As part four of the pre-MRMW speaker interview series, I got to ask Tom Anderson of Anderson Analytics some questions. He goes into a great real world example of how text analytics can be applied to improve the survey taker’s experience and deliver better insights. Sounds like a win/win to me. Here’s our conversation.
RM: What are you most excited to share about with us at MRMW and why? What difference could it make for your audience if they were to implement what you’ll be talking about?
TA: I’ll be speaking about my favorite topic, text analytics. As you may know, since 2005 my firm has been specializing in how to best leverage unstructured (text) data within marketing research. There are so many applications for this, and while social media seems to be top of mind for many I think there are huge opportunities in many other areas as well including survey research.
For instance, just last night I was visiting with my parents. Ironically my father who is 88 was trying to take an online customer satisfaction survey regarding a purchase he had recently made. It was a relatively simple product, but the survey was rather long and asked several unnecessary questions. At his age he has little patience for the computer never mind long winded surveys. I felt a bit embarrassed about it as he knows I’m in the market research business. I think the only reason he completed the survey was because I was there and helped him through it. He said, “Is there a way to tell them this survey is just too long?”
There was in fact one open ended question which he could easily answer. Our research with surveys like this has shown that we can actually do a very good job predicting several of the closed ended variables from the comment question using text analytics. I think one of the many benefits someone would get from implementing a text analytics program in cases like this is to not only better leverage the valuable insights from these rich open ended questions, but also allow them to ask fewer questions thus increasing response rates and creating a better customer and respondent experience.
RM: Great example of how text analytics can be a win/win both for those answering the survey and for those receiving data. I look forward to hearing more. Here’s another question for you, then: what are you most excited to hear about from others at MRMW? Why?
TA: Well I think perhaps at this conference more than most others I speak at, I look forward to hearing what the audience has to say. I understand there will be many there from Procter & Gamble. I’ve always admired P&G for not only being pioneers in new areas of research, but also for doing it right and not sacrificing proper methodological rigor.
Before we had developed the OdinText software platform, and were still focusing mainly on full service text analytics consulting in market research, I had spoken to P&G folks who were also experimenting with text analytics software vendors of the time. Like many of the early pioneers on the client side they hadn’t had the best experience with what was then available. I think we’ve come a long way in the past 6-7 years, not just on the technology side, but as importantly on the best practices on how to best utilize the software and techniques. So I look forward to interesting questions and discussions with members of the audience before, during and after my talk.