Online Surveys - Avoiding the Pitfalls

Online surveys are a fantastic way to get quality feedback from the people who matter most: your prospects and customers. The primary reason to conduct an online survey is to learn more about your customers’ reactions and perceptions. Survey information can apply to present-day scenarios as well as possible future ones. Surveys let you test the waters before you actually invest time, energy and money into change.

Before you even begin to think about the details of a survey, you need to lay down the basics. Don’t just plan to survey an audience for the hell of it — have a reason! Define what you’re trying to learn and then determine exactly what information you’ll need in order to glean your answer(s).

Surveys, unlike quick polls, allow you to ascertain more in-depth information. For example, a survey can help you learn more about your participants’…

• Demographics
• Their reasons for coming to your site
• What they like/dislike about your site

According to Michael Wexler, Director of Research for E-Dialog, an online survey company, some of the many other issues you need to resolve early on are:

• What if the respondent has a question? Who can they turn to for technical concerns or feedback about the survey process?

• Will you provide alternative response channels if the respondent doesn’t want to fill out an online survey?

• How do you plan on using the data you gather, and to what extent do you plan on disclosing this to your participants?

Creating an Online Survey

Once you have decided to go the survey route you pretty much have three choices in terms of creation: 1) write your own survey script in-house; 2) use an off-the-shelf ASP solution; or 3) hire a professional market research firm such as E-Dialog, SurveySite, or BetaResearch. If you’re really attempting to generate unbiased, empirical results, the most appropriate route is to hire a professional firm. As Mr. Wexler puts it, “Survey experts spend years learning careful wording of questions, how to write an invitation, how to handle privacy, and how to control a slew of variables which can all make your survey results non-interpretable.”

Writing your own survey using a survey tool can open up a significant can of worms. Legitimate market research is an incredibly complex field, and a poorly-worded or distributed survey will do nothing but waste your time and money. That being said, if you cannot afford to hire a professional market research firm, there are a variety of outstanding ASP surveys produced by firms that can also and provide assistance in helping you craft a survey. ASP companies include SurveyTrends, Zoomerang, and Insight Express.

Writing A Solid Survey

If you do choose to write your own survey rather than work with a professional company, here are some quick pointers you can use to get started down the right path:

• Your survey must be concise and carefully crafted to gather exactly the information you want. If the survey has too many extraneous questions, a respondent can lose patience and stop filling it out, or worse, fill it in with bogus information just to get through it.

• Questions should be consciously written either “close-ended” or “open-ended.” Use closed-ended questions if you know exactly what you’re seeking feedback on. The object here is to gather quantitative data on a set range of key questions.
  “Yes/No” answers, multiple choice, and “Rate on a scale of 1 to 5″ are examples of close-ended questions.
  If you’re looking for more subjective feedback, use open-ended questions. Open-ended questions cannot be answered in just one word or selection. Instead, open-ended questions such as “What do you think about…?” or “Describe any improvements…” encourage free-flowing opinions. Open-ended questions may reveal information you hadn’t even ever thought to ask.

• Assume nothing! For example, just because YOU think that the shopping experience on your site is confusing, doesn’t mean that your customers do. Your question should be, “Is the shopping experience confusing?” NOT “Please check all the areas in which the shopping experience is confusing.” You don’t KNOW what your customers think about the shopping experience - that’s why you’ve doing a survey! So don’t allow your assumptions to color your questions.

• Speaking of coloring your questions, similiarly, avoid bias which can significantly skew your results. Open-ended questions are different than leading questions which can intentionally or unintentionally lead a respondent to answer a certain way. You may get survey results that sound good to you, but in all honesty would probably not be replicated if presented in an unbiased fashion.

Remember, the more *accurate and truthful* information surveys can reveal, the better you’ll be able to service current customers and reel in new ones. And that’s the name of the game right?

Related Links:

E-Dialog
SurveyTrends
SurveySite
BetaResearch
Zoomerang
Insight Express

 back to top