Nine Steps to Create a Great Customer Survey: Part Two

	Bookmark This Page Email This Page Format for Printing Submit an Article Read More Articles
	Related Tools & Articles Discover Customer Needs With Ethnography How to Think What No One Else Thinks Using Innovation to Address Unmet Customer Needs
	Discussion Forum "My concern that TRIZ has not been widely accepted is based on results from a survey. Responses showed that they did not deny that TRIZ is a powerful and useful technique, however, they all found it not easy to apply TRIZ in their design projects." Contribute to this Discussion

By J. DeLayne Stroud

Many organizations use surveys to gather information and data in order to learn more about their customers. To prepare a successful survey, practitioners should follow nine basic steps. Part One of this series discussed the first three steps: establishing a goal, determining the sample and choosing the methodology. The fourth, fifth and sixth steps, which relate to what types of questions to ask and how they will be asked, are explored here.

Step 4: Choosing the Format

The format or media channel used for a survey has a large effect on how the questionnaire is constructed, the number of surveys sent out, the cost and the timeline. There are several formats available, and each has a number of advantages as well as some disadvantages. Some of the options available:

Face-to-face Interviews

These should last no longer than 30 minutes. Advantages include the ability to let the interviewee use, touch, and test the product or service. In addition, in-person interviewees may allow longer interviews than other media such as the phone.

Disadvantages include that these surveys usually cost more per interview than other formats and training requirements may take a long time. In addition, interviewing at a specific site, such as a shopping mall, may introduce bias into the sample. Training and implementation can take up to 18 weeks.

Telephone Surveys

These should be 10 minutes or less. Advantages include that people can be contacted quickly and the calls can be automated, allowing for logic applications and decision trees to be used. In addition, the surveys can be delivered using voice response units, which can randomly dial numbers to get sample.

The disadvantage of this type of survey is that it can be mistaken by customers as phone solicitation. Unfortunately, telemarketers have made people reluctant to answer the phone. Also, with many more people working, the time window for calls is limited to the 6 p.m. to 9 p.m. area, which sometimes includes interrupting dinner, quality family time or that extended happy hour. In addition, products or services usually cannot be demonstrated over the phone.

Mail Surveys

These should take less than 15 minutes to complete. Advantages include that this media is one of the least expensive formats to use and can be conducted without knowing phone numbers. In addition, the surveys can include pictures or diagrams and allow people to answer at their own pace. Also, they are not as intrusive.
 
Disadvantages are that mail surveys take a long time to get information back and response rates can be affected by education levels of the sample group. Low response rates can cause costs to be driven up.

Email Surveys

These should take no more time than mail surveys – typically less than 15 minutes to complete. Advantages include that email surveys mean speed. Literally thousands of responses can be collected in days with practically no costs after the message is set up. In today's high-tech world of laptops and mobile phones, more people have constant access to email.

Of course, to run this type of survey, the organizations must have email addresses for potential respondents. Other disadvantages include the possibility of people responding to the survey multiple times and messages being filtered out by spam blockers.

Web Page Surveys

These should also take about 15 minutes. Advantages include that they are extremely fast for collecting information with little to no cost after set up. They can include sound and video files, not to mention the fact that designers can utilize decision trees to skip questions based on certain answers or set up paths of questions.

Disadvantages are similar to those of email surveys in that Internet surveys have bias because not everyone uses the web and people can respond multiple times. They also may severely limit responses to simply people using that web page.

Step 5: Creating the Survey

Regardless of the sampling method or survey media chosen, if survey does not have good questions, it will not provide good results. Organizations should design the questions to fit the format. For instance, it is not possible to show pictures over the phone, and mail survey respondents can't ask, "What do you mean by that?" Make certain the questions all link back to the goal of the survey and prioritize questions into three general groups: Must know, helpful to know and nice to know. If resources are constrained, focus on the first two groups.

There are three basic categories of questions used in surveys:

1. Multiple choice questions have two common variants:

• Rating scales: Excellent, good, fair, poor, 1-10 scale
• Agreement scales: Agree, neutral, disagree

Advantages are that you can make the questions very clear and easy to enter the response data. A uniform pool of answers also enables easier analysis.

Disadvantages are that questions may get random answers if they are not clear. The best answer may not be included as a choice or a lack of other choices can frustrate respondents.

2. Open-ended questions, both numeric- and text-based, can be used to supplement multiple choice questions, but should be used sparingly. Uses include as a follow-up to another question, collecting biographical and demographical information at the end of the survey, and wrap-up, opinion type question that allow respondents to add comments. Granted, this type of question is harder to analyze and place into specific categories for graph purposes. And if there are written responses, legibility may also be an issue.

3. Sequential questioning is a technique where questions are asked in a specific order. This method can help establish a flow and encourage people to continue answering questions and complete the survey. The questions should start out easy while they lead the respondent to answer more. Difficult or personal questions can be left to near the end. Demographic information, unless key to the analysis, can also be left until the end.

Step 6: Writing Quality Questions

It is a good rule of thumb to make certain that survey questions will accurately tell the designers what they want to learn. Utilizing single-dimensional questions, and avoiding two-dimensional questions, can assist with this concept.

An example of a two-dimensional question: "Were you satisfied with the quality of our communications and training?" Although people often hear communication and training used in the same sentence, the concepts are two very different things. Using an example of a merger and acquisition, employees may have been happy with the communications they received regarding timelines and training dates, but the actual training they received may not have been appropriate for the obstacles they would soon face. Survey designers should separate the concepts into two questions: "Were you satisfied with the quality of our communications?" and "Were you satisfied with the quality of our training?"

Try to accommodate all possible answers in the choices given and have mutually exclusive choices. Do not guide people to a predetermined answer and stay free of acronyms and jargon that can bias responses.

Before Moving On

As discussed in Part One, designers must have a clear understanding of what they want to learn and from whom they want to learn it, and choose the right methodology for gaining that knowledge. With the audience and method defined, choosing the right format to reach that audience is critical – making it easy for the audience to provide what is needed. The survey questions must be clear, and it can be especially helpful if they are organized into specific groups of knowledge. Finally, understanding the elements of a good question will ensure that surveys return quality results.

About the Author:

. DeLayne Stroud is a Master Black Belt project manager with CSI DeLeeuw. He retired from Bank of America in 2005 with more than 20 years of experience as an executive in project and change management in the banking industry. He has led multiple Design for Six Sigma and Lean initiatives. His most recent client support efforts have included the New York Independent System Operator, JPMorgan Retirement Plan Service, ING and the PNC/National City Bank Merger. Contact J. DeLayne Stroud at jstroud (at) deleeuwinc.com.