Asking in the Age of Listening
Posted by Vovici Blog on Fri, Mar 04, 2011
NewMR is holding the second of its monthly virtual events, Listening 2011, on Monday and Tuesday of next week. The conference agenda looks remarkably pragmatic, avoiding some of the hyperbole that surrounds discussions of text analytics and social media listening at more traditional research conferences.
Both techniques are highly complementary to traditional survey research, offering new ways to field better surveys. Oftentimes, in fact, the results of social listening research will prompt more questions that you will want to explore through the rigor of a survey. As discussed yesterday, Best Buy’s open-ended feedback collection from employees frequently generates ideas for questions to be fielded back to those same employees.
When writing questions, turn to social listening to understand the words and concepts that consumers use to discuss the products and services being researched. Listen to the Voice of the Customer so that you can speak it. Your questions will be less likely to be misunderstood, giving you more accurate data.
Some text analytics vendors will say that questionnaires can now have many more open-ended questions, due to the ease with which they can be analyzed. This can be carried overboard – last night I was invited to a survey and the first page had eight open-ended questions. I closed the window without answering any of them; it just looked like too much work. Like me, most respondents find too many essay questions taxing. That said, with a large enough sample, a single verbatim question can provide you as much data as a complex grid, while reflecting what’s important to the respondent rather than what’s important to the survey author.
When it comes to conducting survey analysis, text analytics makes the verbatim question much more powerful. For ongoing surveys such as a weekly transactional survey or a quarterly relationship survey, recent responses might suggest new ways of thinking about the data. This can be used to modify the coding sheet or taxonomy, which can then be reapplied to all the past responses to update trend information. This is impossible with manual survey coding and can show you patterns in the data that you hadn’t realized were there before.
And, to bring it full circle, sometimes the quantitative nature of a survey raises questions best answered through social listening. Perhaps a product or service attribute scored very low, yet the questionnaire didn’t collect any further details about it. The natural question is: why were respondents dissatisfied with this attribute? Comments mined from social media can provide insight into those low scores.
Listening without asking is an audience, and talking without listening is a monologue. From the standpoint of the research industry, the Age of Listening has not yet arrived. Perhaps we should skip it altogether and go to the Age of Conversation.
See also: