Hackl、Scharitzer & Zuba,1996）是被用来作为在挪威和欧盟测试指标的基础。因此，模型演变的关键是制定最佳模型规格。

5、总结和讨论

一些国内和国际顾客满意度指数晴雨表在过去十年中已经实施。在大多数情况下，这些顾客满意度指标被嵌入在因果联系或满意度模型的系统中。这种模型中用以衡量顾客满意度与关系构建中至关重要的指数的有效性和可靠性在反复的研究和改进。就目前的研究结果和研究趋势，我们提出了一些修改和测试建议，我们通过NCSB调查出的数据发现大众普遍支持拟议的改动。

我们总结和讨论对于每一个拟议改变的结果。变化之一是增加了多重比较基准价格而孤立了知觉物价指数。这种模型成功分离了顾客的知觉价格，成功取代了模型中的“价值”概念，消除了模型中指标的重叠，例如：ACSI、ECSI模型。我们还认为，物价所产生对其中程度的直接影响可能会超过通过满意而产生的间接影响。这是因为满意度作为一种态度型构造，可能只是作用于质量和价格忠诚的局部的影响。满意的价格的直接影响在四个五大行业中是积极、有效的，公共交通例外。忠诚的直接效果在五个行业，航空公司和银行（系数0.096和0.098）是显著的。在某些行业，这些结果和预言，当从忠诚和满意度评价中跳出时，顾客重新发现价格的重要性。价格忠诚在两个价格竞争力强的行业——航空公司和银行的直接影响力是巨大的，这一点是不足为奇的。

建立在最初的NCSB模型之上，我们的模型还包括两个关系模式承诺变数。我们相信公司和客户之间建立起积极情感关系已经有一段时间。在四个五大行业中，双方正在积极建构令人满意的影响。至于对物价的满意效果，唯一的例外是公共汽车运输业。情感承诺（从加油站的0.493至银行的0.652）较之计算承诺（从航空公司的0.1550至火车运输的0.272）对满意度有较大影响。这并不令人惊讶。满意度应该作为增强客户关系及形成客户信任的主要贡献者（Hart&Johnson,1999）。相比之下，顾客的满意度应该影响经济学，顾客可以在经济上住在特定服务提供商，即使是满意度偏低。（Jones&Sasser,1995）

另一个重大变化是，我们以顾客对企业形象与结果感到满意为首因代替客户期望。这个改变是基于全国横断面性质的指数数据。若顾客的消费经历（满足）、他们所公司的看法等都能影响公司的形象。成功的模型是孤立于企业形象建设以及构建企业意料之中的行为。顾客对公司的满意度在每个行业中都持续不断大大影响着公司的形象（从火车运输的0.433至银行的0.575）。这反映了顾客的消费经历、顾客的经验、企业的形象都在顾客对企业的满意度的测评中起着重要作用。在五大行业中，顾客的忠诚度对企业形象的影响是小的但是却显而易见（从燃气站的0.160至航空公司的0.256）。我们相信这将持续影响对于那些最终考虑购买（即审议确定）的企业客户了解公司品牌形象。

顾客的满意度从最低点的火车运输的0.130至最高点的银行的0.289来看，忠诚度的形成对各行业是积极的和显著的。这直接影响对那些公司形象或承诺建构的公司捕捉顾客的满意度的效果表示满意。考虑到我们已经增添了忠诚的动力，那对于我们在任何一种情况下研究整体影响都是有帮助的。整体效应都直接和间接影响着顾客满意度与忠诚度，即0.551、0.627、0.471、0.557、0.458分别为航空公司、银行、公共汽车、火车和加油站。正如我们所想的那样，忠诚顾客总满意效果最大的产业是在那里挪威顾客有更多的选择，最值得注意的是银行、加油站和航空公司。

SERVQUAL的建构和投诉处理这两个范畴的示范演示并没有如我们预期的那样。在ACSI模型，是仿照投诉行为的后果满意。因为投诉处理日益成为提高顾客满意度的重要手段，我们把顾客满意度和顾客忠诚度作为衡量处理客户投诉满意程度的指标。虽然我们成功的把处理顾客投诉体系孤立了起来，但这样对顾客满意度和顾客忠诚度影响不大。最可能的解释是，投诉管理体系的产业并非特别有效或能创造顾客满意、顾客忠诚。这符合（Fornell,1996）ACSI数据分析，分析表明投诉管理系统能够压制投诉。调查结果也符合博尔顿（1999）的看法，他认为，修复服务中的不足对餐饮和旅馆的大多数顾客来说是无效的。

NCSB模型采用作为全国行业服务质量方面标准的SERVQUAL进行的构建（有形性、可靠性、灵敏性、稳定性和同理性）。相比之下，ACSI模型利用了总体素质框架，而ECSI模型运用了整体服务和区分整体服务的产品质量框架。正如有人所说，这主要是个人选择。顾客对于所要订购产品或服务的质量如何取决于细节层次的研究与一般性预期。NCSB模型利用SERVQUAL的标准将展开测评的重点放在了服务行业。然而，我们的结果显示这样的测评标准还有问题，比如有系统模式问题、有具体模型构建问题。我们还发现，大多数的路径，从五个方面对服务品质的满意度并不高。我们的建议是，要么运用国内整体产品型号和/或服务质量建构（如用在ACSI、ECSI的型号），要么增加工业公司或者特定公司的满意（Johnson & Gustafsson,2000）。

但是，从总体上看，我们的成绩是很可观的。如同预计的那样，纯粹的价格构建充当顾客满意度与顾客忠诚度的桥梁会影响顾客对企业形象满意的累计。NCSB新模型在解释顾客满意度和顾客忠诚度上较其他国内指数模型更加显著，并且能作为未来全国性指数模型的基础来服务大众。我们的研究有一个潜在的局限，那就是它是基于一个小型的经济数据。但是。作为一个拥有非常开放竞争的市场的挪威，被认为是很好的测试实验模型的国家。展望未来，这对考验新模型能否在更广泛的工业产业和国家实施来说是很重要的。

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A number of both national and international customer satisfaction barometers and indices have been introduced in the last decade. For the most part, these satisfaction indices are embedded within a system of cause and effect relationships or satisfaction model. Yet there has been little in the way of model development. Of critical importance to the validity and reliability of such indices is that the models and methods used to measure customer satisfaction and related constructs continue to learn, adapt and improve over time. The primary goal of this research is to propose and test a number of modifications and improvements to the national index models. Using survey data from the Norwegian Customer Satisfaction Barometer (NCSB), we find general support for the proposed modifications.

Keywords: Customer satisfaction   National barometers

Customer satisfaction has taken on national and international significance with the development of national satisfaction barometers and indices in Sweden (Fornell, 1992), the US (Fornell, Johnson, Anderson, Cha, & Bryant, 1996) and Norway (Andreessen & Lindestad, 1998). Indices have also been pilot tested in New Zealand, Austria, Korea and the European Union. It remains to be seen whether these indices will develop on a global level and, importantly, in what form. Of critical importance to the validity and reliability of such indices is that the models and methods used to measure customer satisfaction and related constructs continue to learn, adapt, and improve over time.

Viewing satisfaction as a form of consumption utility is also consistent with Poiesz and von Grumbkow's (1988) general framework for understanding economic “well being”. This framework views economic well being as one component of an individual's overall quality of life. Other domains include evaluations of health, socio-cultural context, political freedom and stability. Economic well-being is itself composed of three sub-components, job satisfaction, income evaluation, and consumer or customer satisfaction. At an aggregate level, Poiesz and von Grumbkow equate this customer satisfaction with customer welfare. It is this welfare-based or cumulative view of satisfaction upon which the prominent national satisfaction indices are built.

Established in 1989, the Swedish Customer Satisfaction Barometer (SCSB) was the first truly national customer satisfaction index for domestically purchased and consumed products and services (Fornell, 1992). It has historically included approximately 130 companies from 32 of Sweden's largest industries. The American Customer Satisfaction Index (ACSI) was introduced in the fall of 1994 and reports results for approximately 200 companies from 34 industries (Fornell, 1996). The Norwegian Customer Satisfaction Barometer (Andreassen & Lervik, 1999; Andreassen & Lindestad, 1998) was introduced in 1996 and, as of 1999, reports results for 42 companies in 12 different industries (both business-to-consumer and business-to-business). The most recent development among indices is a pilot test of the European Customer Satisfaction Index (ECSI) across four industries and 11 countries in the European Union (Eklf, 2000).

In reviewing the national indices, we pay particular attention to the ACSI model specification. This model is an evolution of the original Swedish model, has been adopted on a smaller scale in New Zealand and Taiwan (Fornell, 1996) and Austria (Hackl, Scharitzer, Zuba, 1996), and is the basis for the models being used in Norway and the EU. A critical evaluation of the model is, therefore, important to develop the best possible model specification.

A number of both national and international customer satisfaction barometers and indices have been introduced in the last decade, most of which are embedded within a system of cause and effect relationships (satisfaction models). Of critical importance to the validity and reliability of such indices is that the models and methods used to measure customer satisfaction and related constructs continue to learn, adapt, and improve over time. Building on recent findings and current research trends, we propose and test a number of modifications and improvements to the national index models that are now part of the Norwegian Customer Satisfaction Barometer (NCSB) model. We find general support for the proposed modifications using data from the NCSB survey.

We summarize and discuss our findings with respect to each of the proposed changes. One change was to add multiple benchmark comparisons for price to isolate a perceived price index. The model successfully isolates perceived price, and by removing “value” from the model and replacing it with price, we remove the overlap that exists between value and quality in, for example, the ACSI and ECSI models. We also argued that price may have a direct effect on loyalty over and above its indirect effect via satisfaction. This is because satisfaction, as an attitude-type construct, may only partially mediate the effect of quality and price on loyalty. The direct effect of price attractiveness on satisfaction was positive and significant in four of five industries, bus transportation being the exception. The path coefficients range from 0.13 for airlines to 0.30 for banks. The direct effect of price on loyalty is significant in two of the five industries, airlines and banks (path coefficients of 0.096 and 0.098, respectively). These results are consistent with the prediction that, in some industries, customers reweigh the importance of price when moving from satisfaction to loyalty evaluations. It is not surprising that the direct effect of price on loyalty is greatest in two price-competitive industries, airlines and banks.

Building upon the original NCSB model, our proposed model also includes two relationship commitment variables. Affective commitment captures more of the positive (or negative) relationship and trust that has built up between company and customer over time. Calculative commitment captures more of the economic consequences or costs associated with switching product or service providers. Both constructs are positively affected by satisfaction in four of five industries. As for the effect of price on satisfaction, the exception is the bus transportation industry. Satisfaction has a larger effect on affective commitment (ranging from 0.493 for gas stations to 0.652 for banks) than on calculative commitment (ranging from 0.155 for airlines to 0.272 for train transportation). This is not surprising. Satisfaction should be a major contributor to the strength of relationship and resulting customer trust (Hart & Johnson, 1999). In contrast, while satisfaction should influence the economics of switching, customers may be held economically hostage to particular service providers or locations even when satisfaction is low (Jones & Sasser, 1995).

One of the most important findings is the large positive effect that affective commitment has on loyalty. The effect is significant in four of five categories, bus transport again being the exception. In these four industries, affective commitment has a larger effect on loyalty than does satisfaction directly. This suggests that satisfaction affects loyalty largely through its ability to build strong relationships between companies and customers. Adding the commitment variables has the benefit of greatly increasing the model's ability to explain variation in loyalty vis-à-vis the other national index models.

Another major change is that we replace customer expectations, as an antecedent to satisfaction, with corporate image as a consequence of satisfaction. Recall that this change is based on the cross-sectional nature of the national index data, where a customer's consumption experiences (satisfaction) should have some influence on their perceptions of corporate image. The model is successful at isolating the corporate image construct, and the construct behaves as expected. Satisfaction has a consistently large effect on corporate image in each industry (ranging from 0.433 for trains to 0.575 for banks). This reflects the contribution that consumption experiences have on corporate image as well as the consistency between a customer's experiences and corporate image over time. The effect of corporate image on loyalty is smaller but significant in each of the five industries (ranging from 0.160 for gas stations to 0.256 for airlines). We believe that this captures the ongoing inclusion of brands or companies with strong corporate images among those that customers ultimately consider for purchase (the consideration set).

The direct effect of satisfaction on loyalty, which ranges from a low of 0.130 for trains to a high of 0.289 for banks, is also positive and significant for each industry. This direct effect captures the effects of satisfaction on loyalty that are not mediated by the corporate image or commitment constructs. Given that we have added more drivers of loyalty, it is useful to examine the total effect that satisfaction has on loyalty in each case. The total effect is the sum of all direct and indirect effects linking satisfaction and loyalty, which equals 0.551, 0.627, 0.471, 0.557 and 0.458 respectively, for airlines, banks, buses, gas stations and trains. As one would expect, the total effect of satisfaction on loyalty is greatest in those industries where Norwegian customers have greater choice among competitors, most notably banks, gas stations and airlines.

Complaint handling and the SERVQUAL constructs were two areas where the model did not perform as well as expected. In the ACSI model, complaint behavior is modeled as a consequence of satisfaction. Because complaint handing is an increasingly important means of improving satisfaction, we used the quality of complaint handling among complaining customers as a driver of both satisfaction and loyalty. Although we successfully isolate a complaint-handling construct, it has little effect on either satisfaction or loyalty. The most likely explanation is that complaint management systems in the industries are not particularly effective at creating satisfaction or loyalty. This is consistent with Fornell (1996) analysis of ACSI data, which suggests that complaint management systems are only capable of neutralizing complaints. The finding is also consistent with Bolton (1999), who finds that service recovery is generally ineffective for a majority of customers in both a restaurant and hotel setting.

The NCSB model uses a variation on the SERVQUAL constructs (tangibles, reliability, responsiveness, assurance and empathy) as service quality dimensions across industries. In contrast, the ACSI uses an overall quality construct, while the ECSI distinguishes between overall service and overall product quality. As argued earlier, this is largely a matter of choice. How one specifies product or service quality depends on the level of detail versus generality desired in the research. Using the SERVQUAL dimensions is a natural place to start given that the NCSB focuses on service industries. However, our results reveal systematic problems with this part of the model, specifically with the tangibles construct and, to a lesser degree, the assurance construct. We also find that the majority of the paths from the five service quality dimensions to satisfaction are not significant. Our recommendation is that the national models either employ the overall product and/or service quality constructs (as used in the ACSI and ECSI models),or build more industry or firm-specific drivers of satisfaction (following Johnson & Gustafson, 2000).

Overall, however, our results are quite promising. The pure price construct functioned as anticipated with respect to both satisfaction and loyalty. Cumulative satisfaction was found to update corporate image, which in turn impacts customer loyalty. Cumulative satisfaction is also an antecedent to relational commitment, which in turn has a relatively large impact on customer loyalty. As a result, the new NCSB model explains significantly more variance in loyalty than other national index models and can serve as a basis for future national index models. One potential limitation of our study is that it was based on data from a small economy. However, Norway is known to have a very open and competitive economy making it a good context to test the proposed model. But going forward, it will be important to test the new model in a wider range of both industries and countries.