Editor’s note: Jim Stanko is the customer satisfaction manager for one of Rockaway, N.J.-based Hewlett-Packard’s Test and Measurement product lines.

We’ve all seen it before: A promising initiative catches top management’s interest, only to wind up on the office floor when it fails to produce quick returns. So how about this customer satisfaction thing? Is it just another passing fad, or is it a fundamentally different approach to managing a business? Can it really make a difference in the bottom line?

The hypothesis seems sound enough: If we can identify what satisfies and dissatisfies customers, we should be able to use the information to drive internal process improvements that reduce dissatisfaction, and create more loyal customers. This alone would appear to be adequate justification for supporting a customer satisfaction approach. Actually, it’s just scratching the surface of what can be done with an effective program. In fact, the real power of customer satisfaction survey research lies in its application to strategic decisions. This article explores several opportunities that show why you can’t afford not to have a strong customer satisfaction program in your company.

But first, the basics. The starting point for most companies is a survey of their customers, to determine what they like and dislike, and what needs to be improved. A carefully designed survey can also produce information on what factors customers consider most important, and how satisfied they are with your performance on each factor. There have been many examples of four-quadrant plots of importance versus satisfaction, such as the customer value map in Figure 1. It should be a simple matter to identify where customers are not fully satisfied on important factors. Separation of the factors into each quadrant suggests which to act on. Customer comments about what they dislike, or would like to see improved, can help target needed process improvements or programs to address each of the key factors.

Yet some articles have already appeared questioning the effectiveness of this approach. Can we show return on investment (ROI) for these efforts? How far do we need to go in eliminating dissatisfaction? Hold these questions for a moment, and recall the time usage maxim: "It’s small comfort to be more efficient at doing something that you never should have been doing in the first place!" There will always be customers who will complain about any given factor, no matter how high your performance may be, and no matter how much you have invested in improving. The key, then, is to identify the factors that give you an edge on the competition.

So clearly the first improvement needed is to survey both your own customers and your competitor’s customers. This opens the door to a wealth of strategically valuable information. For a start, it shows the relative strengths and weaknesses of your company versus your competitors in each of the factors that affect the customer’s choice. To get unbiased information, of course, the study should be done blind, i.e., so that the customer doesn’t know who is conducting the survey.

The four-quadrant plot of data gathered in this manner is shown in Figure 2. It now highlights specific areas where the competition is vulnerable. This competitive satisfaction map can provide some good ideas for programs not just to build loyalty, but more importantly to build market share. Not only can you identify what to do to keep your customers, you can clearly see some areas where it might be possible to attract the competitor’s customers.

In fact, for many markets it may not be easy to obtain market share data. Competitors may be closely held private companies, or divisions of larger companies. However, a good competitive customer satisfaction survey will gather information on what brand the customer purchased last. This data allows calculation of buyer/chooser share, which can be a good surrogate for market share. True, the absolute accuracy may be lacking, and depends on having a good survey sample. But with annual customer satisfaction surveys, tracking relative share from year to year, as shown in Figure 3, is easy. This relative measure can be more useful than absolute information for determining how your brand is faring relative to competing brands, and it certainly beats having no data. Further, this can provide metrics for the programs you run; but more about this later, when we return to the issue of ROI.

New applications

The basic four-quadrant importance/satisfaction plots have been around for some time, but here are some new applications, showing how it can be used for a variety of tasks. Much has been written on the importance to a successful business of a clear and compelling value proposition. It should state the specific benefits, from the customer’s perspective, that your company offers.

Figure 4 shows a way of relabeling quadrants of the customer value map to help with this task. The upper-right-hand quadrant indicates which values are important to the customer, and which they feel they are receiving satisfactorily. Clearly to be successful you must offer these. The lower-right-hand quadrant shows important factors which the customer is not receiving satisfactorily. This highlights the opportunity to differentiate yourself from competitors in crafting a superior value proposition. Comparing this chart with the competitive satisfaction map of Figure 2 shows not only just which factors the customer considers a benefit (importance), but what you and your competitors are delivering (satisfaction). Of course, you can create these four-quadrant charts not only for the whole market, but for any appropriate subsegment of the data to highlight the difference in value proposition for different segments.

Add to this a study of competitors’ advertising, and you can derive even more value. Taken with the analysis of their ads, the competitive satisfaction map can reveal how competitors are positioning their products and services, and how effective this is in the eyes of your (and their) customers. That could reveal some interesting insights. Why might they emphasize different factors than you would? Do their customers value different things than you are providing? Perhaps their customers represent a different segment than yours, with a different value proposition necessary to attract them. Wouldn’t this knowledge be of strategic value to your business planning?

Competitive landscape

What about the decision to enter a new business; could anything be more strategically critical? Clearly you would like to understand the competitive landscape and what features your product needs to have to displace competitors, before entering a new market. Here’s a situation where you’d like to survey the competitors’ customers before you lock down the design of your new product or the appropriate business strategy.

No problem; you don’t have to have customers in a market to do a customer satisfaction study of it. Furthermore, if you ask customers what brand they plan to purchase next and why, you have all the information you need to construct the customer loyalty plot shown in Figure 5. Here you can clearly see what percentage of each competitor’s customers are vulnerable, i.e., planning to switch brands, or currently undecided. Compare the reason for defection, given by the customers who are vulnerable, with the satisfaction information from Figure 2. This can provide additional insights on the specific areas in which the competitor may be weak, from their customer’s perspective.

But that’s just for starters. The same information that we use to define the value proposition can form the basis for segmentation of the market. Remember, another way to define a segment is a group of customers that have the same value proposition. Cluster analysis can be performed on the same importance/satisfaction data which you used to develop the value proposition. With segments defined from the results, a new version of Figure 2 can be drawn showing the importance/satisfaction data just for the members of each particular segment. The critical factors which differentiate the segments will thus be clearly highlighted. Sorting the survey respondents into their respective segments allows generating crosstabs of the survey data just for the respondents in each segment. This can then be used to define many aspects of the product to meet the needs of the specific segment you decide to target.

Using customer satisfaction data for this task may seem a little surprising at first. However, a customer satisfaction study is very similar to user-needs market research. The difference is that you take a post-sale, versus a pre-sale, perspective to the questions, asking about the customer’s current product. Many of the same questions can be asked, and in fact, customers may find it easier to relate to something concrete (the product they are currently using), than an abstract future product. This brings up a nagging concern. The customer satisfaction survey clearly identifies the critical satisfaction factors, whereas market research is usually concerned with identifying critical choice factors. But what is the difference between choice factors and satisfaction factors?

If you already know that the customer’s choice of your product is heavily influenced by previous product usage, there may be no practical difference between choice factors and satisfaction factors. For many markets there may be little difference; this should be verified with the appropriate surveys of the target segment. But here’s a quick rule of thumb: The more frequently the underlying product is purchased, the greater the likelihood that there will be strong correlation between satisfaction with the current product and choice of the subsequent product. For commodity products, satisfaction clearly wraps around to choice; experience affects purchase directly. For less frequently purchased products, like a new car or home, the basis for satisfaction today may differ substantially from the basis for choice five or more years ago.

Message development

Want even more applications for this survey data? Consider using the customer satisfaction survey data to drive the message development for your advertising programs. The value proposition information is very useful to identify what customers are likely to respond to. And, of course, you have the information to know where competitors are vulnerable. . . seems like it should be relatively easy to identify appropriate advertising themes.

But before we complete this exercise, we could really use two additional pieces of information: awareness of, and preference for, a brand. This data is readily obtained from the survey. It’s easy enough to ask customers what brands they are aware of. This could be either on an aided or unaided basis, depending on the medium you use for the survey, among other factors (see the sidebar for some additional thoughts on this choice).

Now for the remaining question. We have already asked the customer what brand they bought last, and what brand they plan to buy next. Suppose in between these questions we asked the customer what brand they strongly considered, at the time they bought their last brand, but didn’t purchase. This tells us which brands the customer preferred enough to put on their "short list."

Consider a simple model of the stages of the buying process, which can be represented appropriately for many products, as follows:

AWARENESS -- PREFERENCE -- EVALUATION – DECISION

Now we can use the data from these two additional questions to plot a market leakage chart, shown in Figure 6. This indicates where in the buying process you are losing market share. If a large loss is observed in awareness, the remedy would clearly seem to be advertising. If the largest loss occurs in the evaluation process, the Competitive Satisfaction Map, Figure 2, will reveal which factors are pivotal in the decision of the customers you are losing. Further correlation with the comments from the customers who defected at this stage can help indicate what needs to be done. But the data in Figure 6 indicates that, for this market, the largest loss is experienced in the preference stage. A different type of program may be needed to improve brand preference, based on the analysis of the underlying problem. The Market Leakage Chart can help you pinpoint where you are losing share, and how to make your business more effective.

Now, let’s return to the issue raised earlier about the ROI for programs, regardless of whether they are process improvements to increase customer loyalty or more traditional promotional programs. Clearly, deployment of limited marketing funds is one of the more critical strategic decisions. If you could get a reasonable estimate of ROI for marketing programs, you could compare their effectiveness with product development programs, and decide which would grow market share more cost effectively.

Here’s how. The market leakage chart shows how much share there is to be gained for different programs. If a hypothesis is formed, based on the analysis of the problem, on how much share could be gained for a given investment, an expected ROI can readily be calculated. The market leakage chart not only guides the formation of the hypothesis, it can be used to measure the results! If a survey is conducted before and after the program is implemented, the actual ROI can be calculated from the comparison of the "before" and "after" charts as shown in Figures 6 and 7.

In this case, a program designed to increase preference for this HP product line achieved five points of market share, with 80 percent of the improvement resulting from increased preference.

Hopefully, this discussion not only demonstrates some new ways to use customer satisfaction data, but also how it can contribute to successful strategic decision making. All together, it provides a compelling case for establishing, and maintaining, a robust customer satisfaction program.

SIDEBAR

Aided versus unaided awareness

There are many ways to ask respondents about their awareness of a brand. The information thus gathered can help with decisions on advertising investments and promotional programs. Two ways to measure awareness are on an aided and unaided basis. Aided awareness is measured most easily by presenting the respondent with a list of brands that are associated with the product or service being researched and asking them to indicate which ones they are aware of. This technique is most frequently associated with written or mail surveys.

Unaided awareness is gathered without the aid of a list. This approach is common with telephone or personal interview surveys. Typically the respondent is asked what brands they think of when considering the specific product or service. The respondent then answers with as many brands as they can think of. Often there may be some degree of prompting when the respondent pauses, such as: "Can you think of any other brands?" or "Are there any other brands you are aware of?"

As a general rule, aided awareness typically overstates true awareness, as respondents will sometimes associate a brand with a product if that association is plausible to them, even though the brand in question may not offer the specific product or service. On the other hand, unaided awareness may understate true awareness, because many times respondents may fail to remember a brand that they actually have seen or used.

The true level of awareness usually lies in between the values found from these different approaches. For the customer satisfaction surveys discussed in the accompanying article, measuring unaided awareness can cause some difficulties in constructing the market leakage chart. It’s not impossible for respondents to mention a brand later in the survey that they had not included in the list of brands they are aware of. This type of "memory recovery" during the subsequent survey questions requires correction of the answers to previous questions to avoid inaccuracy. For this reason, it is generally preferable to use aided awareness if you plan to construct a market leakage chart from the data.