Contact Us nav toggle
Home > FAQ > Survey Design

FAQ: Survey Design

Could you tell me a little about the things I should consider during survey design?

Surveys should be customized to meet your needs, with questions that address specific concerns. Saying that, surveys should also contain a mix of independent—or predictor—variables and dependent—or outcome—variables. This facilitates identification of the items that have the most impact on your organization’s overall performance, and the items that need to be targeted to improve performance and overall levels of customer satisfaction.

So what are variables? Variables are essentially items or questions, independent variables assess specific elements of the experience, and they tend to be narrow in scope with a niche focus. These might be things like “My account executive is responsive to my requests.” Dependent variables assess the totality of the experience, and tend to be broad—big picture—items. For example, I am satisfied with the company. Having a healthy mix of these two types of questions fosters a well-designed survey.

How long should the survey be?

It is recommended that you include enough questions to obtain all of the information needed at one time. Using items that are short, concise, clearly written statements that get right to the point will only take seconds to read and rate. A 20 item survey should require about 5 minutes for someone with an 8th grade reading level.

Customer surveys should not contain more than 30 questions without offering an incentive to compensate participants for their time. Employee surveys may be longer without impacting response rates or requiring an incentive. A typical employee survey includes 40-60 questions.

Does NBRI support international surveys and foreign languages?

Yes! Our platform allows employee surveys and customer surveys and market research in almost any language anywhere in the world. Our systems support any written language.

How should a survey be constructed?

The order in which survey questions are presented may bias how subsequent questions are interpreted. Generally, the questions on an employee survey are presented in random order while the questions on a customer survey are presented in the order that services or product features are likely to have been experienced.

The wording of survey questions strongly influences the answer. Structuring your survey questions as statements has proven to be the most effective means of evaluating attitudes, opinions, and beliefs, across a wide range of variables.

Professional survey questions written by psychologists, trained in creating psychological tests (surveys) are recommended. Questions should be valid, reliable tools to be used with confidence to fully evaluate the topics selected. And, questions should be tested and standardized. Utilizing standardized questions allows you to compare your scores against other companies who have used identical questions. This comparison is known as benchmarking.

I'm using a ten point survey scale. Is ten points enough?

Ten or more survey scale points are generally too many. We need to assign a weight, or definition, to each point on the scale. To do that for a ten point scale is arduous at best. Respondents may not read all the way through the “weights” and even if they did, would it make sense? For example, do you immediately understand what “slightly more than moderately agree” means? Respondents struggling to come to grips with a scale of this size might simply gravitate toward a middle or neutral point.

More than likely a large ten point survey scale will not have any weights assigned to it. This creates another problem: how do we know a rating of a “4” means roughly the same to two different people? The lack of consistency here means the resulting data will not be reliable. Unfortunately, the use of a ten point scale can undermine the validity of your data from the very start. You can avoid biased data by using a better, balanced scale. Best practice suggests using a forced choice survey scale of approximately six items. A scale of this size allows a range of response, but it also provides clear and distinct criteria or ‘weight’ for each scale point.

Is the survey customizable?

Yes! You can select from our professional survey content and/or we collaborate with you to determine the most appropriate topics and questions for your survey. An NBRI survey ensures that the most meaningful data is collected.

What about open-ended survey questions?

Some customers want to ask one or more general, open-ended questions at the end of the survey. This can prove beneficial, especially when management drives the content of the survey without employee and/or customer input through interviews. The qualitative data from open-ended questions can be used to improve future surveys, as well as provide insight into topics management chose to omit from the present study.

What are the main issues I should consider when developing a guest survey?

When developing a guest survey, consider what deployment methods will allow you to quickly collect guest opinions. Response time is critical to ensure repeat or return business. If guests are unhappy with their stay or their experience at any point in the trip, it is highly likely they will not return or share their negative experience with others. Ultimately, this means the ability to engage in service recovery is particularly critical. For this reason, data should be collected while the guest is still on premises.

A second element to consider is confidentiality. If, for example, a guest has a particularly unpleasant experience with an employee at the front desk, can they be confident that same employee will not handle their complaint? Giving guests the confidence their comments and/or complaints will be addressed appropriately is paramount. Having a trusted and dedicated manager on standby to oversee the guest survey could be one way of building that confidence, or perhaps using an external vendor to manage the data to give clients extra confidence.

These considerations can shape the way data is collected. An online survey is a low cost, convenient, and discreet way of collecting feedback from guests.

What does a good customer survey look like?

A good, well-designed customer survey allows organizations to compare themselves against others and understand the items that have the greatest influence on customer experience.

One of the most important things a customer survey should have is the capability to compare your organization with your competitors: Are our customers happier with our products than our competitor’s products? Compared to other companies, how responsive are we? Understanding your performance can give you a solid frame of reference for improvement initiatives.

A customer survey should assess both causal (and independent) variables like specific elements of the customer experience, and big picture (or dependent) variables, like overall satisfaction or willingness to recommend the company to others. Developing a survey that has an appropriate mix of these questions allows you to understand what issues have the greatest impact on customer experience, and what areas need to be targeted to ensure maximum improvement.

A customer survey is your opportunity to find out if customers would respond well to a new product/service offering. If you have internal service standards, you will be able to determine if you are meeting those standards. A good customer survey will be able to provide answers to these questions and many more.

What does the survey look like?

The heading of the first page of the survey includes your logo, company name, and survey name.

Beneath the heading is an opening paragraph from a Senior Executive, the President, or the CEO about the process and why the survey is being conducted.

Following this introduction are instructions for completing the survey, and any demographic questions you may want to ask.

Then, the survey questions begin.

What factors should I consider when developing a customer service survey?

One of the first things to consider when developing a customer service survey is what you will do with the data when you collect it. Each question should have a purpose, and this purpose should link specifically to measuring or monitoring overall levels of service and/or guiding the organization’s direction and future growth.

Unclear questions can create confusion and inflate customer expectations about what existing product offerings should be. For example, asking if a client would be interested in a specific service if the “price was decreased by 20%,” may lead customers to believe that price decreases are on the immediate horizon. Failure to decrease prices can then create frustration, disappointment, or alienation of existing clients.

Thus, one of the first things to consider when developing a Customer Service Survey is whether you will be able to take reasonable action based on the feedback received from the survey. The second most important element to consider is how your performance compares to that of other organizations in your industry. This is called benchmarking. By understanding your relative position to others, you will be able to evaluate your competitive position and develop a strategy to highlight points of differentiation to your customer base.

What is the general layout of a survey?

The heading of the first page of the survey includes your logo, company name, and survey name.

Beneath the heading is an opening paragraph from a Senior Executive, the President, or the CEO about the process and why the survey is being conducted.

Following this introduction are instructions for completing the survey, and any demographic questions you may want to ask.

Then, the survey questions begin.

Trusted by thousands of businesses since 1982

acushnetacushnet fifth-bankfifth-bank toshibatoshiba ciscocisco audiaudi brightviewbrightview rainbirdrainbird ericssonericsson boeingboeing peterbiltpeterbilt