Best Practices for Designing a Survey

Download a PDF Version of this Article

In a perfect world, you’d know exactly what your customers want and could serve it right up to them without hesitation. Unfortunately, things don’t quite work this way, so you’ve got to use the tools you have available to find out more about your customers. And there’s no better tool for this purpose than a survey.

A well-written and formatted survey will generate high response rates and provide you with quality information that can be use to improve just about any aspect of your business. In this White Paper, we’ll examine some best practices you can use to build a great survey.

Determine Your Objectives

The first step in designing a survey is to consider what kind of information you hope to glean from the survey and how you intend to put this information to use. Knowing your objectives will make it far easier to write your questions, decide which people you need to target, and to take action with the results. A few very broad based survey objectives might be to:

Better understand how customers feel about your sales process. Was the process easy for them? Were they satisfied with the delivery of the product? What factors led to their making a purchase with you? Would they purchase from you again? If yes or no, then why?

Find any pain points in your relationship with your customer. Have they contacted you for support? How did they do so? Were they happy with that support? Are they happy with the available training and information you provide them? Why do they love you? Why are they leaving you for another vendor?

Learn how they feel about your products. Are they satisfied with your product? How do they use it? Do they have suggestions for improvement? What kind of future enhancements interest them? Have they tried other vendors who provide similar products?

Again, these are very broad objectives, but they should give you an idea of the kind of information you can gather through the use of a survey.

Target Your Survey

Once you’ve decided why you’re doing the survey, you should determine your target audience. Putting your survey in front of the right people is key to getting a high response rate and getting the info you need. Knowing your audience will also make it easier to write your questions.

Who are the respondents? You’ll want to think about which customers, users or prospects you need to hear from to get the information you need. Should the survey only be to customers who made a purchase on your website? Or should it go to everyone who’s made a purchase with you through any means? It depends, of course, on the kind of information you hope to collect from the survey.

What are their experiences? As you narrow down the target audience for your survey, consider what kind of interaction they’ve had with your company, as you should try to limit your questions to those they can answer based on that previous experience. As an example, in a survey about interaction on your website, a question about your phone support would likely be out of place (even if it’s relevant to some of the respondents).

Create Your Questions

Now that you know both your goals and who you intend to target with your survey, you’re ready to create your questions. Remember that in order to collect enough feedback to get a valid sample for your survey, it’s important to get a high enough response rate and ensure that respondents complete your questionnaire. The clarity and conciseness of your questions, questionnaire length, number of items per page, and the way you group your questions can all contribute to a better experience for the respondent. This will lead to a higher completion rate and increase the likelihood that people will choose to take your future surveys.

Make sure each question is clear & concise. Keep your questions as short as possible without making them so short that they’re hard to understand. If each question on a page begins with the same phrase, considering using that phrase as the beginning text on the page so that it doesn’t have to be repeated as part of each question. Introductory text that asks, “How likely is it that…” can be followed by questions like “you would use our service again?” or “you would recommend us to a friend?”

There should be no ambiguity around what a question means. If a question is complex in some way, you should provide a brief explanation or provide some clarifying examples within the question itself (in parentheses). Be sure to use terms that will be familiar to your respondent, and avoid technical jargon.

Be careful with acronyms. Don’t use acronyms without an explanation of what the acronym means. If you use an acronym several times, spell it out the first time and place the acronym in parentheses. If I was creating a survey that asked my respondents several questions about their ISP, I’d want to write my first question as “Which Internet Service Provider (ISP) do you currently use?” I could then follow-up with questions like “Are you happy with your ISP?”

Avoid stating questions in the negative. It can be confusing if a phrase is negative and the respondent has to say, “yes,” in order to confirm that negative statement.

Make sure your questions & answers are mutually exclusive. If you have different categories of content, make sure they are distinct from one another. This can also refer to answers that use numerical ranges. For example, a question that asks the age of your respondent should not have overlapping answers like 18-24, 24-30, 30-38, 38-45. Then the 24, 40, and 38 year olds wouldn’t know which group to choose. The answers should instead be written as 18-24, 25-30, 31-38, 39-45, etc.

Be sure not to ask leading questions that suggest a certain answer to your respondents. An example of this would be to state a conclusion in the question and then ask for feedback (“We just redesigned our website to become a leading destination…”). If you present a list of options which allow the respondent to select one or more, be sure that the most common or important items do not appear at the top of that list. A possible way to group such options without bias would be to list them alphabetically.

It’s important that you feel confident the respondents will understand your questions and be able to give you their honest answers. If you are not sure that all respondents will be able to answer each question, be sure to include ‘N/A’ or “Don’t Know” as options. If you ask sensitive questions like age or income range, offer ‘Prefer Not to State’ as an option.

Organize Your Questions

Start with interesting questions to get your respondents immediately engaged in your survey. Then proceed on to specific question areas. If you’re surveying your respondents on a topic that involved a series of steps, try to have the questions mirror that process. I.e., for a training survey, start with questions about registration, then proceed to instructor quality and course content.

You can also group the sections of your survey according to topic. For example, in a Post-Event survey, questions could be grouped by Vendor Exhibits, Program Content and Location. Each section should have a bolded name so the respondents know where they are in the process.

Adjust the Length of Your Survey

There is a fine line between collecting answers on a whole range of interesting topics and generating survey fatigue for your respondents, which could result in them quitting your survey mid-stream. If a survey takes longer than an average of 15 to 20 minutes to complete, then it may be too long (though that doesn’t mean a survey can never be longer than 20 minutes).

And even though you want to keep your survey reasonably short, you should never trade clarity for time if you have to choose between them. If you can’t get clear and actionable information from your survey because you only asked vague questions, then it doesn’t matter how many people complete the survey since the data you’ll have collected won’t be all that useful.

This can also relate to the number of questions per page. Here the trade-off is between the necessity of scrolling on a single page that includes numerous questions and the number of pages a respondent must click-through to complete the survey. If you are using the same rating scale for a group of related subjects, then you could create a single question that allows multiple ratings instead of writing separate questions for each one.

That said, you want to be careful about using too many multiple-rating questions, as longer ones (or numerous such questions) can look intimidating to a respondent.

If you follow these best practices, your survey will provide you much higher value - as I noted at the beginning of this document, a well-written and formatted survey will generate high response rates and give you access to great information that you can use to improve just about any aspect of your business.

Ready to build your own survey?  Go to your account and get started!

Download a PDF Version of this Article