On March 4, 2020, Carmel Dibner from Applied Marketing Science (AMS) presented her talk entitled Unearthing the Unexpected: The power of machine learning at The Quirk’s Event in Brooklyn, NY.
In her talk, Carmel discussed advances in machine learning technology that allow researchers to gain insights for innovation much quicker and less expensively than ever before. The technology, developed in collaboration with researchers in MIT, enables companies to collect insights from existing text-based content – whether public or private – in a matter of a few short weeks. AMS has been employing this revolutionary algorithm, published in the Journal of Marketing Science by Professor John Hauser and Artem Timoshenko, to help our clients gather the insights they need. During the talk, Carmel discussed several case studies where AMS utilized the technique in the consumer, B2B, and medical spaces.
In addition to detailing the machine’s powerful ability to gather insights, Carmel shared a recent quantitative experiment conducted at AMS. The objective of the experiment was to test a novel way of prioritizing customer insights.
To test the new approach, AMS asked 200 respondents to rate the verbatim text associated with the insights uncovered through machine learning on two metrics 1) importance and 2) relatability. In traditional research, we ask respondents to rate insights statements on 1) importance and 2) satisfaction. We employed the traditional way of evaluating needs as the experimental control with 200 respondents.
The experiment showed that there is greater variability in the scores respondents provide when you ask them to rate verbatims (when available) on relatability vs. traditional insights statements measured on performance. In the pilot, this new technique produced data with a wider spread that would allow researchers to more easily observe differences between segments. The technique can uniquely be utilized with machine learning, as it enables the ability to trace each insight back to a specific verbatim which succinctly details the customer insights in their own words.
Our approach can be applied to achieve many different research objectives. For example, in consumer markets we’ve used the technique to understand important unmet needs around purchasing kitchen appliances. In the B2B space, we’ve used the method to help a snowplow company identify hidden opportunities for innovation. And, in the medical space we’ve used the technique to understand potential ways to improve in-home medical devices for diabetes.
Attendees at the Quirk’s Event were eager to learn and apply the novel qualitative and quantitative methods to their own research questions.
Wondering how you can use this method to find insights within your own company or industry? Contact Carmel at cdibner@ams-inc.com.
Tags: Conferences , Machine Learning/AI