By Annie Pettit, FMRIA


Yes, I yelled that. In a manner reminiscent of what mobile research went through, text analytics has had a minor presence at market research conferences for the last ten years at least. There has always been one or two vendors and one or two speakers diligently trying to convince the audience about the wonders of using machines and automation to read and code social media data. This year, an explosion of at least fifteen companies took to the IIeX stage to explain the unique processes and objectives that their systems could tackle. From detailed explanations of how sentiment and content coding is validated and how hierarchical, category relevant  taxonomies are built, anyone who wanted to get a thorough understanding of the intricacies of this method was like me in a candy store. We have finally entered the age of text analytics.

I know we normally think about innovation in terms of crazy complicated technology like text analytics but let’s not forget that innovation is not simply another word for tech. The opening speaker, Ari Popper, described his human approach to creating innovation which entailed asking hundreds of business people to create the most imaginative, fictional stories they could come up with, the crazier and the more unbelievable the better. These stories could then be analyzed to come up with wonderfully innovative ideas. From fun silly stories to innovation.

The second session which caught my attention in terms of being all about innovation and at the same time completely lacking in technology was Piero Procacci’s story telling workshop. If you’ve ever wanted to scream your face off at a professional conference and not be called crazy, this session was for you. Piero worked hard to make people feel comfortable in an uncomfortable situation, thereby giving people permission to be unusually creative and say the craziest things they could think of. And since I mentioned the screaming, may I offer my apologies to the people in the session next door who must have thought Donald Trump was holding his rally there.

Underpinning many of the sessions, even the theoretical sessions, was a theme of automation and machine learning. Whether people were talking about text analytics, survey design, mobile qual, or something else, so many of them were also talking about how automation was being used to streamline systems, speed up processes, and use baseline measurements from humans to plug into machines for scale. Ready or not, be ready for any part of your job that can be automated to be handed over to a machine. Which is great because that will give you a lot more time to think. And speaking of thinking…

I think everyone in the audience thoroughly enjoyed hearing from Elizabeth Merrick, Andrew Jeavons, and Elizabeth Loftus talk about the flaws of human memory. We’ve gotten tastes of these theories over the years as numerous people have shared their expertise about System 1 vs System 2 thinking, and how humans are irrational beings. This particular session expanded the discussion by adding memory to the list of human fallibilities. Loftus spoke about how easy it is to manipulate memories whether deliberately or accidentally. For instance, she described research in which she incorrectly reminded people of their earlier answer on a test and they did end up remembering an answer they didn’t give. It’s not that they were lying or deliberately trying to mislead but rather that they truly believed the suggestion. This is why Loftus recommends collecting recall information as soon as possible after the event and as soon as possible in an interview. Before they can be mislead.

It was a whirlwind two and half days with barely a break to think. I guess I’m ready for some automation.

Annie Pettit, PhD, CMRP, FMRIA is VP Data Awesomeness at Research Now, and Chief Research Officer at Peanut Labs. Annie is a research methodologist who specializes in data quality, sampling, and survey design, and is an invited speaker at marketing research conferences around the world. She won the MRIA Award of Outstanding Merit in 2014, Best Methodological Paper at ESOMAR in 2013, and the 2011 AMA David K. Hardin Award. She was named a Fellow of the MRIA at the 2016 MRIA national conference. She tweets and blogs as @LoveStats.