Q: Why should we measure both satisfaction and loyalty?
Despite this reality, some organizations lump satisfaction and loyalty together by using an overall loyalty index. Unfortunately, this practice obscures the cause and effect relationship between satisfaction and loyalty. This narrow focus on loyalty makes management of customer relationships more difficult and can negatively impact financial performance.

To avoid this happening, CFI strongly recommends measuring satisfaction and loyalty as separate, but related, constructs. Managed appropriately, loyalty is an outcome of (1) customer satisfaction, (2) product/service differentiation, and (3) price variables that are optimized to produce the best results. This approach is fundamental to the relationship between satisfaction as the driver and loyalty as an outcome that influences financial performance.

Price discounting is one example of a narrow loyalty focus and its potential negative impact on financial performance. Relying on price discounts to drive loyalty often leads to ineffective pricing strategies. When unnecessary price discounts are given, the resulting loyalty benefits are often more than offset by the revenue and profit loss from these discounts.

In recent years, the automobile companies relied heavily on price discounts to secure customer loyalty. The discounts certainly kept buyers in the show rooms and propped up market share, but didn't do much to create satisfied customers or improve long-term financial performance.

Satisfy customers first and loyalty will follow.

Q: Why is measurement scale so important?
The debate among advocates of 10-point, 7-point and 5-point scales goes on. Of these, the five-point scale is still perhaps the most widely used today in customer satisfaction research. CFI Group, by contrast, recommends using a ten-point scale with anchors only at the endpoints. But what are the advantages of 10-point versus 5-point scale?

Advocacy for the 5-point scale rests on the assumption that respondents can't discriminate beyond 5 levels. There is actually a preponderance of research to the contrary . Not only is the 10-point scale within the capabilities of respondents, it also provides a number of key advantages.

The ability to discriminate between responses is important because customer satisfaction data is generally positively skewed. In customer satisfaction research, a 5-point scale is really closer to a 3-point scale since respondents typically don't use 1 and 2. With the vast majority of the responses being 3, 4 and 5, there isn't adequate variation in the data for sophisticated statistical analysis. Furthermore, within a 10-point scale respondents are free to choose a number that best represents their true feeling without struggling to decide whether they "agree" or "strongly agree" with a particular statement about a company's products or services. In addition, a ten-point scale inherently provides more variation in responses, which in turn produces more precise modeling results.

So what does all this mean? Using a 10-point scale gives managers:

• Greater precision at lower sample sizes, which means reduced survey costs and less imposition on customers.
• Greater measurement precision, which is crucial to linking to internal performance measures, identifying the drivers of satisfaction and predicting the financial returns associated with their improvement.

• Mean-based analysis is more precise than percentages at any given sample size. Let's compare a CFI Customer Satisfaction Index score of 71 to a "% Satisfied" score of 71%. The CFI Customer Satisfaction Index score would have a 90% confidence interval of +/- 2.7 points for a sample size of 150. In contrast, the "% Satisfied" measure would have a confidence interval of +/- 6.3 points. In other words, the percentage measure is nearly three times less precise than the mean-based measure, resulting in either significantly larger sample size requirements (and greater costs) or compromised decision-making.
• With "% Satisfied" scores, the 1s, 2s and 3s are grouped together, as are the 4s and 5s. Yet, intuitively we know that there is a difference in satisfaction between a response of 1 and a response of 3. When you start collapsing results into "boxes," you are throwing away valuable insight.