Designing an Effective Customer Survey

Once you have determined and defined your target market, you may want to measure customer attitude, preference or satisfaction. The survey method is a common method of measurement. It can be administered via email, website, social media site, telephone, or in person (e.g. mall intercept interview). However, before you design a survey there are several design issues to consider.

Introduction

An important part of the survey tool is the Introduction. In the introduction, clearly state the purpose of the survey and who is being targeted. Also, don’t use the introduction to specify the type of product or service being evaluated as this may cause the respondent to prejudge the survey.

Survey Length

Determine the survey length that is most appropriate for the place it is to be administered. If you are administering the survey at a retail store checkout counter, then the survey shouldn’t exceed 5 or 6 questions. If, however, you are administering the survey in a more comfortable environment (e.g. in a respondent’s home or in a professional interviewing center) the survey could be as long as 100 questions.

The respondent shouldn’t be asked to supply to much detail, especially if a long time has elapsed since behavior of interest occurred (e.g. “In the last five years, exactly how many times have you contacted our office by phone?”).

Clear and Unambiguous Items

It is important to be explicit in your instructions for completing the survey. Explain in detail exactly how to respond to each item. Don’t ask “double-barreled” questions. Each survey item should ask one question and should measure only one variable. Ask for only one piece of information per question. For example, instead of asking “True or false: our associates are polite and knowledgeable” – break the question into two questions, one measuring perception of politeness and another measuring perception of knowledgeability. What if associates are perceived as polite but uninformed about the products sold.

Also, design the survey with your audience in mind. Keep in mind the level of education of respondents and how familiar they are with your product or service. Be sure to keep the questions short and easy to understand. The question should have one meaning and the meaning should be clear. Select words that mean the same thing to everyone.

Don’t Lead the Respondent

Don’t ask the respondent leading questions and be careful not to bias the respondent’s answers with your attitude, tone, or setting. Moreover, the response scale should not slant toward a certain answer (e.g. “How would you rate the fairness of fees charged?”)

More examples of leading questions:

  • “Do you agree that our associates are knowledgeable?”
  • “Do you agree that we have an adequate variety of products?”

Examples of less leading questions:

  • “Overall, how satisfied are you with our store?”
  • “How likely are you to renew your membership?”
  • “Don’t you agree that this is a top-quality law firm?”

Loaded questions

You want to also avoid asking loaded questions. These questions are subtler and tend to offer a reason for responding a particular way (e.g. “Do you advocate higher fees so we can stay in business?”)

Bias

Carefully design and administer the survey in a way that minimizes respondent bias. Reduce chances of bias by:

  • Designing instruments that positively represent the organization and encourages respondent feedback
  • Insisting on a random sample
  • Obtaining a breakdown of subgroups represented in the measurement sample
  • Keeping the survey as short as possible
  • Considering respondent fatigue
  • Asking non-threatening questions

Types of Questions

Two-choice: either/or (Yes, No)

“Are you currently enrolled in a Center for Adult and Graduate Studies academic program?”

  • Yes (Please proceed to Question 2)
  • No (Thank you for participating, please conclude survey)

Multiple-choice

“What resources influenced your decision to enroll in a Center for Adult and Graduate Studies academic program?” (Check all that apply.)

  • Aurora University website
  • Online search
  • Online Advertisement
  • Social Media
  • Word of Mouth
  • College fair
  • Rating Guides
  • Printed Materials from AU
  • Media Sources (e.g. Magazine, Newspaper articles)
  • Academic advisor/Faculty/Staff at your College/University
  • Other (Please specify):

Scale-rated

Example of a Likert Scale question

Likert Scale

Open-ended

These questions are used to elicit detail about a previous question:

Example of an open-ended question:

  • “What suggestions do you have for improving your experience with Aurora University?”

Demographics Section

Lastly, be sure to capture demographic data. Collect as much information as possible (e.g. Age, Gender, Income, Zip Code, Ethnic Group). These questions are usually placed at the end of the survey.

 

About the Author:

Anthony Hawkins is a graduate student in the Masters of Science of Digital Marketing and Analytics program at Aurora University. He is a digital marketing and analytics professional passionate about the natural and organic foods market.


4 thoughts on “Designing an Effective Customer Survey

  1. Great posting in designing an effective customer survey. Research has shown that there will always be a small minority of people who will lie on the survey and sometimes people will give inaccurate answers completely by accident. Do you have solutions to these problems with survey?
    Thanks!

    Liked by 1 person

    1. To minimize lies and maximize survey accuracy it is important to first know why some respondents lie on surveys. People lie on surveys for various reasons. Research has revealed that some people lie to make themselves out to be as good as or better than others. Some lie to appear modern, mainstream or politically correct. In survey research we call this social desirability bias. Some respondents become defensive or even embarrassed by certain questions and consequently lie to conceal their true feelings or beliefs. Some will tell the researcher what she wants to hear just to be polite or to avoid personal embarrassment. There are many other reasons people lie on surveys.
      Other approaches to minimizing lying include limiting the elapsed time between a referenced event and a corresponding survey question, avoiding potentially embarrassing, offensive, intrusive questions; and allowing the respondent sufficient time to respond to questions.
      Another approach is to control for Hypothetical Bias by asking respondents to indicate degree of certainty for certain questions.

      Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.