How to Design a Survey

Part 2

The way in which surveys ask questions is as important as what the survey asks. Thoughtful questions engage respondents while confusing or irrelevant questions can damage a restaurant’s reputation. Preferred question formats even vary within the industry based on the type of restaurant and typical customer. Interactive questions can increase response rates, but these are best used for more upscale restaurants. Simple multiple-choice surveys, however, can work for fast food operations or almost any kind of restaurant.

Creating a survey can be challenging, but restaurateurs can find templates and companies that design professional surveys online. MustHaveMenus offers various templates for guest-satisfaction cards that restaurants can use in-house.

Pros and Cons of Making Surveys Anonymous

People assume that anonymous surveys result in more honest answers. While usually true, some respondents will use anonymity as an excuse to provide funny, irreverent or crazy answers. Restaurants can’t fix problems for customers unless their identities are known.

Of course, Web browsers allow consumers to use various screen names. Restaurants can use various techniques to guarantee anonymity while still being able to communicate with respondents. One method is by updating reports on a delayed basis over for a time period so that data from exact times can’t be used to identify survey takers. Another method is to publish surveys on an anonymous, authenticated website.

  • Surveys can target certain demographic groups to straddle the line between personalization and anonymity.
  • Respondents can choose if they want to personalize their responses or remain anonymous when surveys offer this option.
  • Restaurants can use the responses from an anonymity option to design future surveys.
  • Totally anonymous surveys could generate other problems with the results, so offering a choice provides a greater range of actionable data.

Features for Effective Restaurant Surveys

Qualifying questions are useful for validating survey results and eliminating unnecessary questions. For example, surveys can ask if respondents have ever eaten in the restaurant or target people who only dine at restaurants on weekends. Surveys can also be filtered based on respondents’ answers to qualifying yes-or-no questions. Types of questions that provide useful survey results include:

  • Yes and No Questions
    These questions are great for simple, general surveys where straight yes-or-no answers are needed.
  • Multiple Choice
    Multiple-choice questions can include choosing one answer, all options that apply, a preset number of choices or a ranking in order of preference.
  • Rating Scales
    Useful for determining how strongly diners feel about a subject, rating-scale questions follow a format from one extreme to the other like very pleased, somewhat pleased, neither pleased nor displeased, somewhat displeased or very displeased.
  • Stapel Scale
    These questions ask respondents to rate a menu item or restaurant service on a positive or negative scale—for example, -5 to +5.
  • Open-ended Questions or Comments
    Surveys can ask open-ended questions or allow respondents to check “Other” to open a dialogue where they can explain or qualify their answers or explain negative responses.
  • Demographic Questions
    Surveys can use these questions to sort survey responses based on various criteria such as income, sex, ethnicity, activities, club affiliations or interests.

Too Much Information

Long, complex surveys increase bounce rates in direct proportion to survey length. Keep surveys short, and try to design them so that irrelevant questions are eliminated based on negative answers to qualifying questions. Called skip-logic by survey design companies, this feature makes the questionnaire more responsive.

In longer surveys, customers often quit or begin answering randomly before finishing. One strategy for validating the answers is to randomize the order in which questions are asked. Surveys can provide some accurate data across the range of questions by using this method.

Another method of encouraging respondents to complete longer questionnaires is by allowing them to save their answers and continue later. This method requires putting a cookie on the survey-taker’s browser. Restaurants can send email reminders to respondents to increase completion rates.

Beware of revealing too much internal information in the survey questions, which could generate negative responses among the survey takers. For example, company policy might involve using each customer’s name, if known, three times during a restaurant visit. Don’t spill the bacon grease by asking if someone addressed the customer by name at least three times. Approach the question generically by asking if the service was warm and personal.