5 Tips for Creating an Online Survey

This post originally appeared on the American Express OPEN Forum.

 Would you be willing to take a brief survey after this call? What’s your zip code? How did you hear about us?

Each of these questions probably sounds familiar to you. We’ve all heard them when stuck on the phone with the cable company, or checking out at The Gap, or signing up for a newsletter.

That’s because knowing more about the customer is valuable to the companies you patronize.

Similarly, more information about your customers, including their demographics, preferences and sentiments on your product, can be extraordinarily useful to you as a small business owner or entrepreneur, if you know how to properly gather information and interpret results. Online surveys are one inexpensive and quick way to do this and get great feedback. Here are some steps to get started.

1. Define Your Objectives

First, it’s important to define the business objectives you are attempting to further by conducting a survey. Are you trying to decide whether or not to launch a new product line? Are you attempting to collect information that will bolster pitches to potential advertisers or investors? Are you assessing performance of your marketing campaigns? In order to limit survey length, you’ll have to resist the urge to ask everything that’s on your mind, so having a clear set of objectives will help you to determine which questions should make the cut.

2. Work Backwards

Once you’ve defined your objectives on a high level, you can start to take a more granular approach. Think about the information you’d like to glean from the survey results, and then reverse-engineer questions that address these issues. For example, if you are trying to determine whether or not your customers are satisfied with the level of service at your restaurant, make a list of the various elements of the customer experience — e.g., cleanliness, friendliness of staff — and then frame each individual element in question form. Make sure to avoid what Dr. Phil Garland, vice president of methodology for SurveyMonkey, refers to as “double-barreled” questions; for instance, “how friendly and knowledgeable were the customer services representatives?” Perhaps they were friendly but not knowledgeable, or vice versa, which could affect how a customer responds. The more focused and narrow the scope of your questions, the easier it will be for you to gain useful intelligence when interpreting the results.

3. Check For Bias

Since you have a sense of the answers you’d ideally like to get, it’s a common pitfall to write your questions in a leading manner, i.e., “The customer service representatives were friendly. Do you agree with this statement?” Garland says this can skew your results in a more positive direction, since people naturally tend towards politesse and tact. So, when working backwards to create your questions, take care to make sure that each of your questions is just that — a question — and not a statement. You might also consider randomizing the order of your questions and response choices.

4. Do a Test Drive

As is good practice in any written online medium, have someone with copy editing experience — or, at the least, a strong grasp on language — take a look at your survey once you have built it. What seems intuitive to you could very well be confusing to a reader. Additionally, having a guinea pig take your survey before you let it out into the wild means you have a chance to double check not only for logistical issues but also for how long it takes to complete. If it’s more than 12 minutes, you need to bring an editing eye to your masterpiece to whittle it down — Garland estimates that the optimal length is eight minutes. Additionally, while open-ended questions take longer to answer, they can yield fruitful insights. Melissa Kim, vice president of product for Minted, an online paper goods company, always makes sure to include at least one such question, noting that some of the most useful information she gets from customers is in the free-response areas. “If we start to see something popping up in those types of questions … within a couple of days we can have an entire assortment [of products] sourced,” Kim says. “It’s like absolute gold in there.”

5. Collect Results and Analyze Data

Depending on how you keep track of your customers, you may send out emails soliciting feedback, run interstitial ads on your site, or even utilize third-party sites like Mechanical Turk to target survey respondents. But whatever means you use, Garland suggests keeping your survey open for at least a week, since respondents at different times of day and days of the week may tend towards different answers.

Once the results are in, sites like SurveyMonkey will allow you to compare your results across segments. This is a good time to go back to your core objectives you identified in step 1 and parse the data you’ve collected by those criteria. For instance, Kim highlighted Minted’s “How did you hear about us?” question as a useful “point of triangulation.” By comparing responses based on acquisition source, she is able to gauge which customer acquisition sources are bringing in the most satisfied clients.

There are a lot of factors to consider here, so getting started may seem daunting. But if you are able to make meaningful business decisions, as Kim has, based on the information you gather, it’s worth the extra effort needed to design and execute your survey properly.

Have you learned anything that affected your business from conducting a survey? Let us know in the comments.

); ga('send', 'pageview');