Online Survey Tools -- Valuable When Used Properly

More and more businesses and organizations are using third-party online tools such as SurveyMonkey for their marketing research surveys.  And with good reason -- they bring several advantages, from cost savings to speed.  

As a marketing research company, we are big proponents of online surveys for these very same reasons.  If we can obtain the information our clients need, charge them less for doing so, and give them their survey results in short order, why wouldn't we use that approach to survey research?  

However, for all the value automated platforms deliver, they are NOT a turnkey solution to the entire marketing survey process. 

Any successful marketing research survey process has three main steps:

Step #1 - Designing valid and effective questions

Step #2 - Gathering and tabulating data

Step #3 - Properly analyzing the survey's results

There is no question that online survey tools enable users to accomplish Step #2, the gathering and tabulating of data.

What isn't mentioned with nearly the same emphasis is that online survey platforms aren't always as much help when it comes to Steps #1 and #3. And, in the end, these two steps in the process are more important than automating the gathering and tabulating of data (Step #2). As alluring and as much a novelty as online tools may be, if survey questions aren't properly developed and survey results aren't properly analyzed, the ease with which the data is collected is a moot point.

What do we mean by Step #1, designing valid and effective questions? 

• Ensure that the survey questions are not worded in such a way as to lead or bias the respondent
• Use words that even the least-informed respondent would understand
• Pay attention to the order of questions: make sure that a question asked early in the survey doesn't give the respondent unintended information they can use to answer a later question 
• Don't ask respondents questions that demand too much recall of past events for which their recollection may well be inaccurate
• Resist the temptation to ask lots of open-ended survey questions; such questions are difficult to tabulate and can frustrate respondents when they have to repeatedly type replies
• Begin the survey with relatively insignificant, simple questions; use these early survey questions as a time for the respondent to become familiar with the format of a particular online survey

Step #1 also is about survey length. A lengthy online survey will adversely impact the completion rate.  Respondents will either not begin the survey, or start the survey and abandon mid-survey, or give haphazard answers without giving adequate thought to what is being asked, just to finish the survey.

Regarding Step #3, the proper analysis of survey results, most online tools do little, if anything, to help you determine which survey results are important to addressing your objectives for conducting the survey. Before moving ahead with an internally-administered online survey, consider questions such as the following:

• Will there be an ability to determine which differences in survey results that may seem significant are actually well within the margin of error, and therefore basically meaningless?
• Will the best decisions be made in properly categorizing the varied responses to an open-ended survey question?
• Will someone be able to take all of the survey results and compile an easy-to-understand report to present to decision makers?

While we strongly urge you to consider online tools that can greatly reduce the cost of conducting a survey, we can't emphasize enough the importance of ensuring that time is taken to design a valid survey questionnaire on the front end and to also ensure that survey results are properly and effectively analyzed on the back end. The absence of such attention to these two important steps not only negates the benefits of the cost savings you might realize from using online survey tools,  but, more importantly, likely renders your survey results meaningless and creates the risk of making business decisions that are not based on accurate data.