2.6. Collect and Collate Customer Satisfaction Data Into a Report

Customer satisfaction is in the eye of the customer. To properly measure satisfaction and to be able to effectively leverage this information to enhance profits and branding, it is imperative that customer analysis be conducted in a manner that yields an accurate profile of each individual customer. This includes correctly defining, and then accurately measuring customer satisfaction. While there is no single universal method of doing so, there are certain best practices that are essential to follow when aligning a business with its customers. Once an organisation is able to gauge customer satisfaction, the opportunities for making profitable decisions are endless.

Step One: Defining Customer Satisfaction

The most common mistake companies make when assessing customer satisfaction is asking the wrong questions. Too many companies utilise generic questions such as: “How satisfied are you with Product A?” (Very Dissatisfied, Dissatisfied, Fairly Satisfied, etc.)

Or

“Please rate your level of satisfaction with Service B:” (1 2 3 4 5 6 7…)

Vague questions generally elicit vague responses, and customer satisfaction surveys are no exception. The word satisfaction can mean different things to different people, subsequently diluting the accuracy of data a company receives. Satisfaction with a hotel catering service is very different from satisfaction with a computer purchase.

Furthermore, individual products and services have many dimensions, and thus satisfaction should not be limited to one generalised construct. For example, the quality of food may have been excellent, but the portion quantities and delivery may have been under par. Companies cannot expect respondents to divide these dimensions in their heads, weigh them on importance, and then provide an accurate aggregate rating. The hotel catering service may receive an overall rating of “75% satisfied” but that yields little or no actionable insight. Management can do little with this information to improve the business.

By defining these dimensions, however, the final data becomes much more meaningful. The “75% satisfied” now reads:

• 95% believed the food was excellent
• 60% thought the price was unreasonably high
• 79% responded that the portion sizes were too small

This is actionable insight—a manager can now make knowledgeable decisions, for example, to find more affordable food wholesalers or to make price adjustments. A customer satisfaction survey doesn’t tell you what decisions to make, but a quality survey can give you valuable information to make quality business decisions.

Best Practices:

• Work backwards. Determine what data will help you make decisions first, then create the questions that will accurately yield that data.
• Be specific. Try not to use the word satisfaction if you can help it; determine what dimensions of satisfaction are meaningful to your organisation and customers.
• Be complete. If there are multiple dimensions, measure them separately. You can always aggregate data later.

Step Two: Measuring Customer Satisfaction

Once you have defined the dimensions of satisfaction to measure, special consideration should be taken to properly measure them to yield meaningful data for analysis. Determining the appropriate scale is very important to collect accurate and significant data. For instance, when using a numeric scale (e.g., 1 to 5, 1 to 10), the range needs to be defined clearly and should be utilised consistently throughout the survey. A 10-point range, for example, may not be appropriate when asking about one’s level of satisfaction. Most respondents will not be able to distinguish the difference between a score of 3 or 4; as a result, data quality can suffer. Generally, a 5-point numerical scale or 5- or 7-point Likert scale are accepted as effective measurement techniques for satisfaction questions.

Another important consideration when measuring satisfaction is to only ask questions that are relevant. Not only does it keep your survey length to a minimum (nobody likes long surveys), but it also prevents the risk of mis-measurement. For instance, a customer who purchased copper piping from your website should not be asked for feedback about woodworking tools. In fact, the survey respondent should never even see the question (even with a “not applicable” response choice). The question holds the risk of being randomly answered for the sake of quickly completing the survey. Having a system to display only appropriate questions has the double benefit of facilitating accurate responses and reducing survey length (again, nobody likes long surveys).

Best Practices:

• Use the appropriate scale for each question type. Satisfaction scales work best when limited to 5-point numerical ranges or 5- to 7-point Likert scales.
• Ask only relevant questions. Use a logical routing, or pathing, method to only display questions that are pertinent to the respondent in question.
• Keep it short and simple. Shorter surveys reduce survey fatigue and increase the accuracy of data by allowing respondents to thoughtfully answer the questions. Typically, respondents should be able to complete an entire survey within 8 to 12 minutes.