Surveys, be they for consumer or B2B marketing research, are all about choices. Sometimes it is best to force the respondent to pick one choice, other times we as survey creators need to give the respondent more room to breathe. This in short means allowing them to select multiple options. The graphic below represents a single choice option, which sometimes is easiest, but it also eliminates the respondent’s ability to provide a more robust answer.

The author of the survey had three ‘challenges’ in their mindset, but was kind enough to offer an ‘other’ response option. I wonder if they read the post on the ‘Almighty Other’? Yet I couldn’t help but feel that something was missing.

Upon reflection what was missing was the opportunity as a survey participant to select more than one challenge. In the realm of Big Data, as multi-faceted as this construct is, it is likely that marketers are facing more than one challenge in leveraging the data they are collecting. I could have selected all three of the challenges listed and added a few more in the comment section.

From the perspective of questionnaire design, this topic could have been better served in one of a few ways. First, they could have listed this as a multiple response question, followed by a single select. This approach allows the researcher to gain a picture of the broader set of issues and then drill down to the single most important item for consideration. This technique is also used in measuring the saliency of value propositions, product features or understanding the nature of the competitive set.

Alternatively they could have formulated this as a ranking problem. This approach asks the respondent to rank the concerns according to their perspective. It provides the researcher with ordinal data, which although not as robust as metric data, it is still actionable in the survey data analysis phase.

The key question to consider is should respondents be unnecessarily limited in the data they provide or should they be given the option of providing more robust responses, which ultimately is the goal of quantitative market research. I recommend the latter, because you can always reduce data, but seldom can you inflate it.