1、 外文翻译 原文 Customer Satisfaction and Future Financial Performance Discussion of Are Nonfinancial Measures Leading Indicators of Financial Performance? An Analysis of Customer Satisfaction Material Source:Journal Of Accounting Research Author:Richard A. Lambert 1. Introduction Inner and Larkers paper i
2、s one of the first to empirically examine the relation between customer satisfaction measures and economic variables like customer retention, future sales, and stock price. The theory is that more-satisfied customers will return (with their friends) and buy again in the future. Under generally accep
3、ted accounting principles, neither these future revenues nor their capitalized values appear on firms financial statements. Moreover, the current and past expenditures made to generate customer satisfaction are treated as expenses, not as assets. Customer satisfaction measures therefore are hypothes
4、ized to be indicators of future revenue, and perhaps future profits. It is hard to find fault with this argument. In fact, the paper does not provide any theories or arguments that suggest anything otherwise. The apparent virtues of customer satisfaction measures are well recognized by companies, an
5、d many companies cite customer satisfaction as a primary objective in their mission statements. They expend resources to collect customer satisfaction data for their own internal purposes and increasingly are tying executive compensation to customer satisfaction measures. By packaging three studies
6、into one paper, the authors are able to address a wider range of hypotheses relating to customer satisfaction than they could with any single study. The first study examines the relation between customer satisfaction and customer retention and future revenues at the individual consumer level for a t
7、elecommunications company. The second study looks at customer satisfactions levels aggregated at the branch level for a bank. This enables the authors to examine both spillover effects to other customers, and a possible relation between customer satisfaction and future profitability and expenses, no
8、t just future revenues. The third study reports association tests (whether customer satisfaction measures are associated with stock prices, after controlling for the recorded assets and liabilities) and an event study of customer satisfaction disclosures. However, the differences between the three s
9、tudies make it difficult to integrate their results. The authors do not spend much time comparing or contrasting their results, a criticism which applies to most of the papers in this area. To the extent that results differ, it is difficult to know whether it is because customer satisfaction is meas
10、ured differently, whether “measurement error“ of customer satisfaction at the individual consumer level gets washed out at higher levels of aggregation (e.g., the branch or firm level), or because customer satisfaction is not equally important in all industries (e.g., telecommunications companies an
11、d banks). My discussion addresses each of the issues above, as well as other research design issues. In the next section, I discuss the measurement of customer satisfaction. Sections 3 and 4 discuss the functional form relating customer satisfaction and performance; specifically, the linearity of th
12、e relation and estimation using a levels versus differences specification. I then address other factors that will affect the strength of the relation between customer satisfaction and financial performance. In section 6, I discuss the stock price tests. The last two sections discuss implications for
13、 standard setters and for future research. 2. Measurement of Customer Satisfaction While aggregation across consumers may alleviate some measurement error problems, it can introduce another source of measurement error. In particular, customer satisfaction is typically measured for only a tiny percen
14、tage of the customers, whereas the performance measures are for the aggregated unit as a whole (i.e., all the customers in a branch or the firm).For example, in the second study in the paper, the satisfaction of the customers of a bank branch is calculated based on a survey of only 25 of its custome
15、rs. The customer satisfaction measure is therefore only an estimate of the satisfaction of the customers as a whole. A number of questions were also raised about the sampling procedures used to select customers and about the representativeness of the customers surveyed. For example, in the ASCI data
16、 used in the third study in the paper, the customer satisfaction scores are based only on customers who are willing to answer a telephone survey, only on individual-and not business-consumers, and only on U.S. not foreign customers. The paper does not convey much information about the specifics of t
17、he sampling procedures used by the firms, but it seems likely they do not match the sophistication of the sampling techniques auditors use to test account balances. Customer satisfaction is not measured in a “standardized“ way by companies themselves, across different research papers, or even across
18、 the three stud ies included in this paper. For the telecommunications firm, customer satisfaction is measured based on responses to three questions rated on a ten-point scale. The retail bank measures customer satisfaction based on a weighted aggregation of customers responses to 20 questions on a
19、seven-point scale. The third study uses a satisfaction measure developed by the American Customer Satisfaction Index (ACSI), which is based on 15 questions rated on a ten-point scale. Each study also uses a different method to aggregate the multiple responses into an overall customer satisfaction me
20、asure. The bank calculates, for each question, the percentage of customers who respond with a 6 or a 7 (the two top boxes) and then weights these percentages into an overall score. The ASCI uses a much more sophisticated technique involving latent variable analysis. It is not clear how (or even whet
21、her) the telecommunications firm aggregates its three measures of customer satisfaction. The authors have access to the three individual measures and are therefore able to combine these measures in a number of different ways.2 Some of the techniques are relatively simple ones that are commonly used
22、by companies, and some employ state-of-the-art econometrics (e.g., latent variable analysis with nonlinear estimation). Interestingly, the papers results here are relatively insensitive to the particular way the overall satisfaction score is derived from the individual satisfaction measures. 3. Func
23、tional Form: Linearity The paper spends a lot of time analyzing and discussing whether the relation between customer satisfaction and financial numbers is linear. At the individual customer level, there are a number of reasons to expect this relation to be nonlinear. For example, if there are nonzer
24、o switching costs, a customer might not decrease his purchases if his customer satisfaction falls slightly. However, larger decreases in customer satisfaction may cause the customer to take all his business elsewhere. Below a certain point, further reductions in customer satisfaction have no effect
25、on future revenues from that customer, because the customer is not going to buy from the firm anyway. Above a certain level, there may be diminishing marginal returns to increasing customer satisfaction. The authors are especially interested in investigating this. However, it is not clear that linea
26、rity can be addressed given the limited informational properties of the customer satisfaction numbers. Our current measurement techniques capture customer satisfaction on an ordinal, not a cardinal, scale. That is, we (think we can) measure customer satisfaction so that a score of 80 represents grea
27、ter satisfaction than a score of 60, but we are unable to say whether an increase from 60 to 80 is the same as an increase from 80 to 100. Therefore, we can make a monotonic transformation of the satisfaction numbers and show “linear“ or “convex“ or “concave“ functions. Of course, if the paper finds
28、 that the relation between customer satisfaction and, for example, future revenues decreases (as opposed to increases more slowly) when customer satisfaction increases from 80 to 100, this cannot be explained by the scaling of the customer satisfaction measure. Along these same lines, the real relat
29、ionship of interest is not between customer satisfaction and the future benefits but between the cost to generate customer satisfaction and the benefits. While consulting and corporate hyperbole suggests that 100% customer satisfaction is or should be the goal of organizations, it is difficult to co
30、nstruct an economic argument to support this. After all, if there were no diminishing returns to increasing customer satisfaction, everyone would go to this corner solution. For example, firms could undoubtedly increase their customers satisfaction by providing each customer with a free on-site cons
31、ultant/ operator/repairperson, etc. Obviously, these types of activities are too costly relative to their benefits. An interesting area for future research is to examine the drivers of customer satisfaction and the costs associated with building and maintaining customer satisfaction. 5. Other Factor
32、s That Affect the Customer Satisfaction Performance Relation Many factors affect a customers purchasing decision besides his level of satisfaction. Some of these factors are consumer-specific, such as the consumers wealth, the importance of the product or service in his utility function (if the cons
33、umer is an individual), or production function and investment opportunity set (if the consumer is another firm).Purchasing behavior is also affected by characteristics of the economic environment, such as level of competition from other firms, the availability of substitute products made by competit
34、ors, and the cost of these other products. The availability of substitute products suggests that purchasing behavior will have a significant “relative satisfaction“ component. That is, purchases are not merely a function of the absolute level of satisfaction with that product, but the satisfaction r
35、elative to competitors products. As a result, increasing your customer satisfaction may not win you any new customers if your satisfaction index does not increase as fast as the indexes for your competitors. Another important aspect of the environment is “switching costs,“ which are a function of co
36、ntractual obligations the consumer has with its current supplier, the cost of obtaining information about the customer satisfaction and quality of products from other suppliers, and any upfront costs associated with beginning a relationship with a new supplier. Many of the environmental factors coul
37、d probably be treated as cross-sectional constants in an industry-specific study. However, these factors are likely to be important in intra-industry studies (like the third study in the paper) or in comparing across industry-specific studies. The paper provides some preliminary evidence on inter-in
38、dustry comparisons in the third study. Unfortunately (as the paper notes), the small sample sizes preclude doing much more at this time. This is likely to be a fruitful area for future research. 4. Incorporating Customer Satisfaction into the Financial Statements The authors are careful to avoid pol
39、icy recommendations about how customer satisfaction could or should be included in financial statements. Of their three studies, the last one (cross-sectional regression of stock market values on ASCI scores) is perhaps the most relevant to policymakers. This study uses a standardized customer satis
40、faction measure, and it does not rely on self-reported scores by firms. However, as discussed above, it is limited to large, stable firms, and there are questions about the types of customers surveyed. Nevertheless, it is worthwhile thinking ahead to the recognition/disclosure issue. If we assume re
41、cognition of a “customer satisfaction“ asset on the balance sheet, the next step is choosing a measurement basis (e.g., cost or fair value).Under the cost basis (which is the alternative more consistent with the conceptual framework), we would record the cost of acquiring this customer satisfaction
42、asset. Unfortunately, there are almost no transactions which are traceable solely to the acquisition of customer satisfaction. Indeed, part of the cost of every phase of operations (including research and development, product design, manufacturing, marketing, and customer support) affects customer s
43、atisfaction. The portion of these costs which has future benefits would be capitalized and matched to the future revenues it helps generate. This would obviously be difficult to implement. On the other hand, we are a long way from being able to assess the fair value of customer satisfaction, which m
44、ight depend on the industry and its level of competition as well as other factors that are hard to identify and measure. Moreover, fair valuing customer satisfaction requires an ability to separate it from related assets such as brand names or the company reputation as a whole. The advantage of supp
45、lemental disclosures is that they need not be measured in financial terms, so perhaps we can leave customer satisfaction as a supplemental nonfinancial measure until we figure out how o convert it to a (recognized) financial one. However, even supplemental disclosures should be comparable. A disclos
46、ure standard would have to specify how much flexibility companies would be allowed in the way they measure and present customer satisfaction data. A standard would also have to specify whether customer satisfaction is to be presented for he firm as a whole or separately for business or geographic se
47、gments. The role of auditors in attesting to the customer sampling procedure and measurement methods is also an unresolved issue. Finally, any decision regarding disclosure or recognition should take account of the costs, including proprietary costs (e.g., providing valuable information to competito
48、rs). 5. Summary Inner and Larkers paper makes an important contribution by being among the first to provide evidence on whether customer satisfaction measures are leading indicators of financial performance. While this issue is by no means resolved, I am confident that customer satisfaction, under t
49、he “right circumstances“ and if measured “properly,“ is a leading indicator of financial performance. While it may be tautological to phrase it in this way, it points the challenges for future research toward (1) measuring customer satisfaction; (2) understanding other factors, both consumer-specific and institutional (like competition from other firms products), that affect the strength of the relation; (3) understanding the functional form of the relation, including how this differs at the individual consumer level versus more aggregated levels. The Inner and Larker paper does
